From jrslepak at ccs.neu.edu  Mon Sep  1 11:03:57 2014
From: jrslepak at ccs.neu.edu (Justin R. Slepak)
Date: Mon, 1 Sep 2014 11:03:57 -0400 (EDT)
Subject: [racket] Package installation VERY SLOW
In-Reply-To: <bu060atupnd94fq4rqi97pbjbvvped7q8o@4ax.com>
Message-ID: <17851288.952211409583837453.JavaMail.root@zimbra>

I've also had the 0% CPU hang when trying to build docs for a package I'm working on. Only `raco setup' hangs though. If I use the "Scribble HTML" button in DrRacket, the documentation builds.

---
Justin Slepak
PhD student, Computer Science dept.

----- Original Message -----
From: George Neuner <gneuner2 at comcast.net>
To: users at racket-lang.org
Sent: Sun, 31 Aug 2014 07:44:31 -0400 (EDT)
Subject: Re: [racket] Package installation VERY SLOW


Hi Matthew,


On Sat, 30 Aug 2014 07:41:02 -0600, Matthew Flatt
<mflatt at cs.utah.edu> wrote:

>At Mon, 25 Aug 2014 21:42:08 -0400, George Neuner wrote:
>> Just upgraded to 6.0.1 from 5.3.4.  I know it's a big leap and 6.1 is
>> out already, but I tried 6.1 and quickly experienced several of those
>> non-reproducible crashes I've been reading about in the lists.
>
>Do you mean Evgeny's report on August 1, or did you have something else
>in mind? (I don't have the sense that there are widespread reports of
>problems with v6.1, so I may be missing something.)

I don't recall Evgeny's report offhand and I can't find it just now. I
have seen a different reports of issues with 6.1 in different places.

On Windows 7 I experienced hanging with 64-bit 6.1 - completely
unresponsive, 0% CPU.  This happened numerous times doing different
things: running package manager, editing, even just adjusting
preferences.  It seemed to be a GUI issue as raco worked from the
command line.  I had so much trouble with it that I didn't try very
much.  I uninstalled/reinstalled with fresh downloads a couple of
times before giving up.

32-bit 6.1 did not hang, but it took almost 40 minutes to install
[more than twice as long as 64-bit] and was so slow doing anything
that I was convinced it was broken.  Again I tried
uninstalling/reinstalling without anything changing significantly.

Then I read about the internal changes and decided to fall back to
6.0.1 until everything shakes out.


>If you have time to help and you can quickly provoke crashes --- even
>if you're not sure what causes it --- that sounds promising for
>debugging. I'll ask more about that off-list.

I got your note.  I'd like to help but right now I'm in the middle of
something and don't have a lot of time to muck around with it.

I do have VS2010 available, but Racket was hanging rather than
crashing, so the debugger didn't activate.  Unfortunately I didn't
think to attach to the hung process and see what was happening before
killing it.


>> 6.0.1 is behaving itself, but installing new packages from PLaneT or
>> migrating packages from the older version is taking HOURS ...
>> virtually all of it spent processing .scrbl files.
>
>It's probably no surprise to hear that it worked ok on my machine
>(about 1 minute) in a fresh v6.0.1 install.
>
>Can you say more about the ".scrbl" part? Is the process of installing
>a Planet package rebuilding documentation for the entire installation?
>Or is it getting stuck on just the documentation for the installed
>packages plus the main page and search index?

It's hard to tell exactly because I'm not very familiar with the
package install process, but it does get stuck "--- building
documentation ---" for long periods with no output from raco.

It seems to spend an inordinate amount of time "processing" scrbl
files that are included with the packages.  I don't mean rendering
them - raco shows that separately - but just "processing" them. 

It would be nice if raco could timestamp its output.

I can try to capture a procmon trace of a simple install if that would
be any help.  Ususally I find traces have too much information 8-(

George

____________________
  Racket Users list:
  http://lists.racket-lang.org/users


From dpavlov at ipa.nw.ru  Mon Sep  1 11:12:44 2014
From: dpavlov at ipa.nw.ru (Dmitry Pavlov)
Date: Mon, 01 Sep 2014 19:12:44 +0400
Subject: [racket] fl vs. unsafe-fl
Message-ID: <54048CEC.20109@ipa.nw.ru>

Hello,

I was wondering whether operations with flonums and flvectors work
slower than their unsafe- counterparts in intensive numerical tasks.

Here is an example: res = a+b*c

#lang racket

(require racket/flonum
          racket/unsafe/ops)

(define n 1000)
(define k 100000)

(define a (make-flvector n 1.0))
(define b (make-flvector n 2.0))
(define c (make-flvector n 3.0))
(define res (make-flvector n))

(define (test-perf)
   (for ((i (in-range k)))
     (for ((j (in-range n)))
       (flvector-set!
        res j (fl+ (flvector-ref a j)
                   (fl* (flvector-ref b j)
                        (flvector-ref c j)))))))

(define (test-perf-unsafe-ops)
   (for ((i (in-range k)))
     (for ((j (in-range n)))
       (flvector-set!
        res j (unsafe-fl+ (flvector-ref a j)
                          (unsafe-fl* (flvector-ref b j)
                                      (flvector-ref c j)))))))

(define (test-perf-unsafe-flvec)
   (for ((i (in-range k)))
     (for ((j (in-range n)))
       (unsafe-flvector-set!
        res j (fl+ (unsafe-flvector-ref a j)
                   (fl* (unsafe-flvector-ref b j)
                        (unsafe-flvector-ref c j)))))))

(define (test-perf-unsafe-all)
   (for ((i (in-range k)))
     (for ((j (in-range n)))
       (unsafe-flvector-set!
        res j (unsafe-fl+ (unsafe-flvector-ref a j)
                          (unsafe-fl* (unsafe-flvector-ref b j)
                                      (unsafe-flvector-ref c j)))))))

(time (test-perf))
(time (test-perf-unsafe-ops))
(time (test-perf-unsafe-flvec))
(time (test-perf-unsafe-all))


Result (Racket 6.1.0.2, Linux 64-bit):

cpu time: 1704 real time: 1704 gc time: 0
cpu time: 1692 real time: 1692 gc time: 0
cpu time: 736 real time: 738 gc time: 0
cpu time: 684 real time: 682 gc time: 0


It turns out that replacing arithmetical operations
on flonums with their unsafe- counterparts makes no difference,
unless flvector access is unsafe- too; and even then, making
flvector access unsafe makes a huge improvement, while acceleration
from unsafe ops is about 7%.

I can imagine that the Racket's JIT skips checking the types of
arguments of fl+ and fl* because it figures that since they come
from flvectors, they must be flonums. Am I correct?

If I am correct, then are unsafe-fl arithmetical operations important
at all? Because in numerical programs, nearly every flonum is either
a constant, or a value taken from flvector, or a function argument
that can be traced to the same.


Regards,

Dmitry

From greghendershott at gmail.com  Mon Sep  1 14:04:54 2014
From: greghendershott at gmail.com (Greg Hendershott)
Date: Mon, 1 Sep 2014 14:04:54 -0400
Subject: [racket] fl vs. unsafe-fl
In-Reply-To: <54048CEC.20109@ipa.nw.ru>
References: <54048CEC.20109@ipa.nw.ru>
Message-ID: <CAGspUn3NnZH0vb8JYdcnPVi6QcmRNDavuWRH+sdTSqOXccteWw@mail.gmail.com>

In some applications a 7% speedup in numeric operations is welcome.

I think the main issue with unsafe-fl ops is that they are unsafe.

The best thing about Typed Racket is that you can delegate that
problem. Not only do you not need to write unsafe-fl+, you don't even
need to write fl+. You can write +. Typed Racket will determine if +
can safely be replaced with unsafe-fl+. (Assuming you typed things as
Float a.k.a. Flonum, FlVector, &c.)

From mflatt at cs.utah.edu  Mon Sep  1 15:40:49 2014
From: mflatt at cs.utah.edu (Matthew Flatt)
Date: Mon, 1 Sep 2014 21:40:49 +0200
Subject: [racket] fl vs. unsafe-fl
In-Reply-To: <54048CEC.20109@ipa.nw.ru>
References: <54048CEC.20109@ipa.nw.ru>
Message-ID: <20140901194053.0EBB26501E3@mail-svr1.cs.utah.edu>

Yes, using `fl` operations can often provide much of the benefit of
`unsafe-fl` operations. The JIT knows than a `fl` operation always
produces a flonum, so no check or boxing is needed if the flonum is
immediately consumed by another `fl` operation. (Boxing is a bigger
effect than dynamic checks, and no boxing is needed within the loop for
any of the examples below.)

An `unsafe-flvector` operation avoids array-bounds checks in addition
to fixnum and flonum checks on the second and third arguments. I think
that's why there's a much bigger difference between the second and
third examples below.

At Mon, 01 Sep 2014 19:12:44 +0400, Dmitry Pavlov wrote:
> Hello,
> 
> I was wondering whether operations with flonums and flvectors work
> slower than their unsafe- counterparts in intensive numerical tasks.
> 
> Here is an example: res = a+b*c
> 
> #lang racket
> 
> (require racket/flonum
>           racket/unsafe/ops)
> 
> (define n 1000)
> (define k 100000)
> 
> (define a (make-flvector n 1.0))
> (define b (make-flvector n 2.0))
> (define c (make-flvector n 3.0))
> (define res (make-flvector n))
> 
> (define (test-perf)
>    (for ((i (in-range k)))
>      (for ((j (in-range n)))
>        (flvector-set!
>         res j (fl+ (flvector-ref a j)
>                    (fl* (flvector-ref b j)
>                         (flvector-ref c j)))))))
> 
> (define (test-perf-unsafe-ops)
>    (for ((i (in-range k)))
>      (for ((j (in-range n)))
>        (flvector-set!
>         res j (unsafe-fl+ (flvector-ref a j)
>                           (unsafe-fl* (flvector-ref b j)
>                                       (flvector-ref c j)))))))
> 
> (define (test-perf-unsafe-flvec)
>    (for ((i (in-range k)))
>      (for ((j (in-range n)))
>        (unsafe-flvector-set!
>         res j (fl+ (unsafe-flvector-ref a j)
>                    (fl* (unsafe-flvector-ref b j)
>                         (unsafe-flvector-ref c j)))))))
> 
> (define (test-perf-unsafe-all)
>    (for ((i (in-range k)))
>      (for ((j (in-range n)))
>        (unsafe-flvector-set!
>         res j (unsafe-fl+ (unsafe-flvector-ref a j)
>                           (unsafe-fl* (unsafe-flvector-ref b j)
>                                       (unsafe-flvector-ref c j)))))))
> 
> (time (test-perf))
> (time (test-perf-unsafe-ops))
> (time (test-perf-unsafe-flvec))
> (time (test-perf-unsafe-all))
> 
> 
> Result (Racket 6.1.0.2, Linux 64-bit):
> 
> cpu time: 1704 real time: 1704 gc time: 0
> cpu time: 1692 real time: 1692 gc time: 0
> cpu time: 736 real time: 738 gc time: 0
> cpu time: 684 real time: 682 gc time: 0
> 
> 
> It turns out that replacing arithmetical operations
> on flonums with their unsafe- counterparts makes no difference,
> unless flvector access is unsafe- too; and even then, making
> flvector access unsafe makes a huge improvement, while acceleration
> from unsafe ops is about 7%.
> 
> I can imagine that the Racket's JIT skips checking the types of
> arguments of fl+ and fl* because it figures that since they come
> from flvectors, they must be flonums. Am I correct?
> 
> If I am correct, then are unsafe-fl arithmetical operations important
> at all? Because in numerical programs, nearly every flonum is either
> a constant, or a value taken from flvector, or a function argument
> that can be traced to the same.
> 
> 
> Regards,
> 
> Dmitry
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users

From mb at mbtype.com  Mon Sep  1 21:20:05 2014
From: mb at mbtype.com (Matthew Butterick)
Date: Mon, 1 Sep 2014 18:20:05 -0700
Subject: [racket] How `scribble` resolves relative paths
Message-ID: <06C12393-DBFA-4961-BC81-3B9C6CB96317@mbtype.com>

I am confused by this statement in the docs vs. the actual behavior:

The docs for `image` say that "If path is a relative path, it is relative to the current directory, which is set by `raco setup` and `scribble` to the directory of the main document file." (I'm using `image` as an example but I've seen the same problem with CSS files.)

Suppose I have a project set up like this:

/project
    /scribblings
        main.scrbl
        sample.gif

And "main.scrbl" contains this:

#lang scribble/manual
@image["sample.gif"] 

If I do this:
> cd project
> cd scribblings
> scribble main.html

The HTML will render fine.

But if I do this:

> cd project
> scribble --dest-name doc scribblings/main.scrbl

I'll get this error:

open-input-file: cannot open input file
  path: /project/sample.gif

Thus, the relative path "sample.gif" is being resolved relative to the directory where I invoke the `scribble` command, not the "directory of the main document file", which in both cases seems like it should be  /project/scribblings (i.e., the directory containing "main.scrbl").

What am I misunderstanding?






[1] http://docs.racket-lang.org/scribble/base.html?q=image#%28def._%28%28lib._scribble%2Fbase..rkt%29._image%29%29
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140901/0edb3c2a/attachment.html>

From mflatt at cs.utah.edu  Tue Sep  2 00:59:02 2014
From: mflatt at cs.utah.edu (Matthew Flatt)
Date: Tue, 2 Sep 2014 06:59:02 +0200
Subject: [racket] How `scribble` resolves relative paths
In-Reply-To: <06C12393-DBFA-4961-BC81-3B9C6CB96317@mbtype.com>
References: <06C12393-DBFA-4961-BC81-3B9C6CB96317@mbtype.com>
Message-ID: <20140902045905.B477B65018C@mail-svr1.cs.utah.edu>

I think the docs are wrong. While `raco setup` does set the current
directory, `scribble` does not.

It's better to use `define-runtime-path` instead of relying on the
current directory, since that composes better.

At Mon, 1 Sep 2014 18:20:05 -0700, Matthew Butterick wrote:
> I am confused by this statement in the docs vs. the actual behavior:
> 
> The docs for `image` say that "If path is a relative path, it is relative to 
> the current directory, which is set by `raco setup` and `scribble` to the 
> directory of the main document file." (I'm using `image` as an example but I've 
> seen the same problem with CSS files.)
> 
> Suppose I have a project set up like this:
> 
> /project
>     /scribblings
>         main.scrbl
>         sample.gif
> 
> And "main.scrbl" contains this:
> 
> #lang scribble/manual
> @image["sample.gif"] 
> 
> If I do this:
> > cd project
> > cd scribblings
> > scribble main.html
> 
> The HTML will render fine.
> 
> But if I do this:
> 
> > cd project
> > scribble --dest-name doc scribblings/main.scrbl
> 
> I'll get this error:
> 
> open-input-file: cannot open input file
>   path: /project/sample.gif
> 
> Thus, the relative path "sample.gif" is being resolved relative to the 
> directory where I invoke the `scribble` command, not the "directory of the main 
> document file", which in both cases seems like it should be  
> /project/scribblings (i.e., the directory containing "main.scrbl").
> 
> What am I misunderstanding?
> 
> 
> 
> 
> 
> 
> [1] 
> http://docs.racket-lang.org/scribble/base.html?q=image#%28def._%28%28lib._scribb
> le%2Fbase..rkt%29._image%29%29____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users

From publicityifl at gmail.com  Tue Sep  2 03:23:30 2014
From: publicityifl at gmail.com (publicityifl at gmail.com)
Date: Tue, 02 Sep 2014 00:23:30 -0700 (PDT)
Subject: [racket] Third call for papers, IFL 2014
Message-ID: <54057072.2369b40a.7eb4.ffff977b@mx.google.com>

Hello,

Please, find below the third call for papers for IFL 2014.
The submission page is now open. The submission date has been
delayed to Sep. 8 2014 anywhere on the world.

Please forward these to anyone you think may be interested.
Apologies for any duplicates you may receive.

best regards,
Jurriaan Hage

---

CALL FOR PAPERS

26th SYMPOSIUM ON IMPLEMENTATION AND APPLICATION OF FUNCTIONAL LANGUAGES - IFL 2014

NORTHEASTERN UNIVERSITY/BOSTON, USA

OCTOBER 1-3, 2014

http://ifl2014.github.io

We are pleased to announce that the 26th edition of the IFL series
will be held at Northeastern University in Boston, USA. The symposium
will be held from 1st to 3rd of October 2014.

Scope
-----

The goal of the IFL symposia is to bring together researchers actively
engaged in the implementation and application of functional and
function-based programming languages.  IFL 2014 will be a venue for
researchers to present and discuss new ideas and concepts, work in
progress, and publication-ripe results related to the implementation
and application of functional languages and function-based
programming.

Following the IFL tradition, IFL 2014 will use a post-symposium review
process to produce the formal proceedings. All participants of IFL
2014 are invited to submit either a draft paper or an extended
abstract describing work to be presented at the symposium. At no time
may work submitted to IFL be simultaneously submitted to other venues;
submissions must adhere to ACM SIGPLAN's republication policy:

    http://www.sigplan.org/Resources/Policies/Republication

The submissions will be screened by the program committee chair to
make sure they are within the scope of IFL, and will appear in the
draft proceedings distributed at the symposium. Submissions appearing
in the draft proceedings are not peer-reviewed publications. Hence,
publications that appear only in the draft proceedings do not count as
publication for the ACM SIGPLAN republication policy. After the
symposium, authors will be given the opportunity to incorporate the
feedback from discussions at the symposium and will be invited to
submit a revised full article for the formal review process. From the
revised submissions, the program committee will select papers for the
formal proceedings considering their correctness, novelty,
originality, relevance, significance, and clarity.

Submission Details
------------------

Submission deadline draft papers:                          September 8 
Notification of acceptance for presentation:               September 10
Early registration deadline:                               September 11
Late registration deadline:                                September 17 
Submission deadline for pre-symposium proceedings:         September 24
26th IFL Symposium:                                        October 1-3 
Submission deadline for post-symposium proceedings:        December 15
Notification of acceptance for post-symposium proceedings: January  31 2015
Camera-ready version for post-symposium proceedings:       March    15 2015 

Prospective authors are encouraged to submit papers or extended
abstracts to be published in the draft proceedings and to present them
at the symposium. All contributions must be written in English. Papers
must adhere to the standard ACM two columns conference format. For the
pre-symposium proceedings we adopt a 'weak' page limit of 12
pages. For the post-symposium proceedings the page limit of 12 pages
is firm. A suitable document template for LaTeX can be found at:

    http://www.acm.org/sigs/sigplan/authorInformation.htm
    
Papers should be submitted online at https://easychair.org/conferences/?conf=ifl2014
    

Topics
------

IFL welcomes submissions describing practical and theoretical work as
well as submissions describing applications and tools in the context
of functional programming. If you are not sure whether your work is
appropriate for IFL 2014, please contact the PC chair at
samth at cs.indiana.edu. Topics of interest include, but are not limited
to:

???  language concepts
???  type systems, type checking, type inferencing
???  compilation techniques
???  staged compilation
???  run-time function specialization
???  run-time code generation
???  partial evaluation
???  (abstract) interpretation
???  metaprogramming
???  generic programming
???  automatic program generation
???  array processing
???  concurrent/parallel programming
???  concurrent/parallel program execution
???  embedded systems
???  web applications
???  (embedded) domain specific languages
???  security
???  novel memory management techniques
???  run-time profiling performance measurements
???  debugging and tracing
???  virtual/abstract machine architectures
???  validation, verification of functional programs
???  tools and programming techniques
???  (industrial) applications

Peter Landin Prize
------------------

The Peter Landin Prize is awarded to the best paper presented at the
symposium every year. The honoured article is selected by the program
committee based on the submissions received for the formal review
process. The prize carries a cash award equivalent to 150 Euros.

Programme committee
-------------------

Sam Tobin-Hochstadt, Indiana University (Chair)
Rinus Plasmeijer, Radboud University Nijmegen (Co-Chair)
Atze Dijkstra, Utrecht University
Colin Runciman, University of York
Graham Hutton, University of Nottingham
Mary Sheeran, Chalmers University of Technology
Patricia Johann, Appalachian State University
Matthew Fluet, Rochester Institute of Technology
Josef Svenningsson, Chalmers University of Technology
Ma??gorzata Biernacka, University of Wroclaw
Peter Achten, Radboud Univerity Nijmegen
Laura Castro, University of A Coru??a
Hai Paul Liu, Intel Labs
Kathryn Gray, Cambridge University
Lars Bergstrom, Mozilla Research
Lindsey Kuper, Indiana University
Nicolas Wu, Oxford
T. Stephen Strickland, University of Maryland
Xavier Clerc, INRIA


Venue
-----

The 26th IFL will be held in association with the College of Computer
and Information Science at Northeastern University. It can be reached
quickly and easily by public transport.

From publicityifl at gmail.com  Tue Sep  2 03:23:46 2014
From: publicityifl at gmail.com (publicityifl at gmail.com)
Date: Tue, 02 Sep 2014 00:23:46 -0700 (PDT)
Subject: [racket] Third call for papers, IFL 2014
Message-ID: <54057082.2369b40a.7eb4.ffff9794@mx.google.com>

Hello,

Please, find below the third call for papers for IFL 2014.
The submission page is now open. The submission date has been
delayed to Sep. 8 2014 anywhere on the world.

Please forward these to anyone you think may be interested.
Apologies for any duplicates you may receive.

best regards,
Jurriaan Hage

---

CALL FOR PAPERS

26th SYMPOSIUM ON IMPLEMENTATION AND APPLICATION OF FUNCTIONAL LANGUAGES - IFL 2014

NORTHEASTERN UNIVERSITY/BOSTON, USA

OCTOBER 1-3, 2014

http://ifl2014.github.io

We are pleased to announce that the 26th edition of the IFL series
will be held at Northeastern University in Boston, USA. The symposium
will be held from 1st to 3rd of October 2014.

Scope
-----

The goal of the IFL symposia is to bring together researchers actively
engaged in the implementation and application of functional and
function-based programming languages.  IFL 2014 will be a venue for
researchers to present and discuss new ideas and concepts, work in
progress, and publication-ripe results related to the implementation
and application of functional languages and function-based
programming.

Following the IFL tradition, IFL 2014 will use a post-symposium review
process to produce the formal proceedings. All participants of IFL
2014 are invited to submit either a draft paper or an extended
abstract describing work to be presented at the symposium. At no time
may work submitted to IFL be simultaneously submitted to other venues;
submissions must adhere to ACM SIGPLAN's republication policy:

    http://www.sigplan.org/Resources/Policies/Republication

The submissions will be screened by the program committee chair to
make sure they are within the scope of IFL, and will appear in the
draft proceedings distributed at the symposium. Submissions appearing
in the draft proceedings are not peer-reviewed publications. Hence,
publications that appear only in the draft proceedings do not count as
publication for the ACM SIGPLAN republication policy. After the
symposium, authors will be given the opportunity to incorporate the
feedback from discussions at the symposium and will be invited to
submit a revised full article for the formal review process. From the
revised submissions, the program committee will select papers for the
formal proceedings considering their correctness, novelty,
originality, relevance, significance, and clarity.

Submission Details
------------------

Submission deadline draft papers:                          September 8 
Notification of acceptance for presentation:               September 10
Early registration deadline:                               September 11
Late registration deadline:                                September 17 
Submission deadline for pre-symposium proceedings:         September 24
26th IFL Symposium:                                        October 1-3 
Submission deadline for post-symposium proceedings:        December 15
Notification of acceptance for post-symposium proceedings: January  31 2015
Camera-ready version for post-symposium proceedings:       March    15 2015 

Prospective authors are encouraged to submit papers or extended
abstracts to be published in the draft proceedings and to present them
at the symposium. All contributions must be written in English. Papers
must adhere to the standard ACM two columns conference format. For the
pre-symposium proceedings we adopt a 'weak' page limit of 12
pages. For the post-symposium proceedings the page limit of 12 pages
is firm. A suitable document template for LaTeX can be found at:

    http://www.acm.org/sigs/sigplan/authorInformation.htm
    
Papers should be submitted online at https://easychair.org/conferences/?conf=ifl2014
    

Topics
------

IFL welcomes submissions describing practical and theoretical work as
well as submissions describing applications and tools in the context
of functional programming. If you are not sure whether your work is
appropriate for IFL 2014, please contact the PC chair at
samth at cs.indiana.edu. Topics of interest include, but are not limited
to:

???  language concepts
???  type systems, type checking, type inferencing
???  compilation techniques
???  staged compilation
???  run-time function specialization
???  run-time code generation
???  partial evaluation
???  (abstract) interpretation
???  metaprogramming
???  generic programming
???  automatic program generation
???  array processing
???  concurrent/parallel programming
???  concurrent/parallel program execution
???  embedded systems
???  web applications
???  (embedded) domain specific languages
???  security
???  novel memory management techniques
???  run-time profiling performance measurements
???  debugging and tracing
???  virtual/abstract machine architectures
???  validation, verification of functional programs
???  tools and programming techniques
???  (industrial) applications

Peter Landin Prize
------------------

The Peter Landin Prize is awarded to the best paper presented at the
symposium every year. The honoured article is selected by the program
committee based on the submissions received for the formal review
process. The prize carries a cash award equivalent to 150 Euros.

Programme committee
-------------------

Sam Tobin-Hochstadt, Indiana University (Chair)
Rinus Plasmeijer, Radboud University Nijmegen (Co-Chair)
Atze Dijkstra, Utrecht University
Colin Runciman, University of York
Graham Hutton, University of Nottingham
Mary Sheeran, Chalmers University of Technology
Patricia Johann, Appalachian State University
Matthew Fluet, Rochester Institute of Technology
Josef Svenningsson, Chalmers University of Technology
Ma??gorzata Biernacka, University of Wroclaw
Peter Achten, Radboud Univerity Nijmegen
Laura Castro, University of A Coru??a
Hai Paul Liu, Intel Labs
Kathryn Gray, Cambridge University
Lars Bergstrom, Mozilla Research
Lindsey Kuper, Indiana University
Nicolas Wu, Oxford
T. Stephen Strickland, University of Maryland
Xavier Clerc, INRIA


Venue
-----

The 26th IFL will be held in association with the College of Computer
and Information Science at Northeastern University. It can be reached
quickly and easily by public transport.

From dbastos at toledo.com  Tue Sep  2 11:45:23 2014
From: dbastos at toledo.com (Daniel Bastos)
Date: Tue, 2 Sep 2014 12:45:23 -0300
Subject: [racket] missing solution 20.1.1 ex:sem-funcs
Message-ID: <CAB08-qr3dsphx=M875-g8EBNkrdFN3-BLkvTQnytrZ4i-5uUvg@mail.gmail.com>

A candidate for a solution.

Exercise 20.1.1. Assume the Definitions window in DrScheme contains
(define (f x) x). Identify the values among the following expressions:

(1) (cons f empty)
(2) (f f)
(3) (cons f (cons 10 (cons (f 10) empty)))

Explain why they are values and why the remaining expressions are not
values.

Solution. First we consider (1). Here's the relevant part of the
grammar.

  <lst> = empty | (cons <val> <lst>)

  <val> = <boo> | <sym> | <num> | empty | <lst>
  | <var> (names of defined functions)
  | <prm>

So (1) is a value because it is <lst> of the form (cons <var> empty),
where f is <var> because it is a defined function.

Let's consider (3) before (2). (3) is a list of either <var> or <num>,
so (3) is <val> as well.

Now we consider (2). Here's the relevant part of the grammar.

  <exp>       =    <var>
  | <prm>
  | (<exp> <exp> ...<exp>)

So (2) is (<var> <var>), a list of expressions, which is an <exp>, not
a <val>. So (2) is not a value, although it's a legal expression.

From dpavlov at ipa.nw.ru  Tue Sep  2 11:53:37 2014
From: dpavlov at ipa.nw.ru (Dmitry Pavlov)
Date: Tue, 02 Sep 2014 19:53:37 +0400
Subject: [racket] fl vs. unsafe-fl
In-Reply-To: <CAGspUn3NnZH0vb8JYdcnPVi6QcmRNDavuWRH+sdTSqOXccteWw@mail.gmail.com>
References: <54048CEC.20109@ipa.nw.ru>
 <CAGspUn3NnZH0vb8JYdcnPVi6QcmRNDavuWRH+sdTSqOXccteWw@mail.gmail.com>
Message-ID: <5405E801.9050808@ipa.nw.ru>

Greg,

 > The best thing about Typed Racket is that you can delegate that
> problem. Not only do you not need to write unsafe-fl+, you don't even
> need to write fl+. You can write +. Typed Racket will determine if +
> can safely be replaced with unsafe-fl+. (Assuming you typed things as
> Float a.k.a. Flonum, FlVector, &c.)

I agree with you; however, I think I am in an unlucky situation.
I actually have to write code that works with flonums or (optionally)
with extflonums. As Extflonums are not part of the numeric tower,
Typed Racket will not allow me to use "+" instead of "extfl+".

Regards,

Dmitry


From dbastos at toledo.com  Tue Sep  2 12:05:18 2014
From: dbastos at toledo.com (Daniel Bastos)
Date: Tue, 2 Sep 2014 13:05:18 -0300
Subject: [racket] missing solution 20.1.2 ex:syn-funcs
Message-ID: <CAB08-qqmrq3jesw_GzLbmR0K6PA8WBbtscezT4TGbooL373hZA@mail.gmail.com>

Exercise 20.1.2. Argue why the following sentences are legal
definitions:

(define (f x) (x 10))

(define (f x) f)

(define (f x y) (x 'a y 'b))

Solution. The relevant part of the grammar is the following.

  <def> = (define (<var> <var> ...<var>) <exp>)
        | (define <var> <exp>)
        | (define-struct <var> (<var> <var> ...<var>))

(*) First definition

The LHS is a list of <var>, since we find f and x as members of the
list and they're both variables. The RHS is a list consisting of a
<var> and a <num>. That composes an <exp>, satisfying the first form
of <def>. In other words,

  (define (<var> <var>) (<var> <num)) =
  (define (<var> <var>) (<exp> <exp>) =
  (define (<var> <var>) <exp>).

Therefore it's a legal <def>.

(*) Second definition

Same LHS as the previous, so we need only check the RHS which is a
<var>. <var> is a valid form of <exp>, so we have

  (define (<var> <var>) <var>) =
  (define (<var> <var>) <exp>).

Therefore it's a legal <def>.

(*) Third definition

The LHS is (<var> <var> <var>), while the RHS is (<var> <sym> <var>
<sym>), so we have

  (define (<var> <var> <var>) (<var> <sym> <var> <sym>)) =
  (define (<var> <var> <var>) (<exp> <exp> <exp> <exp>)) =
  (define (<var> <var> <var>) <exp>).

Therefore it's a legal <def>.

From samth at cs.indiana.edu  Tue Sep  2 12:05:45 2014
From: samth at cs.indiana.edu (Sam Tobin-Hochstadt)
Date: Tue, 2 Sep 2014 12:05:45 -0400
Subject: [racket] fl vs. unsafe-fl
In-Reply-To: <5405E801.9050808@ipa.nw.ru>
References: <54048CEC.20109@ipa.nw.ru>
 <CAGspUn3NnZH0vb8JYdcnPVi6QcmRNDavuWRH+sdTSqOXccteWw@mail.gmail.com>
 <5405E801.9050808@ipa.nw.ru>
Message-ID: <CAK=HD+ZQZkCuMLzVNnygi9ASgeFsHXqj708wMqMqVzAHqtQXbA@mail.gmail.com>

On Tue, Sep 2, 2014 at 11:53 AM, Dmitry Pavlov <dpavlov at ipa.nw.ru> wrote:
>
>
>> The best thing about Typed Racket is that you can delegate that
>>
>> problem. Not only do you not need to write unsafe-fl+, you don't even
>> need to write fl+. You can write +. Typed Racket will determine if +
>> can safely be replaced with unsafe-fl+. (Assuming you typed things as
>> Float a.k.a. Flonum, FlVector, &c.)
>
>
> I agree with you; however, I think I am in an unlucky situation.
> I actually have to write code that works with flonums or (optionally)
> with extflonums. As Extflonums are not part of the numeric tower,
> Typed Racket will not allow me to use "+" instead of "extfl+".


But `fl+` and `unsafe-fl+` also do not work on Extflonums, so you
should be able to use TR and then just replace uses of unsafe-fl+ with
+ as appropriate, and your extfl+ will still be there.

Sam

From dpavlov at ipa.nw.ru  Tue Sep  2 12:13:50 2014
From: dpavlov at ipa.nw.ru (Dmitry Pavlov)
Date: Tue, 02 Sep 2014 20:13:50 +0400
Subject: [racket] fl vs. unsafe-fl
In-Reply-To: <CAK=HD+ZQZkCuMLzVNnygi9ASgeFsHXqj708wMqMqVzAHqtQXbA@mail.gmail.com>
References: <54048CEC.20109@ipa.nw.ru>
 <CAGspUn3NnZH0vb8JYdcnPVi6QcmRNDavuWRH+sdTSqOXccteWw@mail.gmail.com>
 <5405E801.9050808@ipa.nw.ru>
 <CAK=HD+ZQZkCuMLzVNnygi9ASgeFsHXqj708wMqMqVzAHqtQXbA@mail.gmail.com>
Message-ID: <5405ECBE.4010402@ipa.nw.ru>


>> I agree with you; however, I think I am in an unlucky situation.
>> I actually have to write code that works with flonums or (optionally)
>> with extflonums. As Extflonums are not part of the numeric tower,
>> Typed Racket will not allow me to use "+" instead of "extfl+".
>
>
> But `fl+` and `unsafe-fl+` also do not work on Extflonums, so you
> should be able to use TR and then just replace uses of unsafe-fl+ with
> + as appropriate, and your extfl+ will still be there.

Right. I was just dreaming about Typed Racket replacing "+"
with "fl+" or "extfl+" basing on compile-time type inference.

Regards,

Dmitry


From samth at cs.indiana.edu  Wed Sep  3 08:30:24 2014
From: samth at cs.indiana.edu (Sam Tobin-Hochstadt)
Date: Wed, 3 Sep 2014 08:30:24 -0400
Subject: [racket] fl vs. unsafe-fl
In-Reply-To: <5405ECBE.4010402@ipa.nw.ru>
References: <54048CEC.20109@ipa.nw.ru>
 <CAGspUn3NnZH0vb8JYdcnPVi6QcmRNDavuWRH+sdTSqOXccteWw@mail.gmail.com>
 <5405E801.9050808@ipa.nw.ru>
 <CAK=HD+ZQZkCuMLzVNnygi9ASgeFsHXqj708wMqMqVzAHqtQXbA@mail.gmail.com>
 <5405ECBE.4010402@ipa.nw.ru>
Message-ID: <CAK=HD+bapMcvZ78ELGP6wkVY4cU5YbtG03CDnS28AZ+ZvYoQaw@mail.gmail.com>

On Tue, Sep 2, 2014 at 12:13 PM, Dmitry Pavlov <dpavlov at ipa.nw.ru> wrote:
>
>>> I agree with you; however, I think I am in an unlucky situation.
>>> I actually have to write code that works with flonums or (optionally)
>>> with extflonums. As Extflonums are not part of the numeric tower,
>>> Typed Racket will not allow me to use "+" instead of "extfl+".
>>
>>
>>
>> But `fl+` and `unsafe-fl+` also do not work on Extflonums, so you
>> should be able to use TR and then just replace uses of unsafe-fl+ with
>> + as appropriate, and your extfl+ will still be there.
>
>
> Right. I was just dreaming about Typed Racket replacing "+"
> with "fl+" or "extfl+" basing on compile-time type inference.

Ah, ok. I don't think we'll do that, since we want to keep the same
Racket runtime semantics. However, if there was a `generic-+` with
that behavior (which you could then rename to `+` :) then TR could
optimize as you desired (maybe).

Sam

From dbastos at toledo.com  Wed Sep  3 09:29:32 2014
From: dbastos at toledo.com (Daniel Bastos)
Date: Wed, 3 Sep 2014 10:29:32 -0300
Subject: [racket] missing solution 20.2.1 ex:arrows-dd
Message-ID: <CAB08-qqeTXAweBU_nHuQ06DKeVD9f1Lr==S7WnVgQu9JvWH5tA@mail.gmail.com>

A candidate for a solution.

Exercise 20.2.1. Explain the following classes of functions:

1. (number  ->  boolean),
2. (boolean symbol  ->  boolean),
3. (number number number  ->  number),
4. (number  ->  (listof number)), and
5. ((listof number)  ->  boolean).

Solution.

  (1) consumes a number and produces a boolean;

  (2) consumes a boolean and a symbol and produces a boolean;

  (3) consumes three numbers and produces a number;

  (4) consumes a number and produces a list of numbers;

  (5) consumes a list of numbers and produces a boolean.

From dbastos at toledo.com  Wed Sep  3 09:30:40 2014
From: dbastos at toledo.com (Daniel Bastos)
Date: Wed, 3 Sep 2014 10:30:40 -0300
Subject: [racket] missing solution 20.2.2 ex:fancy-contracts
Message-ID: <CAB08-qo--gMj5qs2qPjfZWcoNhtdjrREqH3moWtLMUgCY9ZtdA@mail.gmail.com>

A candidate for a solution.

Exercise 20.2.2. Formulate contracts for the following functions:

1. sort, which consumes a list of numbers and a function that consumes
two numbers (from the list) and produces a boolean; sort produces a
list of numbers.

2. map, which consumes a function from numbers to numbers and a list
of numbers; it also produces a list of numbers.

3. project, which consumes a list of lists of symbols and a function
from lists of symbols to symbols; it produces a list of symbols.

Solution.

sort: (listof number) (number number -> boolean) -> (listof number)

map: (number -> number) (listof number) -> (listof number)

project: (listof (listof symbol)) ((listof symbol) -> symbol) -> (listof symbol)

From dbastos at toledo.com  Wed Sep  3 09:31:25 2014
From: dbastos at toledo.com (Daniel Bastos)
Date: Wed, 3 Sep 2014 10:31:25 -0300
Subject: [racket] missing solution 20.2.3 ex:filter-contract
Message-ID: <CAB08-qp1HT6jVwVuuO4wkmL7pLtkUKo41kBgQyQOjwCWrQPHgw@mail.gmail.com>

A candidate for a solution.

;; Exercise 20.2.3. Use filter1 to develop a function that
;; consumes a list of symbols and extracts all those that
;; are not equal to 'car. Give filter1's corresponding contract.

;; Solution.

(define (filter1 rel-op alon t)
  (cond
    [(empty? alon) empty]
    [else (cond
            [(rel-op (first alon) t)
             (cons (first alon)
                   (filter1 rel-op (rest alon) t))]
            [else
             (filter1 rel-op (rest alon) t)])]))

;; f: (listof symbol) -> (listof symbol)
;; consumes a (listof symbol), producing a (listof symbol) without
;; any occurrence of 'car
(define (f ls-of-sym)
  (local ((define (cmp sym1 sym2)
            (not (symbol=? sym1 sym2))))
    (filter1 cmp ls-of-sym 'car)))

(check-expect (f (list 'a 'b 'car 'x)) (list 'a 'b 'x))
(check-expect (f (list 'car 'car 'x 'car)) (list 'x))

;; The contract of filter1
;; filter1: (symbol symbol -> boolean) (listof symbol) symbol -> (listof symbol)

From gil.martinez242 at gmail.com  Wed Sep  3 12:01:21 2014
From: gil.martinez242 at gmail.com (Gilbert Martinez)
Date: Wed, 3 Sep 2014 11:01:21 -0500
Subject: [racket] Issue reading bytes from TCP port
Message-ID: <CAOyHz4vWHwGqeyJ7PBGXHskGb24dP1O4M0H-WauiSEaDtQ5qpg@mail.gmail.com>

I'm having issues reading the response from a TCP server.  Specifically,
any attempt to read the input port does not terminate.  I've used
port->bytes, read-byte, and read-line. In all cases the effect is the same.

I thought that if there were no bytes available on an input port that the
read attempt would just return <eof>.  Is this some kind of exclusivity
issue?  Can I not read from the port until the connection with the server
is closed or something?

My traffic monitor shows that the server is receiving the request and
responding (here is the exchange):

   Time Src. & Dst. Type Function Code Data  908.875    10.10.10.10:49967<- TCP
Req. 0x04 [0x000 - 0x027]     00 01 00 00 00 06 0B 04    00 3B 00 03
908.90510.10.10.10:49967-> TCP Resp. 0x04 [0x000 - 0x027]     00 01 00 00
00 09 0B 04    06 00 00 00 00 00 00


Here is the code I am using to query:

#lang racket


(define (hex-bytes->bytes hex-bytes-list)
  (define (hex-byte->dec-byte hex-byte)
    (string->number (string-append "#x" ((if (symbol? hex-byte)
                                             symbol->string
                                             number->string)
                                         hex-byte))))
  (apply bytes (for/list ((hex-byte (in-list hex-bytes-list)))
                 (hex-byte->dec-byte hex-byte))))

(define-values (c-in c-out) (tcp-connect "10.10.10.11" 502))

(write-bytes (hex-bytes->bytes '(00 01 00 00 00 06 0B 04 00 3B 00 03))
c-out)
(flush-output c-out)


I run the code above and execute the following in the interactions pane:


Welcome to DrRacket, version 6.1.0.5--2014-08-25(32ae3f8/a) [3m].
Language: racket [custom].
12
>(for ((byte (in-bytes (port->bytes c-in))))
  (printf "~x " byte))


The entry above above hangs on port->bytes.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140903/84b5391e/attachment.html>

From matthias at ccs.neu.edu  Wed Sep  3 12:18:07 2014
From: matthias at ccs.neu.edu (Matthias Felleisen)
Date: Wed, 3 Sep 2014 18:18:07 +0200
Subject: [racket] Issue reading bytes from TCP port
In-Reply-To: <CAOyHz4vWHwGqeyJ7PBGXHskGb24dP1O4M0H-WauiSEaDtQ5qpg@mail.gmail.com>
References: <CAOyHz4vWHwGqeyJ7PBGXHskGb24dP1O4M0H-WauiSEaDtQ5qpg@mail.gmail.com>
Message-ID: <ACACF8E1-3481-4246-92D0-0837377ACBFE@ccs.neu.edu>


Have you tried closing the ports? -- Matthias



On Sep 3, 2014, at 6:01 PM, Gilbert Martinez wrote:

> I'm having issues reading the response from a TCP server.  Specifically, any attempt to read the input port does not terminate.  I've used port->bytes, read-byte, and read-line. In all cases the effect is the same.
> 
> I thought that if there were no bytes available on an input port that the read attempt would just return <eof>.  Is this some kind of exclusivity issue?  Can I not read from the port until the connection with the server is closed or something?
> 
> My traffic monitor shows that the server is receiving the request and responding (here is the exchange):
> 
> Time	Src. & Dst.	Type	Function Code	Data
> 908.875   	10.10.10.10:49967<-	TCP Req.	0x04	[0x000 - 0x027]     00 01 00 00 00 06 0B 04    00 3B 00 03
> 908.905	10.10.10.10:49967->	TCP Resp.	0x04	[0x000 - 0x027]     00 01 00 00 00 09 0B 04    06 00 00 00 00 00 00
> 
> Here is the code I am using to query:
> 
> #lang racket
> 
> 
> (define (hex-bytes->bytes hex-bytes-list)
>   (define (hex-byte->dec-byte hex-byte)
>     (string->number (string-append "#x" ((if (symbol? hex-byte)
>                                              symbol->string
>                                              number->string)
>                                          hex-byte))))
>   (apply bytes (for/list ((hex-byte (in-list hex-bytes-list)))
>                  (hex-byte->dec-byte hex-byte))))
> 
> (define-values (c-in c-out) (tcp-connect "10.10.10.11" 502))
> 
> (write-bytes (hex-bytes->bytes '(00 01 00 00 00 06 0B 04 00 3B 00 03)) c-out)
> (flush-output c-out)
> 
> 
> I run the code above and execute the following in the interactions pane:
> 
> Welcome to DrRacket, version 6.1.0.5--2014-08-25(32ae3f8/a) [3m].
> Language: racket [custom].
> 12
> >(for ((byte (in-bytes (port->bytes c-in))))
>   (printf "~x " byte))
> 
> 
> The entry above above hangs on port->bytes.
> ____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140903/b340eb67/attachment.html>

From neil at neilvandyke.org  Wed Sep  3 12:23:10 2014
From: neil at neilvandyke.org (Neil Van Dyke)
Date: Wed, 03 Sep 2014 12:23:10 -0400
Subject: [racket] Issue reading bytes from TCP port
In-Reply-To: <CAOyHz4vWHwGqeyJ7PBGXHskGb24dP1O4M0H-WauiSEaDtQ5qpg@mail.gmail.com>
References: <CAOyHz4vWHwGqeyJ7PBGXHskGb24dP1O4M0H-WauiSEaDtQ5qpg@mail.gmail.com>
Message-ID: <5407406E.9010603@neilvandyke.org>


Gilbert Martinez wrote at 09/03/2014 12:01 PM:
>
> I run the code above and execute the following in the interactions pane:
>
>
>     Welcome to DrRacket, version 6.1.0.5--2014-08-25(32ae3f8/a) [3m].
>     Language: racket [custom].
>     12
>     >(for ((byte (in-bytes (port->bytes c-in))))
>       (printf "~x " byte))
>
>
> The entry above above hangs on port->bytes.

Unless "in-bytes" is doing more magic than I feared, I believe that here 
"port->bytes" is a procedure that reads all the input to the end-of-file 
(which won't happen until the TCP connection is closed), before 
"in-bytes" or "for" or anyone else sees anything from the port.

"port->bytes" is a convenience, mainly for reading files.  You instead 
probably want to read one byte at a time from the port, or (a little 
harder) to do block reads of available bytes from the port.

BTW, I think this is another example of how "for" and friends are 
confusing to people.  I still think that people are better off first 
learning how to do things using named-"let" and mutually-recursive 
procedures *before* being introduced to "for". (My biggest objection is 
that people introduced to "for" first then end up shoehorning 
control-flow into "for" even when it doesn't fit.  But I'm saving this 
new kind of anecdote, about syntactic sugar obscuring evaluation model, 
as additional evidence of "for"'s crimes against humanity.)

Neil V.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140903/5af31ec5/attachment-0001.html>

From david.vanderson at gmail.com  Wed Sep  3 17:48:36 2014
From: david.vanderson at gmail.com (David Vanderson)
Date: Wed, 03 Sep 2014 17:48:36 -0400
Subject: [racket] Issue reading bytes from TCP port
In-Reply-To: <CAOyHz4vWHwGqeyJ7PBGXHskGb24dP1O4M0H-WauiSEaDtQ5qpg@mail.gmail.com>
References: <CAOyHz4vWHwGqeyJ7PBGXHskGb24dP1O4M0H-WauiSEaDtQ5qpg@mail.gmail.com>
Message-ID: <54078CB4.50207@gmail.com>

On 09/03/2014 12:01 PM, Gilbert Martinez wrote:
> I'm having issues reading the response from a TCP server. 
>  Specifically, any attempt to read the input port does not terminate. 
>  I've used port->bytes, read-byte, and read-line. In all cases the 
> effect is the same.
>
> I thought that if there were no bytes available on an input port that 
> the read attempt would just return <eof>.  Is this some kind of 
> exclusivity issue?  Can I not read from the port until the connection 
> with the server is closed or something?
This is not true.  Most functions like 'read-byte' block until there are 
bytes available to read.  You don't get <eof> until the port 
(connection) is closed.  If you want a non-blocking call, look at 
'read-bytes-avail!*'.

#lang racket

(define-values (in out) (make-pipe))

(thread
  (lambda ()
    (for ((i 5))
      (write-byte i out)
      (sleep 1))
    (close-output-port out)))

(for ((b in))
   (printf "~x " b))

'for' is pretty smart in my experience.  I believe in this case it's 
using a sequence from 'in-port'.  But in any case, you should see 
numbers print out before the port is closed.

Does this help?

Thanks,
Dave
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140903/638b41e8/attachment.html>

From clements at brinckerhoff.org  Wed Sep  3 19:00:00 2014
From: clements at brinckerhoff.org (John Clements)
Date: Wed, 3 Sep 2014 16:00:00 -0700
Subject: [racket] racket search: a compelling reason to use duckduckgo.com
Message-ID: <89769A17-173D-487C-8127-029CCA5D7EA1@brinckerhoff.org>

Once I have duckduckgo set as my default browser search engine, I can search

!racket add1

? and I get to <http://docs.racket-lang.org/search/index.html?q=add1>

not bad.  I think I?m ready to switch to this search engine permanently.

John



From jhemann at umail.iu.edu  Wed Sep  3 23:09:07 2014
From: jhemann at umail.iu.edu (Jason Hemann)
Date: Wed, 3 Sep 2014 23:09:07 -0400
Subject: [racket] [ANN] *Extended Deadline* CFP for 2014 Scheme and
 Functional Programming Workshop
Message-ID: <CACHJ6hivr2L2KXWQk2KoNcrLZrYE8sKOW6HPevyLJWvtMB_8vg@mail.gmail.com>

The paper submission deadline has been extended one week, to Friday,
Sept. 12 23:59 (UTC-12).

[Apologies for duplication from cross-postings.]

NEW DEADLINE: 12 September 2014, 23:59 (UTC-12)
WEBSITE: http://homes.soic.indiana.edu/jhemann/scheme-14/
LOCATION: Washington, DC (co-located with Clojure/conj)
DATE: 19 November 2014

The 2014 Scheme and Functional Programming Workshop is calling for
submissions.

Submissions related to Scheme and functional programming are welcome and
encouraged. Topics of interest include but are not limited to:

- Program-development environments, debugging, testing
- Implementation (interpreters, compilers, tools, benchmarks, etc)
- Syntax, macros, and hygiene
- Distributed computing, concurrency, parallelism
- Interoperability with other languages, FFIs
- Continuations, modules, object systems, types
- Theory, formal semantics, correctness
- History, evolution and standardization of Scheme
- Applications, experience and industrial uses of Scheme
- Education
- Scheme pearls (elegant, instructive uses of Scheme)

We also welcome papers related to dynamic or multiparadigmatic languages
and programming techniques.

Full papers are due 12 September 2014.
Authors will be notified by 10 October 2014.
Camera-ready versions are due 24 October 2014.
All deadlines are 23:59 (UTC-12, "Anywhere on Earth").

For more information, please see:
http://homes.soic.indiana.edu/jhemann/scheme-14/

Jason Hemann
Organizer, Scheme '14
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140903/0cd8602b/attachment.html>

From tarmo at cs.ioc.ee  Tue Sep  2 06:28:35 2014
From: tarmo at cs.ioc.ee (Tarmo Uustalu)
Date: Tue, 2 Sep 2014 13:28:35 +0300 (EEST)
Subject: [racket] ETAPS 2015 2nd call for papers
Message-ID: <alpine.DEB.1.10.1409021324350.4604@fookus.kybi>


******************************************************************

                   CALL FOR PAPERS: ETAPS 2015

18th European Joint Conferences on Theory And Practice of Software

                  London, UK,  11-18 April 2015

                    http://www.etaps.org/2015

******************************************************************

-- ABOUT ETAPS --

ETAPS is the primary European forum for academic and industrial
researchers working on topics relating to software science. ETAPS,
established in 1998, is a confederation of six main annual
conferences, accompanied by satellite workshops. ETAPS 2015 is the
eighteenth event in the series.


-- MAIN CONFERENCES (13-17 April) --

    * CC: Compiler Construction
        (PC chair Bj?rn Franke, University of Edinburgh, UK)
    * ESOP: European Symposium on Programming
        (PC chair Jan Vitek, Northeastern University, USA)
    * FASE: Fundamental Approaches to Software Engineering
        (PC chairs Alexander Egyed, Johannes Kepler U Linz, Austria,
         and Ina Schaefer, Technische Universit?t Braunschweig, Germany)
    * FOSSACS: Foundations of Software Science
        and Computation Structures
        (PC chair Andrew Pitts, University of Cambridge, UK)
    * POST: Principles of Security and Trust
        (PC chairs Riccardo Focardi, Universit? Ca' Foscari Venezia,
         Italy, and Andrew Myers, Cornell University, USA)
    * TACAS: Tools and Algorithms for
        the Construction and Analysis of Systems
        (PC chairs Christel Baier, Technische Univ Dresden, Germany,
         and Cesare Tinelli, The University of Iowa, USA)

TACAS '15 will host the 4rd Competition on Software Verification
(SV-COMP).


-- INVITED SPEAKERS --

    * Unifying speakers:
      Robert Harper (Carnegie Mellon University, USA)
      Catuscia Palamidessi (INRIA Saclay and LIX, France)

    * CC invited speaker:
      Keshav Pingali (University of Texas, USA)
    * FoSSaCS invited speaker:
      Frank Pfenning (Carnegie Mellon University, USA)
    * TACAS invited speaker:
      Wang Yi (Uppsala Universitet, Sweden)


-- IMPORTANT DATES --

    * 10 October 2014: Submission deadline for abstracts
    * 17 October 2014: Submission deadline for full papers
    * 3-5 December 2014: Author response period (ESOP and FoSSaCS only)
    * 19 December 2014: Notification of acceptance
    * 16 January  2015: Camera-ready versions due


-- SUBMISSION INSTRUCTIONS --

ETAPS conferences accept two types of contributions: research papers
and tool demonstration papers. Both types will appear in the
proceedings and have presentations during the conference.

ESOP and FoSSaCS accept only research papers. TACAS has more paper
categories (see http://www.etaps.org/2015/tacas).

A condition of submission is that, if the submission is accepted, one
of the authors attends the conference to give the presentation.

Submitted papers must be in English presenting original
research. They must be unpublished and not submitted for publication
elsewhere. In particular, simultaneous submission of the same
contribution to multiple ETAPS conferences is forbidden. The
proceedings will be published in the Advanced Research in Computing
and Software Science (ARCoSS) subline of Springer's Lecture Notes in
Computer Science series.

Papers must follow the formatting guidelines specified by Springer at

http://www.springer.de/comp/lncs/authors.html

and be submitted electronically in pdf through the EasyChair author
interface of the respective conference (HotCRP for ESOP).

Submissions not adhering to the specified format and length may be
rejected immediately.


- Research papers

FASE, FOSSACS and TACAS have a page limit of 15 pages for research
papers, whereas CC, POST allow at most 20 pages and ESOP 25 pages.

Additional material intended for the referees but not for publication
in the final version - for example, details of proofs - may be placed
in a clearly marked appendix that is not included in the page
limit. ETAPS referees are at liberty to ignore appendices and papers
must be understandable without them.

In addition to regular research papers, TACAS solicits also case study
papers (at most 15 pages).

Both TACAS and FASE solicit also regular tool papers (at most 15
pages).


- Tool demonstration papers

Submissions should consist of two parts:

* The first part, at most 4 pages, should describe the tool
    presented. Please include the URL of the tool (if available) and
    provide information that illustrates the maturity and robustness of
    the tool. (This part will be included in the proceedings.)

* The second part, at most 6 pages, should explain how the
    demonstration will be carried out and what it will show, including
    screen dumps and examples. (This part will be not be included in the
    proceedings, but will be evaluated.)

ESOP and FOSSACS do not accept tool demonstration papers.

TACAS has a page limit of 6 pages for tool demonstrations.


-- SATELLITE EVENTS (11-12 April, 18 April) --

Around 20 satellite workshops will take place before and after the
main conferences.


-- HOST CITY --

London, the capital city of England and the UK, is a leading global
city, with strengths in the arts, commerce, education, entertainment,
fashion, finance, healthcare, media, professional services, research
and development, tourism and transport all contributing to its
prominence. It is one of the world's leading financial centers and a
world cultural capital. It is the world's most-visited city as
measured by international arrivals and has the world's largest city
airport system measured by passenger traffic. In 2012, London became
the first city to host the modern Summer Olympic Games three times.


-- HOST INSTITUTION --

ETAPS 2015 is hosted by the School of Electrical Engineering and
Computer Science of the Queen Mary University of London.

The main campus is located in the Mile End area of the East End of
London.


-- ORGANIZERS

* General chairs: Pasquale Malacaria, Nikos Tzevelekos
* Workshops chair: Paulo Oliva


-- FURTHER INFORMATION --

Please do not hesitate to contact the organizers at
p.malacaria at qmul.ac.uk, nikos.tzevelekos at qmul.ac.uk.


From dbastos at toledo.com  Thu Sep  4 12:54:57 2014
From: dbastos at toledo.com (Daniel Bastos)
Date: Thu, 4 Sep 2014 13:54:57 -0300
Subject: [racket] missing solution 20.2.4 ex:fancy-contracts-poly
Message-ID: <CAB08-qrbvXik1gT7R_gcH+yUjCaY-bAZQF+T=LWFj7EB2DuMug@mail.gmail.com>

A candidate for a solution. (I'm not sure I'm correct regarding the
function project. The description says "a lists of lists" (which I
translate to (listof ITEM)) and "a function from lists to Xs", so I
wonder if the domain of this function-argument must be (listof ITEM)
or it could be more general such as (listof Y). I decided to imagine
that the domain of the function-argument would be items of the "list
of lists" and so I made it into (listof ITEM).

Thank you for your attention.

Exercise 20.2.4. Formulate general contracts for
the following functions:

1. sort, which consumes a list of items and a
function that consumes two items (from the list)
and produces a boolean; it produces a list of items.

2. map, which consumes a function from list items
to Xs and a list; it produces a list of Xs.

3. project, which consumes a list of lists and a
function from lists to Xs; it produces a list of Xs.

Compare with exercise 20.2.2.

Solution.

sort: (listof ITEM) (ITEM ITEM -> boolean) -> (listof ITEM)

map: ((listof ITEM) -> X) (listof ITEM) -> (listof X)

project: (listof (listof ITEM)) ((listof ITEM) -> (listof X)) -> (listof X)

(*) Comparison with exercise 20.2.2

For sort, we replaced number with X. So we generalized the type
of the list which we sort. The new procedure can sort a generic list.

For map, we generalized both the domain and co-domain of
the function consumed by map. Before, the consumed function
was only a mapping from numbers to numbers and now it is
a mapping from a generic domain ITEM to a generic co-domain X.

For project, we generalized the type of the list of lists. It was of
type symbol and now it can be a list of lists of any kind of item.

From johnbclements at gmail.com  Thu Sep  4 13:45:48 2014
From: johnbclements at gmail.com (John Clements)
Date: Thu, 4 Sep 2014 10:45:48 -0700
Subject: [racket] Best typed/racket representation for bitfield of width 125?
Message-ID: <CAAwfr=Wh0-vJib+tzaSpaTcsqHnrtMr3kirnp7w5Mjxp5vJdng@mail.gmail.com>

I want to represent a bitfield of 125 bits, using typed racket. I?ll be
combining them using bitwise operations, and checking for zero-ness. I
think that?s about it.

I could just use type Integer, but it seems like that would impose
additional checking, since the type system doesn?t know the size.  In
untyped racket I used a vector of four integers in the range 0-2^32, but
those don?t fit in typed Racket?s ?Fixnum? because it uses a signed
representation (and indeed, untyped racket may have been falling prey to
the same problem. Perhaps I should break it up into four integers in the
range 0-2^31? That seems weird, but maybe that?s my best choice? Or? I
could use a byte string.

Apologies if I?m missing something in the docs, here; I?ve spent a few
minutes searching, without turning anything up.

Thanks for any advice,

John
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140904/152fe003/attachment.html>

From greghendershott at gmail.com  Thu Sep  4 14:41:31 2014
From: greghendershott at gmail.com (Greg Hendershott)
Date: Thu, 4 Sep 2014 14:41:31 -0400
Subject: [racket] Best typed/racket representation for bitfield of width
	125?
In-Reply-To: <CAAwfr=Wh0-vJib+tzaSpaTcsqHnrtMr3kirnp7w5Mjxp5vJdng@mail.gmail.com>
References: <CAAwfr=Wh0-vJib+tzaSpaTcsqHnrtMr3kirnp7w5Mjxp5vJdng@mail.gmail.com>
Message-ID: <CAGspUn27Q7tVV19Os+KHX-Jebxjj++EiLFy1fxxhLNfYBv22wA@mail.gmail.com>

Although I haven't tried data/bit-vector in Typed Racket, could you do
something like the following?

#lang typed/racket

(require/typed data/bit-vector
  [#:opaque BitVector bit-vector?]
  [make-bit-vector (-> Index BitVector)]
  ... TO-DO ...)

(define bv (make-bit-vector 125))


On Thu, Sep 4, 2014 at 1:45 PM, John Clements <johnbclements at gmail.com> wrote:
> I want to represent a bitfield of 125 bits, using typed racket. I?ll be
> combining them using bitwise operations, and checking for zero-ness. I think
> that?s about it.
>
> I could just use type Integer, but it seems like that would impose
> additional checking, since the type system doesn?t know the size.  In
> untyped racket I used a vector of four integers in the range 0-2^32, but
> those don?t fit in typed Racket?s ?Fixnum? because it uses a signed
> representation (and indeed, untyped racket may have been falling prey to the
> same problem. Perhaps I should break it up into four integers in the range
> 0-2^31? That seems weird, but maybe that?s my best choice? Or? I could use a
> byte string.
>
> Apologies if I?m missing something in the docs, here; I?ve spent a few
> minutes searching, without turning anything up.
>
> Thanks for any advice,
>
> John
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>


From stamourv at ccs.neu.edu  Thu Sep  4 15:35:57 2014
From: stamourv at ccs.neu.edu (Vincent St-Amour)
Date: Thu, 04 Sep 2014 15:35:57 -0400
Subject: [racket] Best typed/racket representation for bitfield of width
	125?
In-Reply-To: <CAAwfr=Wh0-vJib+tzaSpaTcsqHnrtMr3kirnp7w5Mjxp5vJdng@mail.gmail.com>
References: <CAAwfr=Wh0-vJib+tzaSpaTcsqHnrtMr3kirnp7w5Mjxp5vJdng@mail.gmail.com>
Message-ID: <878ulzji02.wl%stamourv@ccs.neu.edu>

Would an fxvector of length 5 (to account for the tag bits) work?

If all you're doing is bitwise ops and comparisons, it should be easy to
convince TR that you're staying within fixnum range.

Vincent




At Thu, 4 Sep 2014 10:45:48 -0700,
John Clements wrote:
> 
> [1  <multipart/alternative (7bit)>]
> [1.1  <text/plain; UTF-8 (quoted-printable)>]
> I want to represent a bitfield of 125 bits, using typed racket. I?ll be
> combining them using bitwise operations, and checking for zero-ness. I
> think that?s about it.
> 
> I could just use type Integer, but it seems like that would impose
> additional checking, since the type system doesn?t know the size.  In
> untyped racket I used a vector of four integers in the range 0-2^32, but
> those don?t fit in typed Racket?s ?Fixnum? because it uses a signed
> representation (and indeed, untyped racket may have been falling prey to
> the same problem. Perhaps I should break it up into four integers in the
> range 0-2^31? That seems weird, but maybe that?s my best choice? Or? I
> could use a byte string.
> 
> Apologies if I?m missing something in the docs, here; I?ve spent a few
> minutes searching, without turning anything up.
> 
> Thanks for any advice,
> 
> John
> [1.2  <text/html; UTF-8 (quoted-printable)>]
> 
> [2  <text/plain; us-ascii (7bit)>]
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users

From steve at stevelloyd.net  Thu Sep  4 15:48:43 2014
From: steve at stevelloyd.net (Steve Lloyd)
Date: Thu, 4 Sep 2014 20:48:43 +0100
Subject: [racket] Snapshot builds
Message-ID: <F3438AEA-6E9D-44FB-B63C-677A8EA68D97@stevelloyd.net>

I recently ?discovered? (thanks to a tweet from @dkvansnickajr) the snapshot builds for ARM6 built on Raspbian. 
Previously I?d been recommending people try out Racket on RasPi via the source+built packages route, but realise they might baulk at the build times and this other route seems very attractive in this respect. Can I just sanity-check that this new approach is sensible?

Thanks.
 

From marco at neniu.org  Fri Sep  5 06:04:45 2014
From: marco at neniu.org (Marco Monteiro)
Date: Fri, 5 Sep 2014 11:04:45 +0100
Subject: [racket] Understanding error in use of submodule in my language
In-Reply-To: <20140822153851.1C28965018F@mail-svr1.cs.utah.edu>
References: <CALGOKDeD7se=j0z60NhqzWhonUDtpdrvQWiFpiOs+UkH8LQMjA@mail.gmail.com>
 <20140822153851.1C28965018F@mail-svr1.cs.utah.edu>
Message-ID: <CALGOKDcm60rHYHUg6iaBv2+3KqEP9XwLAPn+rTGPwuXKSgUgPQ@mail.gmail.com>

Thanks, Matthew.

I fixed it by not expanding the module* form in the local-expand,
i.e. I used (list #'module*) as the third argument to local-expand.

Cheers,
Marco


On Fri, Aug 22, 2014 at 4:38 PM, Matthew Flatt <mflatt at cs.utah.edu> wrote:

> I think this may be a problem in macro expansion with submodules.
>
> The `#%module-begin` from `racket/base` adds a submodule declaration
> for `configure-runtime` when it doesn't see an existing submodule
> declaration for `configure-runtime` --- but it looks for an existing
> declaration as an immediate one, before macro-expanding the module
> body.
>
> In your code, the `mute` wrapper hides the declaration of
> `configure-runtime`, so a conflicting one gets added. The tricky part
> is that this happens the second time that the `test` submodule is
> expanded:
>
>  * The overall module's body is forced via `local-expand`.
>
>  * As a part of that local expansion, the `test` submodule is expanded'
>    its expansion includes a declaration of a `configure-runtime`
>    submodule.
>
>  * The overall module's expanded body is wrapped back in
>    `#%module-begin`, which will re-expand all of the content, including
>    the `test` submodule.
>
>  * The `test` submodule this second time around has a `#%module-begin`
>    that is supposed to be bound to `#%plain-module-begin`, but it ends
>    up being captured[*] by `#%module-begin` from "bootstrap-lang.rkt",
>    so that the submodule's body is again local-expanded and again send
>    to `#%module-begin` from `racket/base` with a `mute` wrapper around
>    the `configure-runtime` sub-submodule.
>
> The [*] step seems wrong to me. I think this may be an example of a
> problem deep in the macro expander, and solving that is my next
> project.
>
>
> Since the repair (assuming that I have correctly diagnosed the cause)
> may be a while, do you have a workaround already? The macro name `mute`
> suggests that you may want to suppress the printing of expression
> results at the top-level of a module, and that might be as easy as
> expanding `#%module-begin` to `#%plain-module-begin`. Or maybe
> `syntax/wrap-modbeg` can help, or maybe the strategy used by
> `syntax/wrap-modbeg` would work better for your macro.
>
>
> At Thu, 21 Aug 2014 17:22:29 +0100, Marco Monteiro wrote:
> > Hello!
> >
> > I'm using Racket 6.1
> >
> > I'm trying to understand this error. I'm using submodules in a language I
> > defined. The language file is here:
> >
> > http://pastebin.com/7RiHHRmi
> >
> > Notice the commented out line 28.
> >
> > I use that language in the following code
> >
> > http://pastebin.com/Qa6RD9wF
> >
> > raco expanding this last file fails with
> >
> > module: duplicate submodule definition
> >   at: configure-runtime
> >   in: module
> >   context...:
> >
> /usr/share/racket/pkgs/compiler-lib/compiler/commands/expand.rkt:29:11:
> > loop
> >    /usr/share/racket/pkgs/compiler-lib/compiler/commands/expand.rkt:
> > [running body]
> >    /usr/share/racket/collects/raco/raco.rkt: [running body]
> >    /usr/share/racket/collects/raco/main.rkt: [running body]
> >
> > If I uncomment line 28 it no longer fails. If I use (module ...) instead
> of
> > (module* ...) it no longer fails independently of line 28 being commented
> > out or not.
> >
> > Trying to macro expand that in drracket gives me "Internal error: car:
> > contract violation ." error.
> >
> > What is happending here?
> >
> > Thanks,
> > Marco
> > ____________________
> >   Racket Users list:
> >   http://lists.racket-lang.org/users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140905/97a3894f/attachment-0001.html>

From acarrico at memebeam.org  Fri Sep  5 11:12:40 2014
From: acarrico at memebeam.org (Anthony Carrico)
Date: Fri, 05 Sep 2014 11:12:40 -0400
Subject: [racket] typed racket cps, state machines
Message-ID: <5409D2E8.2070903@memebeam.org>

Thanks for occasional help learning typed racket on IRC. I'm still
finding the pitfalls that await those of us used to dynamically typed
programs.

The other day I posted some cps patterns I figured out:
  http://pasterack.org/pastes/49759

Extending that to simple state machines that loop through various states
gets me into trouble. The simplest example is just looping through one
state, or not:

(define #:? (B0V B1V)
  (p0 (branch0 : (-> B0V))
      (branch1 : (-> B1V)))
  : (U B0V B1V)

  ;; This one is fine:
  (define (no-loop) : (U B0V B1V)
    (branch0))

  ;; insufficient type information to typecheck this one:
  (define (loop) : (U B0V B1V)
    (loop))

  ;; insufficient type information to typecheck this one too:
  (define (loop*) : (U B0V B1V)
    ((inst loop* B0V B1V)))

  ;; insufficient type information to typecheck this one too:
  (define (loop**) : (U B0V B1V)
    ((inst loop** (U B0V B1V) (U B0V B1V))))

  (no-loop)
  (loop)
  (loop*)
  (loop**))

I'm still thinking about this problem, but since I have distilled it to
the simplest cases, I thought I'd post. I have the intuition that this
has something to do with the nested instantiation of polymorphic types.
The unions would "obviously" (to a human) collapse, but I suppose the
type checker isn't picking up on this.

-- 
Anthony Carrico

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 181 bytes
Desc: OpenPGP digital signature
URL: <http://lists.racket-lang.org/users/archive/attachments/20140905/95d33a87/attachment.sig>

From matthias at ccs.neu.edu  Fri Sep  5 11:48:40 2014
From: matthias at ccs.neu.edu (Matthias Felleisen)
Date: Fri, 5 Sep 2014 11:48:40 -0400
Subject: [racket] typed racket cps, state machines
In-Reply-To: <5409D2E8.2070903@memebeam.org>
References: <5409D2E8.2070903@memebeam.org>
Message-ID: <02BB644D-7722-412C-8319-BD440CC03561@ccs.neu.edu>


1. Your programs don't look like conventional cps. Perhaps you're thinking 2-way continuations (as in Prolog implementations). 

2. Don't use polymorphism. Just use (define (halt) (printf "done") (halt)) as the proper halt continuation, often falsely called initial continuation. 

3. Think about the algorithm. This hint comes in the form of a cps algorithm: 

  http://pasterack.org/pastes/55294

#lang typed/racket

(require typed/rackunit)

(define-type ?               ;; e = (is one of)
  (U Symbol                  ;; x
     (List 'lambda Symbol ?) ;; ?x.e
     (List ? ?)              ;; (e e)
     ))

(: e1 ?)
(define e1 'x)

(: e2 ?)
(define e2 '(lambda x x))

(: e3 ?)
(define e3 `(,e1 ,e2))

(: e1-cps ?)
(define e1-cps '(lambda k (k x)))

(: e2-cps ?)
(define e2-cps '(lambda k (k (lambda x (lambda k (k x))))))

(: e3-cps ?)
(define e3-cps `(lambda k (,e1-cps (lambda k-e1 (,e2-cps (lambda k-e2 ((k-e1 k-e2) k)))))))

(: cps (-> ? Symbol ?))

(define (cps e k)
  (match e 
    [(? symbol? x) `(lambda ,k (,k ,x))]
    [`(lambda ,(? symbol? x) ,e) `(lambda ,k (,k (lambda ,x ,(cps e 'k))))]
    [`(,f ,a) `(lambda ,k (,(cps f 'k) (lambda k-e1 (,(cps a 'k) (lambda k-e2 ((k-e1 k-e2) ,k))))))]))

(check-equal? (cps e1 'k) e1-cps)
(check-equal? (cps e2 'k) e2-cps)
(check-equal? (cps e3 'k) e3-cps)

;; ---------------------------------------------------------------------------------------------------
;; now write an evaluator for ? and run the examples 

(: halt ?)
(define halt '(lambda x x))

(: e1-run-cps ?)
(define e1-run-cps `(,e1 ,halt))

(: e2-run-cps ?)
(define e2-run-cps `(,e2 ,halt))

(: e3-run-cps ?)
(define e3-run-cps `(,e3 ,halt))




On Sep 5, 2014, at 11:12 AM, Anthony Carrico <acarrico at memebeam.org> wrote:

> Thanks for occasional help learning typed racket on IRC. I'm still
> finding the pitfalls that await those of us used to dynamically typed
> programs.
> 
> The other day I posted some cps patterns I figured out:
>  http://pasterack.org/pastes/49759
> 
> Extending that to simple state machines that loop through various states
> gets me into trouble. The simplest example is just looping through one
> state, or not:
> 
> (define #:? (B0V B1V)
>  (p0 (branch0 : (-> B0V))
>      (branch1 : (-> B1V)))
>  : (U B0V B1V)
> 
>  ;; This one is fine:
>  (define (no-loop) : (U B0V B1V)
>    (branch0))
> 
>  ;; insufficient type information to typecheck this one:
>  (define (loop) : (U B0V B1V)
>    (loop))
> 
>  ;; insufficient type information to typecheck this one too:
>  (define (loop*) : (U B0V B1V)
>    ((inst loop* B0V B1V)))
> 
>  ;; insufficient type information to typecheck this one too:
>  (define (loop**) : (U B0V B1V)
>    ((inst loop** (U B0V B1V) (U B0V B1V))))
> 
>  (no-loop)
>  (loop)
>  (loop*)
>  (loop**))
> 
> I'm still thinking about this problem, but since I have distilled it to
> the simplest cases, I thought I'd post. I have the intuition that this
> has something to do with the nested instantiation of polymorphic types.
> The unions would "obviously" (to a human) collapse, but I suppose the
> type checker isn't picking up on this.
> 
> -- 
> Anthony Carrico
> 
> ____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users



From skanaley at gmail.com  Fri Sep  5 15:14:44 2014
From: skanaley at gmail.com (Sean Kanaley)
Date: Fri, 5 Sep 2014 15:14:44 -0400
Subject: [racket] Submodules can't be imported in same file?
Message-ID: <CA+APS0D0D8iLa2m5nnaFqNePzs2YDTj2TJ_07zRRT=yu21BBHg@mail.gmail.com>

Hello,

For ease of exporting wrapper functions I am trying to do something like

#lang racket

<class system stuff to wrap>

(module wrap racket or module+ wrap
  (provide (all-defined-out))
  <plain function wrappers>)

(require 'wrap)
(provide (all-from-out 'wrap))

With (module wrap racket ...), (require 'wrap) works but the stuff to be
wrapped isn't visible in the module, which is what the module+ form is for,
but then there doesn't seem to be a way to require wrap then. It seems to
require REPL or another file using (submod "<file>" wrap).

Obviously I can just split the code into two files in the first place, but
then what was the intended use for modules or submodules vs. just plain
files?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140905/07798dec/attachment.html>

From acarrico at memebeam.org  Fri Sep  5 15:48:57 2014
From: acarrico at memebeam.org (Anthony Carrico)
Date: Fri, 05 Sep 2014 15:48:57 -0400
Subject: [racket] typed racket cps, state machines
In-Reply-To: <02BB644D-7722-412C-8319-BD440CC03561@ccs.neu.edu>
References: <5409D2E8.2070903@memebeam.org>
 <02BB644D-7722-412C-8319-BD440CC03561@ccs.neu.edu>
Message-ID: <540A13A9.50204@memebeam.org>

On 09/05/2014 11:48 AM, Matthias Felleisen wrote:>
> 1. Your programs don't look like conventional cps...

I have distilled my "cps-ish" examples too far away from any context. A
concrete example of this pattern would be searching. The direct style is:

(findf proc lst) -> any/c

When I say "cps-ish", I'm talking about the following style:

(find proc lst found-tail not-found-tail)

the type turns out to be something like this:

(: find (Any (Element FoundV NotFoundV)
   (-> (-> Element Boolean) (Listof Element)
       (-> Element FoundV)
       (-> NotFoundV)
       (U FoundV NotFoundV)))

In my mind, the computation is essentially "returning" to either the
found-tail, or the not-found-tail. The result types of the tails are
irrelevant to the search computation; they don't show up in direct style
version. Another way to express this would be with two return points, in
the multi-return lambda calculus of Shivers/Fisher, where findf could be:

(: findf (Any (Element) (-> Element Boolean) (Listof Element) <Element,
Void>))

Here are there are two return points, <Element, Void>, found and not found.

This pattern does show up occasionally in plain old racket code.
Sometimes the pattern translates easily to TR, but sometimes not. In my
looping "state machine" example, I had a scanner or parser in mind which
would exit through one of the tails.

> 2. Don't use polymorphism.

Apparently. I was trying to figure out if this is something I don't know
how to annotate properly, or something that really requires a code
transformation to type.

> 3. Think about the algorithm. This hint comes in the form of a cps
> algorithm:

Thank you for the cps converter. I'm not sure if it is relevant to the
intended question, but I think I will write the evaluator you suggest as
an exercise.

-- 
Anthony Carrico


-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 181 bytes
Desc: OpenPGP digital signature
URL: <http://lists.racket-lang.org/users/archive/attachments/20140905/c38dac4c/attachment.sig>

From acarrico at memebeam.org  Fri Sep  5 15:53:59 2014
From: acarrico at memebeam.org (Anthony Carrico)
Date: Fri, 05 Sep 2014 15:53:59 -0400
Subject: [racket] Submodules can't be imported in same file?
In-Reply-To: <CA+APS0D0D8iLa2m5nnaFqNePzs2YDTj2TJ_07zRRT=yu21BBHg@mail.gmail.com>
References: <CA+APS0D0D8iLa2m5nnaFqNePzs2YDTj2TJ_07zRRT=yu21BBHg@mail.gmail.com>
Message-ID: <540A14D7.30605@memebeam.org>

Look at the grammer in
http://docs.racket-lang.org/reference/require.html, you use something
like (submod "." wrap).

-- 
Anthony Carrico

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 181 bytes
Desc: OpenPGP digital signature
URL: <http://lists.racket-lang.org/users/archive/attachments/20140905/bae3c43c/attachment.sig>

From skanaley at gmail.com  Fri Sep  5 16:11:43 2014
From: skanaley at gmail.com (Sean Kanaley)
Date: Fri, 5 Sep 2014 16:11:43 -0400
Subject: [racket] Submodules can't be imported in same file?
In-Reply-To: <540A14D7.30605@memebeam.org>
References: <CA+APS0D0D8iLa2m5nnaFqNePzs2YDTj2TJ_07zRRT=yu21BBHg@mail.gmail.com>
 <540A14D7.30605@memebeam.org>
Message-ID: <CA+APS0D729cox=Yg0ESOd1Lq9_Q1MX+yTbK1=5gXBCXm1ey36g@mail.gmail.com>

(reply to all)
Right but that doesn't work from within the file, only at REPL or another
file.

#lang racket
(module+ invisible
  (provide x)
  (define x 3))

(require (submod "." invisible))

doesn't work
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140905/f60028d9/attachment.html>

From mflatt at cs.utah.edu  Fri Sep  5 16:44:23 2014
From: mflatt at cs.utah.edu (Matthew Flatt)
Date: Fri, 5 Sep 2014 14:44:23 -0600
Subject: [racket] Snapshot builds
In-Reply-To: <F3438AEA-6E9D-44FB-B63C-677A8EA68D97@stevelloyd.net>
References: <F3438AEA-6E9D-44FB-B63C-677A8EA68D97@stevelloyd.net>
Message-ID: <20140905204425.BCC9965018E@mail-svr1.cs.utah.edu>

At Thu, 4 Sep 2014 20:48:43 +0100, Steve Lloyd wrote:
> I recently ?discovered? (thanks to a tweet from @dkvansnickajr) the snapshot 
> builds for ARM6 built on Raspbian. 
> Previously I?d been recommending people try out Racket on RasPi via the 
> source+built packages route, but realise they might baulk at the build times 
> and this other route seems very attractive in this respect. Can I just 
> sanity-check that this new approach is sensible?

Yes, definitely.



From gustavo at oma.org.ar  Fri Sep  5 17:24:59 2014
From: gustavo at oma.org.ar (Gustavo Massaccesi)
Date: Fri, 5 Sep 2014 18:24:59 -0300
Subject: [racket] Continuation marks
Message-ID: <CAPaha9MOb7W76JuvHKAGnas0ts2D5PRRpM_W86bh+W1E_yoY0w@mail.gmail.com>

I'm trying to add a continuation mark inside a function, but I want
that the new mark to be visible after the functions returns. (Mostly
for curiosity, not an actual problem.)

I want something like this:

;--
#lang racket/base

(define (display-cm-x)
  (displayln
   (continuation-mark-set->list
    (current-continuation-marks) 'x)))

(define (add-cm-x v)
  '???)

(define-syntax-rule (with-cm-x v body ...)
  (with-continuation-mark 'x v
    body ...))

(let ()
  (display-cm-x) ;==>()
  (with-cm-x 7
    (display-cm-x)) ;==>(7)
  (display-cm-x) ;==>()
  (add-cm-x 7)
  (display-cm-x) ;actual==>(), expected==>(7)
  )
;--

The last line prints (), but I wish it prints (7).

Gustavo

From greghendershott at gmail.com  Fri Sep  5 18:29:27 2014
From: greghendershott at gmail.com (Greg Hendershott)
Date: Fri, 5 Sep 2014 18:29:27 -0400
Subject: [racket] Submodules can't be imported in same file?
In-Reply-To: <CA+APS0D729cox=Yg0ESOd1Lq9_Q1MX+yTbK1=5gXBCXm1ey36g@mail.gmail.com>
References: <CA+APS0D0D8iLa2m5nnaFqNePzs2YDTj2TJ_07zRRT=yu21BBHg@mail.gmail.com>
 <540A14D7.30605@memebeam.org>
 <CA+APS0D729cox=Yg0ESOd1Lq9_Q1MX+yTbK1=5gXBCXm1ey36g@mail.gmail.com>
Message-ID: <CAGspUn2Tu1Vr_8byCyLRmZUaJtzE0hdAPbEWtunHHh5tA_r5bQ@mail.gmail.com>

The following approach works for me. Would that do what you need?

;; ----- mod.rkt -----
#lang racket

(module private racket
  (define (secret)
    'foo)
  (provide secret))

(require 'private)

(define (wrapper)
  (secret))

(provide (all-defined-out))
;; Note that all-defined-out will not provide stuff from 'private.
;; For that to happen, you would need to use (require (all-from-out
;; 'private)). So ... don't do that.


;; ----- use-mod.rkt -----
#lang racket

(require "mod.rkt")
(wrapper) ;; => 'foo
(secret)  ;; => error -- as desired

From skanaley at gmail.com  Fri Sep  5 19:23:53 2014
From: skanaley at gmail.com (Sean Kanaley)
Date: Fri, 5 Sep 2014 19:23:53 -0400
Subject: [racket] Submodules can't be imported in same file?
In-Reply-To: <CAGspUn2Tu1Vr_8byCyLRmZUaJtzE0hdAPbEWtunHHh5tA_r5bQ@mail.gmail.com>
References: <CA+APS0D0D8iLa2m5nnaFqNePzs2YDTj2TJ_07zRRT=yu21BBHg@mail.gmail.com>
 <540A14D7.30605@memebeam.org>
 <CA+APS0D729cox=Yg0ESOd1Lq9_Q1MX+yTbK1=5gXBCXm1ey36g@mail.gmail.com>
 <CAGspUn2Tu1Vr_8byCyLRmZUaJtzE0hdAPbEWtunHHh5tA_r5bQ@mail.gmail.com>
Message-ID: <CA+APS0ABT+6zLMgg0TG9SEkm335Y1gACQP4xd71nfKnNJP0OKg@mail.gmail.com>

Oh of course, I had inverted the hierarchy of modules. Your way is much
more logical, thanks. I got too sidetracked into what exactly module+
does...


On Fri, Sep 5, 2014 at 6:29 PM, Greg Hendershott <greghendershott at gmail.com>
wrote:

> The following approach works for me. Would that do what you need?
>
> ;; ----- mod.rkt -----
> #lang racket
>
> (module private racket
>   (define (secret)
>     'foo)
>   (provide secret))
>
> (require 'private)
>
> (define (wrapper)
>   (secret))
>
> (provide (all-defined-out))
> ;; Note that all-defined-out will not provide stuff from 'private.
> ;; For that to happen, you would need to use (require (all-from-out
> ;; 'private)). So ... don't do that.
>
>
> ;; ----- use-mod.rkt -----
> #lang racket
>
> (require "mod.rkt")
> (wrapper) ;; => 'foo
> (secret)  ;; => error -- as desired
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140905/8e5ab2b7/attachment.html>

From matthias at ccs.neu.edu  Fri Sep  5 20:21:21 2014
From: matthias at ccs.neu.edu (Matthias Felleisen)
Date: Fri, 5 Sep 2014 20:21:21 -0400
Subject: [racket] Continuation marks
In-Reply-To: <CAPaha9MOb7W76JuvHKAGnas0ts2D5PRRpM_W86bh+W1E_yoY0w@mail.gmail.com>
References: <CAPaha9MOb7W76JuvHKAGnas0ts2D5PRRpM_W86bh+W1E_yoY0w@mail.gmail.com>
Message-ID: <EE3456FB-058B-481D-9AD0-F03A48EFCEC7@ccs.neu.edu>


A continuation represents the rest of the computation with respect to a sub-computation, e.g., a procedure call. After a return from a procedure call, anything having to do is gone. 

In Racket, you can grab this continuation, including pieces from sub-sub-computations within the extent of the procedure call. Grabing means you turn them into a value within the language. And on those continuations you can find the marks. By reinstalling these continuations, you can get the marks back. 




On Sep 5, 2014, at 5:24 PM, Gustavo Massaccesi wrote:

> I'm trying to add a continuation mark inside a function, but I want
> that the new mark to be visible after the functions returns. (Mostly
> for curiosity, not an actual problem.)
> 
> I want something like this:
> 
> ;--
> #lang racket/base
> 
> (define (display-cm-x)
>  (displayln
>   (continuation-mark-set->list
>    (current-continuation-marks) 'x)))
> 
> (define (add-cm-x v)
>  '???)
> 
> (define-syntax-rule (with-cm-x v body ...)
>  (with-continuation-mark 'x v
>    body ...))
> 
> (let ()
>  (display-cm-x) ;==>()
>  (with-cm-x 7
>    (display-cm-x)) ;==>(7)
>  (display-cm-x) ;==>()
>  (add-cm-x 7)
>  (display-cm-x) ;actual==>(), expected==>(7)
>  )
> ;--
> 
> The last line prints (), but I wish it prints (7).
> 
> Gustavo
> ____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users



From matthias at ccs.neu.edu  Fri Sep  5 21:37:30 2014
From: matthias at ccs.neu.edu (Matthias Felleisen)
Date: Fri, 5 Sep 2014 21:37:30 -0400
Subject: [racket] typed racket cps, state machines
In-Reply-To: <540A13A9.50204@memebeam.org>
References: <5409D2E8.2070903@memebeam.org>
 <02BB644D-7722-412C-8319-BD440CC03561@ccs.neu.edu>
 <540A13A9.50204@memebeam.org>
Message-ID: <F698721A-1684-49F1-B30D-9FF38731D838@ccs.neu.edu>


Would this do for you: 

#lang typed/racket

(require typed/rackunit)

(define-type Answer Boolean)

(: find (All (?) (-> (-> ? Boolean) (Listof ?) (-> (Listof ?) Answer) (-> Answer) Answer)))
(define (find pred? l sk fk)
  (cond
    [(empty? l) (fk)]
    [(pred? (first l)) (sk l)]
    [else (find pred? (rest l) sk fk)]))
                       
(check-equal? (find odd? '(2 3 4) (? (x) #t) (? () #f)) #t)





On Sep 5, 2014, at 3:48 PM, Anthony Carrico wrote:

> On 09/05/2014 11:48 AM, Matthias Felleisen wrote:>
>> 1. Your programs don't look like conventional cps...
> 
> I have distilled my "cps-ish" examples too far away from any context. A
> concrete example of this pattern would be searching. The direct style is:
> 
> (findf proc lst) -> any/c
> 
> When I say "cps-ish", I'm talking about the following style:
> 
> (find proc lst found-tail not-found-tail)
> 
> the type turns out to be something like this:
> 
> (: find (Any (Element FoundV NotFoundV)
>   (-> (-> Element Boolean) (Listof Element)
>       (-> Element FoundV)
>       (-> NotFoundV)
>       (U FoundV NotFoundV)))
> 
> In my mind, the computation is essentially "returning" to either the
> found-tail, or the not-found-tail. The result types of the tails are
> irrelevant to the search computation; they don't show up in direct style
> version. Another way to express this would be with two return points, in
> the multi-return lambda calculus of Shivers/Fisher, where findf could be:
> 
> (: findf (Any (Element) (-> Element Boolean) (Listof Element) <Element,
> Void>))
> 
> Here are there are two return points, <Element, Void>, found and not found.
> 
> This pattern does show up occasionally in plain old racket code.
> Sometimes the pattern translates easily to TR, but sometimes not. In my
> looping "state machine" example, I had a scanner or parser in mind which
> would exit through one of the tails.
> 
>> 2. Don't use polymorphism.
> 
> Apparently. I was trying to figure out if this is something I don't know
> how to annotate properly, or something that really requires a code
> transformation to type.
> 
>> 3. Think about the algorithm. This hint comes in the form of a cps
>> algorithm:
> 
> Thank you for the cps converter. I'm not sure if it is relevant to the
> intended question, but I think I will write the evaluator you suggest as
> an exercise.
> 
> -- 
> Anthony Carrico
> 
> 



From johnbclements at gmail.com  Sat Sep  6 02:36:24 2014
From: johnbclements at gmail.com (John Clements)
Date: Fri, 5 Sep 2014 23:36:24 -0700
Subject: [racket] Best typed/racket representation for bitfield of width
	125?
In-Reply-To: <878ulzji02.wl%stamourv@ccs.neu.edu>
References: <CAAwfr=Wh0-vJib+tzaSpaTcsqHnrtMr3kirnp7w5Mjxp5vJdng@mail.gmail.com>
 <878ulzji02.wl%stamourv@ccs.neu.edu>
Message-ID: <CAAwfr=Ws0=MvUV-HJHTCPWPJJnTmefcgs7vOYQ1aG5pqMky6LQ@mail.gmail.com>

Is this a bug, or just a known limitation of typed racket? It looks like a
fixnum can be up to 2^61, but the type checker for integer literals doesn't
like them. This program:

#lang typed/racket

(require racket/fixnum)

(define (ensure-fixnum i)
  (cond [(fixnum? i) i]
        [else (error 'ensure-fixnum)]))

(fixnum? 281474976710656) ;; -> #t
(: i Fixnum)
(define i (ensure-fixnum 281474976710656))

works fine, but requires the ensure-fixnum, because the type checker does
not believe that the literal 281474976710656 lives in fixnum.

You can also see this at the top level, using (:print-type 281474976710656)

Is this a bug, or is this just a known compromise to simplify the
type-checker's portability and numeric tower?

John




On Thu, Sep 4, 2014 at 12:35 PM, Vincent St-Amour <stamourv at ccs.neu.edu>
wrote:

> Would an fxvector of length 5 (to account for the tag bits) work?
>
> If all you're doing is bitwise ops and comparisons, it should be easy to
> convince TR that you're staying within fixnum range.
>
> Vincent
>
>
>
>
> At Thu, 4 Sep 2014 10:45:48 -0700,
> John Clements wrote:
> >
> > [1  <multipart/alternative (7bit)>]
> > [1.1  <text/plain; UTF-8 (quoted-printable)>]
> > I want to represent a bitfield of 125 bits, using typed racket. I?ll be
> > combining them using bitwise operations, and checking for zero-ness. I
> > think that?s about it.
> >
> > I could just use type Integer, but it seems like that would impose
> > additional checking, since the type system doesn?t know the size.  In
> > untyped racket I used a vector of four integers in the range 0-2^32, but
> > those don?t fit in typed Racket?s ?Fixnum? because it uses a signed
> > representation (and indeed, untyped racket may have been falling prey to
> > the same problem. Perhaps I should break it up into four integers in the
> > range 0-2^31? That seems weird, but maybe that?s my best choice? Or? I
> > could use a byte string.
> >
> > Apologies if I?m missing something in the docs, here; I?ve spent a few
> > minutes searching, without turning anything up.
> >
> > Thanks for any advice,
> >
> > John
> > [1.2  <text/html; UTF-8 (quoted-printable)>]
> >
> > [2  <text/plain; us-ascii (7bit)>]
> > ____________________
> >   Racket Users list:
> >   http://lists.racket-lang.org/users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140905/0500f5ee/attachment.html>

From samth at cs.indiana.edu  Sat Sep  6 02:41:28 2014
From: samth at cs.indiana.edu (Sam Tobin-Hochstadt)
Date: Sat, 6 Sep 2014 02:41:28 -0400
Subject: [racket] Best typed/racket representation for bitfield of width
	125?
In-Reply-To: <CAAwfr=Ws0=MvUV-HJHTCPWPJJnTmefcgs7vOYQ1aG5pqMky6LQ@mail.gmail.com>
References: <CAAwfr=Wh0-vJib+tzaSpaTcsqHnrtMr3kirnp7w5Mjxp5vJdng@mail.gmail.com>
 <878ulzji02.wl%stamourv@ccs.neu.edu>
 <CAAwfr=Ws0=MvUV-HJHTCPWPJJnTmefcgs7vOYQ1aG5pqMky6LQ@mail.gmail.com>
Message-ID: <CAK=HD+Z7_R-XpEOQnEZN=9Yrf7sgb8RmOG4mbaeeTW+YK7Fosw@mail.gmail.com>

If you type check the file on a 64-bit system like yours, but then ran it
on a 32-bit system, the fixnum range would change, making that number into
a bignum.

Sam
On Sep 6, 2014 8:37 AM, "John Clements" <johnbclements at gmail.com> wrote:

> Is this a bug, or just a known limitation of typed racket? It looks like a
> fixnum can be up to 2^61, but the type checker for integer literals doesn't
> like them. This program:
>
> #lang typed/racket
>
> (require racket/fixnum)
>
> (define (ensure-fixnum i)
>   (cond [(fixnum? i) i]
>         [else (error 'ensure-fixnum)]))
>
> (fixnum? 281474976710656) ;; -> #t
> (: i Fixnum)
> (define i (ensure-fixnum 281474976710656))
>
> works fine, but requires the ensure-fixnum, because the type checker does
> not believe that the literal 281474976710656 lives in fixnum.
>
> You can also see this at the top level, using (:print-type 281474976710656)
>
> Is this a bug, or is this just a known compromise to simplify the
> type-checker's portability and numeric tower?
>
> John
>
>
>
>
> On Thu, Sep 4, 2014 at 12:35 PM, Vincent St-Amour <stamourv at ccs.neu.edu>
> wrote:
>
>> Would an fxvector of length 5 (to account for the tag bits) work?
>>
>> If all you're doing is bitwise ops and comparisons, it should be easy to
>> convince TR that you're staying within fixnum range.
>>
>> Vincent
>>
>>
>>
>>
>> At Thu, 4 Sep 2014 10:45:48 -0700,
>> John Clements wrote:
>> >
>> > [1  <multipart/alternative (7bit)>]
>> > [1.1  <text/plain; UTF-8 (quoted-printable)>]
>> > I want to represent a bitfield of 125 bits, using typed racket. I?ll be
>> > combining them using bitwise operations, and checking for zero-ness. I
>> > think that?s about it.
>> >
>> > I could just use type Integer, but it seems like that would impose
>> > additional checking, since the type system doesn?t know the size.  In
>> > untyped racket I used a vector of four integers in the range 0-2^32, but
>> > those don?t fit in typed Racket?s ?Fixnum? because it uses a signed
>> > representation (and indeed, untyped racket may have been falling prey to
>> > the same problem. Perhaps I should break it up into four integers in the
>> > range 0-2^31? That seems weird, but maybe that?s my best choice? Or? I
>> > could use a byte string.
>> >
>> > Apologies if I?m missing something in the docs, here; I?ve spent a few
>> > minutes searching, without turning anything up.
>> >
>> > Thanks for any advice,
>> >
>> > John
>> > [1.2  <text/html; UTF-8 (quoted-printable)>]
>> >
>> > [2  <text/plain; us-ascii (7bit)>]
>> > ____________________
>> >   Racket Users list:
>> >   http://lists.racket-lang.org/users
>>
>
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140906/d7ea0bd8/attachment-0001.html>

From gustavo at oma.org.ar  Sat Sep  6 10:36:37 2014
From: gustavo at oma.org.ar (Gustavo Massaccesi)
Date: Sat, 6 Sep 2014 11:36:37 -0300
Subject: [racket] Continuation marks
In-Reply-To: <4A40AA5E-88DB-4E91-99BB-D6702125E951@brinckerhoff.org>
References: <CAPaha9MOb7W76JuvHKAGnas0ts2D5PRRpM_W86bh+W1E_yoY0w@mail.gmail.com>
 <4A40AA5E-88DB-4E91-99BB-D6702125E951@brinckerhoff.org>
Message-ID: <CAPaha9OQxrjHUH+yMx5-CFwRXZSmQyYan4Cu4wYKkJdXZs-UZA@mail.gmail.com>

I agree that to get this output, using continuation marks is a bad
idea. It'd be much better to use a parameter (or set!ing a global
variable at a last resort).

I'm trying to understand continuations marks. I'm reading some parts
of the Racket compiler, and internally the functions have some flags
that are not visible from the Racket code.

One of the flags is 'preserves-marks. From
http://docs.racket-lang.org/raco/decompile.html

> The preserves-marks? field is true if calling the function is expected to leave continuation marks unchanged by the time it returns.

So: Can a function change the continuation marks after it returns?
What does that exactly mean? Can I use that to complete the code in my
original e-mail? Is that a good idea for a real program?

(The last question is easy: No. :) )


I also read the answer from Matthias Felleisen. I'm not sure that I
understood it completely, but somehow it give me a new idea to solve
this. But I need to try it before I'm sure it works.


Gustavo

On Fri, Sep 5, 2014 at 7:04 PM, John Clements <clements at brinckerhoff.org> wrote:
>
> On Sep 5, 2014, at 2:24 PM, Gustavo Massaccesi <gustavo at oma.org.ar> wrote:
>
>> I'm trying to add a continuation mark inside a function, but I want
>> that the new mark to be visible after the functions returns. (Mostly
>> for curiosity, not an actual problem.)
>>
>> I want something like this:
>>
>> ;--
>> #lang racket/base
>>
>> (define (display-cm-x)
>>  (displayln
>>   (continuation-mark-set->list
>>    (current-continuation-marks) 'x)))
>>
>> (define (add-cm-x v)
>>  '???)
>>
>> (define-syntax-rule (with-cm-x v body ...)
>>  (with-continuation-mark 'x v
>>    body ...))
>>
>> (let ()
>>  (display-cm-x) ;==>()
>>  (with-cm-x 7
>>    (display-cm-x)) ;==>(7)
>>  (display-cm-x) ;==>()
>>  (add-cm-x 7)
>>  (display-cm-x) ;actual==>(), expected==>(7)
>>  )
>> ;--
>>
>> The last line prints (), but I wish it prints (7).
>
> This is going to sound flip, but ? it sounds like you don?t want continuation marks. The whole point of continuation marks is that they?re associated with continuations, and will disappear when those continuations are gone.
>
> The example you give isn?t quite enough to figure out exactly what you want; it sounds like some combination of fluid-let and ? something else.
>
> Maybe you can describe why it is that you want this behavior?
>
> John
>


From clements at brinckerhoff.org  Fri Sep  5 18:04:12 2014
From: clements at brinckerhoff.org (John Clements)
Date: Fri, 5 Sep 2014 15:04:12 -0700
Subject: [racket] Continuation marks
In-Reply-To: <CAPaha9MOb7W76JuvHKAGnas0ts2D5PRRpM_W86bh+W1E_yoY0w@mail.gmail.com>
References: <CAPaha9MOb7W76JuvHKAGnas0ts2D5PRRpM_W86bh+W1E_yoY0w@mail.gmail.com>
Message-ID: <4A40AA5E-88DB-4E91-99BB-D6702125E951@brinckerhoff.org>


On Sep 5, 2014, at 2:24 PM, Gustavo Massaccesi <gustavo at oma.org.ar> wrote:

> I'm trying to add a continuation mark inside a function, but I want
> that the new mark to be visible after the functions returns. (Mostly
> for curiosity, not an actual problem.)
> 
> I want something like this:
> 
> ;--
> #lang racket/base
> 
> (define (display-cm-x)
>  (displayln
>   (continuation-mark-set->list
>    (current-continuation-marks) 'x)))
> 
> (define (add-cm-x v)
>  '???)
> 
> (define-syntax-rule (with-cm-x v body ...)
>  (with-continuation-mark 'x v
>    body ...))
> 
> (let ()
>  (display-cm-x) ;==>()
>  (with-cm-x 7
>    (display-cm-x)) ;==>(7)
>  (display-cm-x) ;==>()
>  (add-cm-x 7)
>  (display-cm-x) ;actual==>(), expected==>(7)
>  )
> ;--
> 
> The last line prints (), but I wish it prints (7).

This is going to sound flip, but ? it sounds like you don?t want continuation marks. The whole point of continuation marks is that they?re associated with continuations, and will disappear when those continuations are gone.

The example you give isn?t quite enough to figure out exactly what you want; it sounds like some combination of fluid-let and ? something else.

Maybe you can describe why it is that you want this behavior?

John



From mflatt at cs.utah.edu  Sat Sep  6 12:19:53 2014
From: mflatt at cs.utah.edu (Matthew Flatt)
Date: Sat, 6 Sep 2014 10:19:53 -0600
Subject: [racket] Continuation marks
In-Reply-To: <CAPaha9OQxrjHUH+yMx5-CFwRXZSmQyYan4Cu4wYKkJdXZs-UZA@mail.gmail.com>
References: <CAPaha9MOb7W76JuvHKAGnas0ts2D5PRRpM_W86bh+W1E_yoY0w@mail.gmail.com>
 <4A40AA5E-88DB-4E91-99BB-D6702125E951@brinckerhoff.org>
 <CAPaha9OQxrjHUH+yMx5-CFwRXZSmQyYan4Cu4wYKkJdXZs-UZA@mail.gmail.com>
Message-ID: <20140906161955.6AE2E650190@mail-svr1.cs.utah.edu>

The 'preserves-marks flag makes sense only in the context of the JIT's
implementation.

At the Racket level, there's no way to write a function `f` so that

 (with-continuation-mark
  'x
  'my-value
  (begin
   (f)
   (continuation-mark-set-first #f 'x)))

produces anything other than 'my-value. Even if `f` contains a
`with-continuation-mark` form to set a mark with the key 'x, returning
from `f` necessarily means leaving the dynamic extent of that form, so
the mark will not be visible.

It turns out that the JITted form of `with-continuation-mark` is not
responsible for popping a mark stack as it returns. Instead, every
non-tail position is wrapped with a kind of push and pop of the
continuation-mark stack. Since `(f)` appears in a non-tail position
within with `begin` form, the call with be wrapped.

If `f` is known to not have a `with-continuation-mark` as a tail form,
however, then the push and pop of the continuation-mark stack will not
be necessary around the `(f)` call. The 'preserves-marks flag tells the
JIT that it can make that assumption. (In retrospect, a better name
would be 'has-no-wcm-in-tail-position.)


At Sat, 6 Sep 2014 11:36:37 -0300, Gustavo Massaccesi wrote:
> I agree that to get this output, using continuation marks is a bad
> idea. It'd be much better to use a parameter (or set!ing a global
> variable at a last resort).
> 
> I'm trying to understand continuations marks. I'm reading some parts
> of the Racket compiler, and internally the functions have some flags
> that are not visible from the Racket code.
> 
> One of the flags is 'preserves-marks. From
> http://docs.racket-lang.org/raco/decompile.html
> 
> > The preserves-marks? field is true if calling the function is expected to 
> leave continuation marks unchanged by the time it returns.
> 
> So: Can a function change the continuation marks after it returns?
> What does that exactly mean? Can I use that to complete the code in my
> original e-mail? Is that a good idea for a real program?
> 
> (The last question is easy: No. :) )
> 
> 
> I also read the answer from Matthias Felleisen. I'm not sure that I
> understood it completely, but somehow it give me a new idea to solve
> this. But I need to try it before I'm sure it works.
> 
> 
> Gustavo
> 
> On Fri, Sep 5, 2014 at 7:04 PM, John Clements <clements at brinckerhoff.org> wrote:
> >
> > On Sep 5, 2014, at 2:24 PM, Gustavo Massaccesi <gustavo at oma.org.ar> wrote:
> >
> >> I'm trying to add a continuation mark inside a function, but I want
> >> that the new mark to be visible after the functions returns. (Mostly
> >> for curiosity, not an actual problem.)
> >>
> >> I want something like this:
> >>
> >> ;--
> >> #lang racket/base
> >>
> >> (define (display-cm-x)
> >>  (displayln
> >>   (continuation-mark-set->list
> >>    (current-continuation-marks) 'x)))
> >>
> >> (define (add-cm-x v)
> >>  '???)
> >>
> >> (define-syntax-rule (with-cm-x v body ...)
> >>  (with-continuation-mark 'x v
> >>    body ...))
> >>
> >> (let ()
> >>  (display-cm-x) ;==>()
> >>  (with-cm-x 7
> >>    (display-cm-x)) ;==>(7)
> >>  (display-cm-x) ;==>()
> >>  (add-cm-x 7)
> >>  (display-cm-x) ;actual==>(), expected==>(7)
> >>  )
> >> ;--
> >>
> >> The last line prints (), but I wish it prints (7).
> >
> > This is going to sound flip, but ? it sounds like you don?t want continuation 
> marks. The whole point of continuation marks is that they?re associated with 
> continuations, and will disappear when those continuations are gone.
> >
> > The example you give isn?t quite enough to figure out exactly what you want; 
> it sounds like some combination of fluid-let and ? something else.
> >
> > Maybe you can describe why it is that you want this behavior?
> >
> > John
> >
> 
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users


From jensaxel at soegaard.net  Sat Sep  6 12:28:19 2014
From: jensaxel at soegaard.net (=?UTF-8?Q?Jens_Axel_S=C3=B8gaard?=)
Date: Sat, 6 Sep 2014 18:28:19 +0200
Subject: [racket] Continuation marks
In-Reply-To: <CAPaha9OQxrjHUH+yMx5-CFwRXZSmQyYan4Cu4wYKkJdXZs-UZA@mail.gmail.com>
References: <CAPaha9MOb7W76JuvHKAGnas0ts2D5PRRpM_W86bh+W1E_yoY0w@mail.gmail.com>
 <4A40AA5E-88DB-4E91-99BB-D6702125E951@brinckerhoff.org>
 <CAPaha9OQxrjHUH+yMx5-CFwRXZSmQyYan4Cu4wYKkJdXZs-UZA@mail.gmail.com>
Message-ID: <CABefVgzXK3WAem73BLcc_rZjk8DOsHC374R1cYRyAEw6h952CQ@mail.gmail.com>

2014-09-06 16:36 GMT+02:00 Gustavo Massaccesi <gustavo at oma.org.ar>:
> I'm trying to understand continuations marks.

The second chapter of Johns dissertation
    "Portable and high-level access to the stack with Continuation Marks"
gives a nice introduction.

http://www.brinckerhoff.org/clements/papers/dissertation.pdf

-- 
Jens Axel S?gaard


From acarrico at memebeam.org  Sat Sep  6 13:46:38 2014
From: acarrico at memebeam.org (Anthony Carrico)
Date: Sat, 06 Sep 2014 13:46:38 -0400
Subject: [racket] typed racket cps, state machines
In-Reply-To: <F698721A-1684-49F1-B30D-9FF38731D838@ccs.neu.edu>
References: <5409D2E8.2070903@memebeam.org>
 <02BB644D-7722-412C-8319-BD440CC03561@ccs.neu.edu>
 <540A13A9.50204@memebeam.org>
 <F698721A-1684-49F1-B30D-9FF38731D838@ccs.neu.edu>
Message-ID: <540B487E.10705@memebeam.org>

Even something like this won't type check:

(define #:? (V) (fn (r : V)) : V
  (define (loop) : Void (loop))
  r)

So all that stuff about cps-ish code may be irrelevant.

-- 
Anthony Carrico

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 181 bytes
Desc: OpenPGP digital signature
URL: <http://lists.racket-lang.org/users/archive/attachments/20140906/3056c7fa/attachment.sig>

From matthias at ccs.neu.edu  Sat Sep  6 13:58:47 2014
From: matthias at ccs.neu.edu (Matthias Felleisen)
Date: Sat, 6 Sep 2014 13:58:47 -0400
Subject: [racket] typed racket cps, state machines
In-Reply-To: <540B487E.10705@memebeam.org>
References: <5409D2E8.2070903@memebeam.org>
 <02BB644D-7722-412C-8319-BD440CC03561@ccs.neu.edu>
 <540A13A9.50204@memebeam.org>
 <F698721A-1684-49F1-B30D-9FF38731D838@ccs.neu.edu>
 <540B487E.10705@memebeam.org>
Message-ID: <4DB683A6-0A63-43DE-AA7A-30C6CA34816C@ccs.neu.edu>


This type checks fine in v6.1 though I can't read the dang syntax and I thought I was someone who should be able to do just that :-)


On Sep 6, 2014, at 1:46 PM, Anthony Carrico wrote:

> Even something like this won't type check:
> 
> (define #:? (V) (fn (r : V)) : V
>  (define (loop) : Void (loop))
>  r)
> 
> So all that stuff about cps-ish code may be irrelevant.
> 
> -- 
> Anthony Carrico
> 



From matthias at ccs.neu.edu  Sat Sep  6 14:00:09 2014
From: matthias at ccs.neu.edu (Matthias Felleisen)
Date: Sat, 6 Sep 2014 14:00:09 -0400
Subject: [racket] typed racket cps, state machines
In-Reply-To: <4DB683A6-0A63-43DE-AA7A-30C6CA34816C@ccs.neu.edu>
References: <5409D2E8.2070903@memebeam.org>
 <02BB644D-7722-412C-8319-BD440CC03561@ccs.neu.edu>
 <540A13A9.50204@memebeam.org>
 <F698721A-1684-49F1-B30D-9FF38731D838@ccs.neu.edu>
 <540B487E.10705@memebeam.org>
 <4DB683A6-0A63-43DE-AA7A-30C6CA34816C@ccs.neu.edu>
Message-ID: <F3B7C386-F66B-41D2-9EB9-98BFA6439CD9@ccs.neu.edu>


p.s. I have figured out the syntax. I am not sure what you want with 'loop' but in any case, this all type checks. 



On Sep 6, 2014, at 1:58 PM, Matthias Felleisen wrote:

> 
> This type checks fine in v6.1 though I can't read the dang syntax and I thought I was someone who should be able to do just that :-)
> 
> 
> On Sep 6, 2014, at 1:46 PM, Anthony Carrico wrote:
> 
>> Even something like this won't type check:
>> 
>> (define #:? (V) (fn (r : V)) : V
>> (define (loop) : Void (loop))
>> r)
>> 
>> So all that stuff about cps-ish code may be irrelevant.
>> 
>> -- 
>> Anthony Carrico
>> 
> 
> 
> ____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users



From gustavo at oma.org.ar  Sat Sep  6 19:31:57 2014
From: gustavo at oma.org.ar (Gustavo Massaccesi)
Date: Sat, 6 Sep 2014 20:31:57 -0300
Subject: [racket] Continuation marks
In-Reply-To: <20140906161955.6AE2E650190@mail-svr1.cs.utah.edu>
References: <CAPaha9MOb7W76JuvHKAGnas0ts2D5PRRpM_W86bh+W1E_yoY0w@mail.gmail.com>
 <4A40AA5E-88DB-4E91-99BB-D6702125E951@brinckerhoff.org>
 <CAPaha9OQxrjHUH+yMx5-CFwRXZSmQyYan4Cu4wYKkJdXZs-UZA@mail.gmail.com>
 <20140906161955.6AE2E650190@mail-svr1.cs.utah.edu>
Message-ID: <CAPaha9N2_U+9kUstHBm780ed+evnTkfNtQth=z5gRpA=12XT_w@mail.gmail.com>

Thanks to everyone. It's much clearer now.

I tried to get the same output anyway. This is the closest I could
get. The example involves a little of "cheating", because the add-cm-x
function actually never return in the traditional sense, but it use
call/cc and abort/cc to get the same effect.

Perhaps this can be "improved" with more work, but it was only a
curiosity question to understand more about the continuation marks.

;--
#lang racket/base
(define (display-cm-x)
  (display
   (continuation-mark-set->list
    (current-continuation-marks) 'x))
  (display "  first: ")
  (displayln
   (continuation-mark-set-first
    (current-continuation-marks) 'x)))

(define prompt (make-continuation-prompt-tag 'temp))

(define (add-cm-x v)
  (call-with-composable-continuation
   (lambda (xx)
     (with-cm-x 7
       (let ([r (xx (void))])
         (abort-current-continuation prompt r))))))

(define-syntax-rule (with-cm-x v body ...)
  (with-continuation-mark 'x v
    body ...))

(call-with-continuation-prompt
 (lambda ()
   (with-cm-x 0
     (let ()
       (display-cm-x) ;==>(0) first: 0
       (with-cm-x 7
         (display-cm-x)) ;==>(7 0) first: 7
       (display-cm-x) ;==>(0) first: 0
       (add-cm-x 7)
       (display "!")
       (display-cm-x) ;==>(0 7 0) first: 0 :(
       333)))
 prompt
 (lambda (x) x))

(newline)

(call-with-continuation-prompt
 (lambda ()
   (let ()
     (display-cm-x) ;==>() first: #f
     (with-cm-x 7
       (display-cm-x)) ;==>(7) first: 7
     (display-cm-x) ;==>() first: #f
     (add-cm-x 7)
     (display "!")
     (display-cm-x) ;==>(7) first: 7 :)
     333))
 prompt
 (lambda (x) x))

;--

Gustavo

On Sat, Sep 6, 2014 at 1:19 PM, Matthew Flatt <mflatt at cs.utah.edu> wrote:
> The 'preserves-marks flag makes sense only in the context of the JIT's
> implementation.
>
> At the Racket level, there's no way to write a function `f` so that
>
>  (with-continuation-mark
>   'x
>   'my-value
>   (begin
>    (f)
>    (continuation-mark-set-first #f 'x)))
>
> produces anything other than 'my-value. Even if `f` contains a
> `with-continuation-mark` form to set a mark with the key 'x, returning
> from `f` necessarily means leaving the dynamic extent of that form, so
> the mark will not be visible.
>
> It turns out that the JITted form of `with-continuation-mark` is not
> responsible for popping a mark stack as it returns. Instead, every
> non-tail position is wrapped with a kind of push and pop of the
> continuation-mark stack. Since `(f)` appears in a non-tail position
> within with `begin` form, the call with be wrapped.
>
> If `f` is known to not have a `with-continuation-mark` as a tail form,
> however, then the push and pop of the continuation-mark stack will not
> be necessary around the `(f)` call. The 'preserves-marks flag tells the
> JIT that it can make that assumption. (In retrospect, a better name
> would be 'has-no-wcm-in-tail-position.)
>
>
> At Sat, 6 Sep 2014 11:36:37 -0300, Gustavo Massaccesi wrote:
>> I agree that to get this output, using continuation marks is a bad
>> idea. It'd be much better to use a parameter (or set!ing a global
>> variable at a last resort).
>>
>> I'm trying to understand continuations marks. I'm reading some parts
>> of the Racket compiler, and internally the functions have some flags
>> that are not visible from the Racket code.
>>
>> One of the flags is 'preserves-marks. From
>> http://docs.racket-lang.org/raco/decompile.html
>>
>> > The preserves-marks? field is true if calling the function is expected to
>> leave continuation marks unchanged by the time it returns.
>>
>> So: Can a function change the continuation marks after it returns?
>> What does that exactly mean? Can I use that to complete the code in my
>> original e-mail? Is that a good idea for a real program?
>>
>> (The last question is easy: No. :) )
>>
>>
>> I also read the answer from Matthias Felleisen. I'm not sure that I
>> understood it completely, but somehow it give me a new idea to solve
>> this. But I need to try it before I'm sure it works.
>>
>>
>> Gustavo
>>
>> On Fri, Sep 5, 2014 at 7:04 PM, John Clements <clements at brinckerhoff.org> wrote:
>> >
>> > On Sep 5, 2014, at 2:24 PM, Gustavo Massaccesi <gustavo at oma.org.ar> wrote:
>> >
>> >> I'm trying to add a continuation mark inside a function, but I want
>> >> that the new mark to be visible after the functions returns. (Mostly
>> >> for curiosity, not an actual problem.)
>> >>
>> >> I want something like this:
>> >>
>> >> ;--
>> >> #lang racket/base
>> >>
>> >> (define (display-cm-x)
>> >>  (displayln
>> >>   (continuation-mark-set->list
>> >>    (current-continuation-marks) 'x)))
>> >>
>> >> (define (add-cm-x v)
>> >>  '???)
>> >>
>> >> (define-syntax-rule (with-cm-x v body ...)
>> >>  (with-continuation-mark 'x v
>> >>    body ...))
>> >>
>> >> (let ()
>> >>  (display-cm-x) ;==>()
>> >>  (with-cm-x 7
>> >>    (display-cm-x)) ;==>(7)
>> >>  (display-cm-x) ;==>()
>> >>  (add-cm-x 7)
>> >>  (display-cm-x) ;actual==>(), expected==>(7)
>> >>  )
>> >> ;--
>> >>
>> >> The last line prints (), but I wish it prints (7).
>> >
>> > This is going to sound flip, but ? it sounds like you don?t want continuation
>> marks. The whole point of continuation marks is that they?re associated with
>> continuations, and will disappear when those continuations are gone.
>> >
>> > The example you give isn?t quite enough to figure out exactly what you want;
>> it sounds like some combination of fluid-let and ? something else.
>> >
>> > Maybe you can describe why it is that you want this behavior?
>> >
>> > John
>> >
>>
>> ____________________
>>   Racket Users list:
>>   http://lists.racket-lang.org/users


From neil at neilvandyke.org  Sat Sep  6 20:34:00 2014
From: neil at neilvandyke.org (Neil Van Dyke)
Date: Sat, 06 Sep 2014 20:34:00 -0400
Subject: [racket] gnucash and racket
Message-ID: <540BA7F8.4060400@neilvandyke.org>

What all have people done with using Racket with GnuCash?

I know that John B. Clements wrote a package for doing reports from the 
".gnucash" file.

(My immediate motivation is that I'd like to be able to use my 
"html-template" Racket package to do reports that are integrated into 
the GnuCash interface, like the Guile reports are.  Otherwise, I'll 
probably use this "eguile" HTML template thing.)

Neil V.


From hendrik at topoi.pooq.com  Sat Sep  6 20:56:25 2014
From: hendrik at topoi.pooq.com (Hendrik Boom)
Date: Sat, 6 Sep 2014 20:56:25 -0400
Subject: [racket] gnucash and racket
In-Reply-To: <540BA7F8.4060400@neilvandyke.org>
References: <540BA7F8.4060400@neilvandyke.org>
Message-ID: <20140907005625.GA7496@topoi.pooq.com>

On Sat, Sep 06, 2014 at 08:34:00PM -0400, Neil Van Dyke wrote:
> What all have people done with using Racket with GnuCash?

Racket I don't know.  But the scripting language for gnucash, from 
the very beginning ages ago was guile, whch is a dialect of Scheme.

In fact, at one time gnucash was a guile program that invoked code 
written in C (or C++?). That is, of course, how a scripting language 
is supposed to work.  

I gather that the current gnucash developer(s) have decided that 
they would like to distance themselves from guile.  But it is still 
alive and well as its report generator.

The main problem I have with using guile for generating reports is 
that I can find no documentation about the guile-gnucash interface.

-- hendrik

From lysseus at gmail.com  Sat Sep  6 21:55:40 2014
From: lysseus at gmail.com (Kevin Forchione)
Date: Sat, 6 Sep 2014 18:55:40 -0700
Subject: [racket]  recursion style question
Message-ID: <F3AF6EB7-306F-4D2B-A69E-F48EEB1B5EB3@gmail.com>

Hi guys,
Which is preferable? 

(define (foo let ? (acc empty)) ? (foo (rest let) ... (cons ?. acc))

or 

(define (foo lst ...) (let loop ([lst lst] ?[acc empty]) ? (loop (rest let) ? (cons ?. acc)))

On the one hand we eliminate the named let construct, but on the other hand we expose the acc in the function / contract.  If the acc is not something the caller would ever need to know about, should we hide it inside the definition?

-Kevin

From neil at neilvandyke.org  Sat Sep  6 23:38:29 2014
From: neil at neilvandyke.org (Neil Van Dyke)
Date: Sat, 06 Sep 2014 23:38:29 -0400
Subject: [racket] gnucash and racket
In-Reply-To: <20140907005625.GA7496@topoi.pooq.com>
References: <540BA7F8.4060400@neilvandyke.org>
 <20140907005625.GA7496@topoi.pooq.com>
Message-ID: <540BD335.7060204@neilvandyke.org>

Hendrik Boom wrote at 09/06/2014 08:56 PM:
> I gather that the current gnucash developer(s) have decided that
> they would like to distance themselves from guile.  But it is still
> alive and well as its report generator.

I was using Guile back when GnuCash was being developed in it. Guile 
seemed like a great choice for extension language for GnuCash, and still 
could be.  (RMS was intending Guile to be the GNU extension/application 
language for all apps, which I suspect would've been Emacs-times-100 
awesome, for how people make and share software enhancements, but events 
not Guile's fault ended that.)

Anyway, this evening, having just made my first GnuCash report, by 
adapting "balsheet-eg"... I have to say that the eguile report scripts 
aren't as pleasant to work with as they could be.  It's easy to imagine 
someone thinking, ``This is scary and painful, and would be better in 
Python.''  GnuCash *could* be made better to provide a better API in 
Guile than is possible in Python, but is not currently.

I'm very fond of GnuCash overall, however.

Neil V.


From acarrico at memebeam.org  Sat Sep  6 23:41:47 2014
From: acarrico at memebeam.org (Anthony Carrico)
Date: Sat, 06 Sep 2014 23:41:47 -0400
Subject: [racket] typed racket cps, state machines
In-Reply-To: <F3B7C386-F66B-41D2-9EB9-98BFA6439CD9@ccs.neu.edu>
References: <5409D2E8.2070903@memebeam.org>
 <02BB644D-7722-412C-8319-BD440CC03561@ccs.neu.edu>
 <540A13A9.50204@memebeam.org>
 <F698721A-1684-49F1-B30D-9FF38731D838@ccs.neu.edu>
 <540B487E.10705@memebeam.org>
 <4DB683A6-0A63-43DE-AA7A-30C6CA34816C@ccs.neu.edu>
 <F3B7C386-F66B-41D2-9EB9-98BFA6439CD9@ccs.neu.edu>
Message-ID: <540BD3FB.8080308@memebeam.org>

On 09/06/2014 02:00 PM, Matthias Felleisen wrote:
>
> p.s. I have figured out the syntax. I am not sure what you want with
> 'loop' but in any case, this all type checks.

It is just a very degenerate case of my problems, with everything else 
gone, an infinite loop nested in the most basic polymorphic type I could 
think of.

> On Sep 6, 2014, at 1:58 PM, Matthias Felleisen wrote:
>> This type checks fine in v6.1 though I can't read the dang syntax
>> and I thought I was someone who should be able to do just that :-)

LOL

I had been using, Racket v6.0.0.5, but here is a machine with 6.0.1, and 
actually I just downloaded 6.1. All give the same answer:

$ racket --version
Welcome to Racket v6.1.

$ cat loop.rkt
#lang typed/racket/base

(define #:? (V) (fn (r : V)) : V
  (define (loop) : Void (loop))
  r)

$ racket loop.rkt
loop.rkt:4:10: Type Checker: insufficient type information to typecheck. 
please add more type annotations
   in: loop
   context...:
    /home/acarrico/src/racket/collects/racket/private/map.rkt:21:13: map
 
/home/acarrico/src/racket/share/pkgs/typed-racket-lib/typed-racket/typecheck/tc-let-unit.rkt:147:26


...etc.

-- 
Anthony Carrico

From matthias at ccs.neu.edu  Sat Sep  6 23:44:05 2014
From: matthias at ccs.neu.edu (Matthias Felleisen)
Date: Sat, 6 Sep 2014 23:44:05 -0400
Subject: [racket] typed racket cps, state machines
In-Reply-To: <540BD3FB.8080308@memebeam.org>
References: <5409D2E8.2070903@memebeam.org>
 <02BB644D-7722-412C-8319-BD440CC03561@ccs.neu.edu>
 <540A13A9.50204@memebeam.org>
 <F698721A-1684-49F1-B30D-9FF38731D838@ccs.neu.edu>
 <540B487E.10705@memebeam.org>
 <4DB683A6-0A63-43DE-AA7A-30C6CA34816C@ccs.neu.edu>
 <F3B7C386-F66B-41D2-9EB9-98BFA6439CD9@ccs.neu.edu>
 <540BD3FB.8080308@memebeam.org>
Message-ID: <CA202D56-1E70-4B3B-9D86-C59B6B2B0EB7@ccs.neu.edu>


I am working with git head. -- Matthias



On Sep 6, 2014, at 11:41 PM, Anthony Carrico wrote:

> On 09/06/2014 02:00 PM, Matthias Felleisen wrote:
>> 
>> p.s. I have figured out the syntax. I am not sure what you want with
>> 'loop' but in any case, this all type checks.
> 
> It is just a very degenerate case of my problems, with everything else gone, an infinite loop nested in the most basic polymorphic type I could think of.
> 
>> On Sep 6, 2014, at 1:58 PM, Matthias Felleisen wrote:
>>> This type checks fine in v6.1 though I can't read the dang syntax
>>> and I thought I was someone who should be able to do just that :-)
> 
> LOL
> 
> I had been using, Racket v6.0.0.5, but here is a machine with 6.0.1, and actually I just downloaded 6.1. All give the same answer:
> 
> $ racket --version
> Welcome to Racket v6.1.
> 
> $ cat loop.rkt
> #lang typed/racket/base
> 
> (define #:? (V) (fn (r : V)) : V
> (define (loop) : Void (loop))
> r)
> 
> $ racket loop.rkt
> loop.rkt:4:10: Type Checker: insufficient type information to typecheck. please add more type annotations
>  in: loop
>  context...:
>   /home/acarrico/src/racket/collects/racket/private/map.rkt:21:13: map
> /home/acarrico/src/racket/share/pkgs/typed-racket-lib/typed-racket/typecheck/tc-let-unit.rkt:147:26
> 
> 
> ...etc.
> 
> -- 
> Anthony Carrico



From greghendershott at gmail.com  Sat Sep  6 23:50:12 2014
From: greghendershott at gmail.com (Greg Hendershott)
Date: Sat, 6 Sep 2014 23:50:12 -0400
Subject: [racket] recursion style question
In-Reply-To: <F3AF6EB7-306F-4D2B-A69E-F48EEB1B5EB3@gmail.com>
References: <F3AF6EB7-306F-4D2B-A69E-F48EEB1B5EB3@gmail.com>
Message-ID: <CAGspUn0XR2JuW9GapWzxq3_c3zgvmSJt9UTmE-QScHZYahbWEg@mail.gmail.com>

Well, I think you answer your own question with good reasons in the
last paragraph. :)

If `acc` is an implementation detail, let's not expose it as a parameter.[1]

At least, let's not do this for a function provided by a module.
Especially not a function with a contract and/or documentation.

But if it's a helper function that's not provided? Especially in a
small module? Then I'd say it doesn't matter so much.


[1]: Not even a parameter with a default value of '(), which is a 3rd
possibility -- a way to try to have your cake and eat it, too. You
could even have provide/contract insist that the value be '(), for
"outsiders"... but inside the module the contract is N/A. I actually
did this once upon a time. I changed my mind when I started to write
documentation for the function, got to that parameter, and realized...
yeah, nope. :)


On Sat, Sep 6, 2014 at 9:55 PM, Kevin Forchione <lysseus at gmail.com> wrote:
> Hi guys,
> Which is preferable?
>
> (define (foo let ? (acc empty)) ? (foo (rest let) ... (cons ?. acc))
>
> or
>
> (define (foo lst ...) (let loop ([lst lst] ?[acc empty]) ? (loop (rest let) ? (cons ?. acc)))
>
> On the one hand we eliminate the named let construct, but on the other hand we expose the acc in the function / contract.  If the acc is not something the caller would ever need to know about, should we hide it inside the definition?
>
> -Kevin
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users


From m4burns at uwaterloo.ca  Sun Sep  7 00:41:51 2014
From: m4burns at uwaterloo.ca (Marc Burns)
Date: Sun, 7 Sep 2014 00:41:51 -0400
Subject: [racket] recursion style question
In-Reply-To: <CAGspUn0XR2JuW9GapWzxq3_c3zgvmSJt9UTmE-QScHZYahbWEg@mail.gmail.com>
References: <F3AF6EB7-306F-4D2B-A69E-F48EEB1B5EB3@gmail.com>
 <CAGspUn0XR2JuW9GapWzxq3_c3zgvmSJt9UTmE-QScHZYahbWEg@mail.gmail.com>
Message-ID: <04AE2352-877E-4616-B878-ED932B04E7F0@uwaterloo.ca>

As Greg explained, the named let form simplifies the interface to other code. It is much more common in practice.

> On Sep 6, 2014, at 23:50, Greg Hendershott <greghendershott at gmail.com> wrote:
> 
> Well, I think you answer your own question with good reasons in the
> last paragraph. :)
> 
> If `acc` is an implementation detail, let's not expose it as a parameter.[1]
> 
> At least, let's not do this for a function provided by a module.
> Especially not a function with a contract and/or documentation.
> 
> But if it's a helper function that's not provided? Especially in a
> small module? Then I'd say it doesn't matter so much.
> 
> 
> [1]: Not even a parameter with a default value of '(), which is a 3rd
> possibility -- a way to try to have your cake and eat it, too. You
> could even have provide/contract insist that the value be '(), for
> "outsiders"... but inside the module the contract is N/A. I actually
> did this once upon a time. I changed my mind when I started to write
> documentation for the function, got to that parameter, and realized...
> yeah, nope. :)
> 
> 
>> On Sat, Sep 6, 2014 at 9:55 PM, Kevin Forchione <lysseus at gmail.com> wrote:
>> Hi guys,
>> Which is preferable?
>> 
>> (define (foo let ? (acc empty)) ? (foo (rest let) ... (cons ?. acc))
>> 
>> or
>> 
>> (define (foo lst ...) (let loop ([lst lst] ?[acc empty]) ? (loop (rest let) ? (cons ?. acc)))
>> 
>> On the one hand we eliminate the named let construct, but on the other hand we expose the acc in the function / contract.  If the acc is not something the caller would ever need to know about, should we hide it inside the definition?
>> 
>> -Kevin
>> ____________________
>>  Racket Users list:
>>  http://lists.racket-lang.org/users
> 
> ____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users


From alexander at knauth.org  Sun Sep  7 08:20:17 2014
From: alexander at knauth.org (Alexander D. Knauth)
Date: Sun, 7 Sep 2014 08:20:17 -0400
Subject: [racket] typed racket cps, state machines
In-Reply-To: <540BD3FB.8080308@memebeam.org>
References: <5409D2E8.2070903@memebeam.org>
 <02BB644D-7722-412C-8319-BD440CC03561@ccs.neu.edu>
 <540A13A9.50204@memebeam.org>
 <F698721A-1684-49F1-B30D-9FF38731D838@ccs.neu.edu>
 <540B487E.10705@memebeam.org>
 <4DB683A6-0A63-43DE-AA7A-30C6CA34816C@ccs.neu.edu>
 <F3B7C386-F66B-41D2-9EB9-98BFA6439CD9@ccs.neu.edu>
 <540BD3FB.8080308@memebeam.org>
Message-ID: <1401901D-66FC-4CB8-B1D7-872E592B4624@knauth.org>


On Sep 6, 2014, at 11:41 PM, Anthony Carrico <acarrico at memebeam.org> wrote:

> On 09/06/2014 02:00 PM, Matthias Felleisen wrote:
>> 
>> p.s. I have figured out the syntax. I am not sure what you want with
>> 'loop' but in any case, this all type checks.
> 
> It is just a very degenerate case of my problems, with everything else gone, an infinite loop nested in the most basic polymorphic type I could think of.
> 
>> On Sep 6, 2014, at 1:58 PM, Matthias Felleisen wrote:
>>> This type checks fine in v6.1 though I can't read the dang syntax
>>> and I thought I was someone who should be able to do just that :-)
> 
> LOL
> 
> I had been using, Racket v6.0.0.5, but here is a machine with 6.0.1, and actually I just downloaded 6.1. All give the same answer:
> 
> $ racket --version
> Welcome to Racket v6.1.
> 
> $ cat loop.rkt
> #lang typed/racket/base
> 
> (define #:? (V) (fn (r : V)) : V
> (define (loop) : Void (loop))
> r)
> 
> $ racket loop.rkt
> loop.rkt:4:10: Type Checker: insufficient type information to typecheck. please add more type annotations
>  in: loop
>  context...:
>   /home/acarrico/src/racket/collects/racket/private/map.rkt:21:13: map
> /home/acarrico/src/racket/share/pkgs/typed-racket-lib/typed-racket/typecheck/tc-let-unit.rkt:147:26

Well, that doesn?t type check for me on v6.1.0.5, but using (: loop : [-> Void]) fixes it:
#lang typed/racket/base

(: fn : (All (V) [V -> V]))
(define (fn r)
  (: loop : [-> Void])
  (define (loop) (loop))
  r)


> 
> 
> ...etc.
> 
> -- 
> Anthony Carrico
> ____________________
> Racket Users list:
> http://lists.racket-lang.org/users



From mflatt at cs.utah.edu  Sun Sep  7 08:50:12 2014
From: mflatt at cs.utah.edu (Matthew Flatt)
Date: Sun, 7 Sep 2014 06:50:12 -0600
Subject: [racket] v6.1 freeze on Windows: help needed
Message-ID: <20140907125014.05B6665018E@mail-svr1.cs.utah.edu>

Several users have reported problems with Racket v6.1 on Windows where
DrRacket freezes. The reports have several details in common, I think:

 * When DrRacket freezes, it uses 0% CPU.

 * On machine where the problem happens, it happens consistently, and
   it doesn't take long after starting DrRacket.

 * No particular action within DrRacket is identified as a trigger, but
   in some cases, it might be related to switching the foreground
   window.

 * The previous release, v6.0.1, works fine on the same machine and
   configuration.

The Windows variant doesn't seem to matter; we've received reports for
64-bit Windows 7 and 32-bit Windows XP, at least.

Unfortunately, I haven't been able to provoke the freezing behavior on
any of the Windows machines/installations that I have tried (about 6 of
them, using various OS flavors, many different locations and ways of
installing, administrator vs. non-administrator accounts, etc.). I
think I must be missing a crucial ingredient --- maybe a particular
kind of network filesystem or a particular other piece of software.

If anyone else can provoke the freezing behavior, I'd appreciate
information about your configuration and observations. Most useful,
probably, would be a stack trace obtained by attaching Visual Studio to
the process when DrRacket is frozen. Otherwise, if DrRacket freezes on
some machines and not others, anything that you can say about the
difference in the machines could be helpful.

Thanks!


From jensaxel at soegaard.net  Sun Sep  7 10:02:04 2014
From: jensaxel at soegaard.net (=?UTF-8?Q?Jens_Axel_S=C3=B8gaard?=)
Date: Sun, 7 Sep 2014 16:02:04 +0200
Subject: [racket] An FFI question
Message-ID: <CABefVgyf5LcGLkB879suxJFCeVDjn9Cd4MqSXo0LXJ8Xo3u3BA@mail.gmail.com>

Hi All,

While reading the source of draw-lib/racket/draw/unsafe/cairo.rkt
which contains the bindings of Cairo, I noticed the following note:

;; ALLOCATION NOTE: drawing to a Cairo surface might call back to
;; Racket, because a drawing suface might be a PDF or SVG file
;; that is written through a callback to Racket. Consequently,
;; all GC-allocated arguments to Cairo functions must be allocated
;; a 'atomic-interior, so that they do not move in case of a GC.

This explains why _cairo_matrix_t is defined as:

(define-cstruct _cairo_matrix_t ([xx _double*] [yx _double*]  [xy
_double*] [yy _double*] [x0 _double*] [y0 _double*])
   #:malloc-mode 'atomic-interior)

However _cairo_glyph_t is defined as:

(define-cstruct _cairo_glyph_t ([index _long] [x _double*] [y _double*]))

Why is the    #:malloc-mode 'atomic-interior not needed here?

https://github.com/plt/racket/blob/c18f6e8d6dd2afc095be6764c8276b30f7c8da39/pkgs/draw-pkgs/draw-lib/racket/draw/unsafe/cairo.rkt#L16

--
Jens Axel S?gaard


From mflatt at cs.utah.edu  Sun Sep  7 10:18:45 2014
From: mflatt at cs.utah.edu (Matthew Flatt)
Date: Sun, 7 Sep 2014 08:18:45 -0600
Subject: [racket] An FFI question
In-Reply-To: <CABefVgyf5LcGLkB879suxJFCeVDjn9Cd4MqSXo0LXJ8Xo3u3BA@mail.gmail.com>
References: <CABefVgyf5LcGLkB879suxJFCeVDjn9Cd4MqSXo0LXJ8Xo3u3BA@mail.gmail.com>
Message-ID: <20140907141847.609E265018E@mail-svr1.cs.utah.edu>

The function that consumes a `_cairo_glyph_t-pointer` is passed an
array of glyphs, instead of a single one. It would be fine to add the
`#:malloc-mode 'atomic-interior` annotation, but that annotation will
have no effect on the one use, because the array is not allocated with
`make-cairo_glyph_t`.

At Sun, 7 Sep 2014 16:02:04 +0200, Jens Axel S?gaard wrote:
> Hi All,
> 
> While reading the source of draw-lib/racket/draw/unsafe/cairo.rkt
> which contains the bindings of Cairo, I noticed the following note:
> 
> ;; ALLOCATION NOTE: drawing to a Cairo surface might call back to
> ;; Racket, because a drawing suface might be a PDF or SVG file
> ;; that is written through a callback to Racket. Consequently,
> ;; all GC-allocated arguments to Cairo functions must be allocated
> ;; a 'atomic-interior, so that they do not move in case of a GC.
> 
> This explains why _cairo_matrix_t is defined as:
> 
> (define-cstruct _cairo_matrix_t ([xx _double*] [yx _double*]  [xy
> _double*] [yy _double*] [x0 _double*] [y0 _double*])
>    #:malloc-mode 'atomic-interior)
> 
> However _cairo_glyph_t is defined as:
> 
> (define-cstruct _cairo_glyph_t ([index _long] [x _double*] [y _double*]))
> 
> Why is the    #:malloc-mode 'atomic-interior not needed here?
> 
> https://github.com/plt/racket/blob/c18f6e8d6dd2afc095be6764c8276b30f7c8da39/pkgs
> /draw-pkgs/draw-lib/racket/draw/unsafe/cairo.rkt#L16
> 
> --
> Jens Axel S?gaard
> 
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users


From edu500ac at gmail.com  Sun Sep  7 12:54:01 2014
From: edu500ac at gmail.com (Eduardo Costa)
Date: Sun, 7 Sep 2014 13:54:01 -0300
Subject: [racket] usb/serial port
Message-ID: <CAKxnX2J4V0QYPYskOVg+_e5S7Qr0OLasLH60OyCCQHruf+cCKw@mail.gmail.com>

Racketeers.

I work with a small research team that is building a navigation system for
blind people. The hardware is similar to the one used in robotic vision:
Cameras, sonar, GPS, laser measurement sensors, and the like. There are two
sonar systems: (a) A sonar belt around the head to avoid hitting walls,
light poles, etc. This sonar belt has 8 sonars. There is also a pair of
sonars to avoid irregularities on the sidestep. This watch your step sonar
also detects curbs, and activates a camera in charge of recognising zebra
crossings.

The Artificial Intelligence programs are mostly written in Lush. In fact,
the idea is to use the deep learning libraries and image processing
facilities provided in Lush. There are also Common Lisp libraries, like
cl-viola-jones and opticl. The probability of hitting are calculated by
using the Rasch method for each Artificial Intelligence library. Then the
probabilities are combined through the Condorcet's theorem. Once one has
the probability distribution, it is easy to design a gh-filter.

For the time being, the servlets are written in Racket, except for the
Rasch model that is written in Bigloo. As Junia reported to this list, the
Rasch model written in Racket stopped working. Therefore Junia ported the
Rasch model to Bigloo to perform the calibration of the system.

The cameras use the image capture provided by Lush. However, our team
designed custom hardware for the GPS, sonars and other sensors. This system
sends the signal as a string via serial USB to a computer (OS-X or Linux),
that feeds the servlets. Here is how I use it in OS-X 10.9.4:

1- I connect the sonar to a USB port. In this example, I will use the watch
your step sonar. The sonar is supposed to rotate, in order to create a
synthetic aperture  sonar. However I fixed it to simplify the example.

2- I discover where is the sonar:

Eduardos-Mac-mini:~ edu500ac$ ls /dev/tty.*

/dev/tty.Bluetooth-Incoming-Port /dev/tty.usbmodem1421

/dev/tty.Bluetooth-Modem


3- screen /dev/tty.usbmodem1421 9600


4- If I send D, I get a single reading of the sonar. If I send C, I get
continuous reading.


D

Distancia = 00141 mm

D

Distancia = 00176 mm

Dist. Error

D

Distancia = 01261 mm

D

Distancia = 01716 mm


Distancia = 01452 mm

Distancia = 01445 mm

Distancia = 01447 mm

Distancia = 01170 mm

Distancia = 01713 mm

Distancia = 00011 mm

Distancia = 00085 mm

Distancia = 00119 mm

Distancia = 02698 mm

Distancia = 00149 mm

Distancia = 01852 mm

Distancia = 01851 mm


As you can see, we are trying to employ pr?t ? porter systems, like Lush
(provided by Yann Lecun), deep learning (eblearn, provided by Koray
Kavukcuoglu), opticl (provided by Cyrus Harmon), etc.


Is there a Racket library for USB/serial communication with our equipment?
I mean, is there something like OS-X screen written in Racket? I found the
following answer to a similar question:


I am not sure if there a library specifically written in racket for
reading/writing to serial ports, however, you can read/write to character
devices using standard i/o routines.


Could a member of this list elaborate on this answer?  For instance, how
can I open a device from Racket? I tried:


> (define out (open-output-file "/dev/tty.usbmodem1421"))

. . ../../Applications/Racket
v6.1/share/pkgs/drracket/drracket/private/rep.rkt:1123:24:
open-output-file: cannot open output file

  path: /dev/tty.usbmodem1421

  system error: Permission denied; errno=13

>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140907/1823b4bb/attachment.html>

From mb at mbtype.com  Sun Sep  7 14:18:53 2014
From: mb at mbtype.com (Matthew Butterick)
Date: Sun, 7 Sep 2014 11:18:53 -0700
Subject: [racket] How `scribble` resolves relative paths
In-Reply-To: <20140902045905.B477B65018C@mail-svr1.cs.utah.edu>
References: <06C12393-DBFA-4961-BC81-3B9C6CB96317@mbtype.com>
 <20140902045905.B477B65018C@mail-svr1.cs.utah.edu>
Message-ID: <1EAAAD42-3C2C-4DCA-B343-104BCAC1EC47@mbtype.com>

I made a rudimentary helper macro called `image/rp` that upgrades a relative path-string into a runtime path:

(define-syntax (image/rp stx)
  (syntax-case stx ()
    [(_ name) #'(image/rp name 1.0)]
    [(_ name scale) #'(begin
                        (define-runtime-path rp name)
                        (image rp #:scale scale))]))

Of course, the problem is that it disregards the interface for `image`, which includes keyword arguments.

Is there a standard idiom for matching keyword arguments with `syntax-case` ...

1) To capture all keyword arguments and pass them through unaltered?
2) To unpack keyword arguments, alter them, and reapply them?



On Sep 1, 2014, at 9:59 PM, Matthew Flatt <mflatt at cs.utah.edu> wrote:

> I think the docs are wrong. While `raco setup` does set the current
> directory, `scribble` does not.
> 
> It's better to use `define-runtime-path` instead of relying on the
> current directory, since that composes better.
> 
> At Mon, 1 Sep 2014 18:20:05 -0700, Matthew Butterick wrote:
>> I am confused by this statement in the docs vs. the actual behavior:
>> 
>> The docs for `image` say that "If path is a relative path, it is relative to 
>> the current directory, which is set by `raco setup` and `scribble` to the 
>> directory of the main document file." (I'm using `image` as an example but I've 
>> seen the same problem with CSS files.)
>> 
>> Suppose I have a project set up like this:
>> 
>> /project
>>    /scribblings
>>        main.scrbl
>>        sample.gif
>> 
>> And "main.scrbl" contains this:
>> 
>> #lang scribble/manual
>> @image["sample.gif"] 
>> 
>> If I do this:
>>> cd project
>>> cd scribblings
>>> scribble main.html
>> 
>> The HTML will render fine.
>> 
>> But if I do this:
>> 
>>> cd project
>>> scribble --dest-name doc scribblings/main.scrbl
>> 
>> I'll get this error:
>> 
>> open-input-file: cannot open input file
>>  path: /project/sample.gif
>> 
>> Thus, the relative path "sample.gif" is being resolved relative to the 
>> directory where I invoke the `scribble` command, not the "directory of the main 
>> document file", which in both cases seems like it should be  
>> /project/scribblings (i.e., the directory containing "main.scrbl").
>> 
>> What am I misunderstanding?
>> 
>> 
>> 
>> 
>> 
>> 
>> [1] 
>> http://docs.racket-lang.org/scribble/base.html?q=image#%28def._%28%28lib._scribb
>> le%2Fbase..rkt%29._image%29%29____________________
>>  Racket Users list:
>>  http://lists.racket-lang.org/users



From greghendershott at gmail.com  Sun Sep  7 17:11:32 2014
From: greghendershott at gmail.com (Greg Hendershott)
Date: Sun, 7 Sep 2014 17:11:32 -0400
Subject: [racket] How `scribble` resolves relative paths
In-Reply-To: <1EAAAD42-3C2C-4DCA-B343-104BCAC1EC47@mbtype.com>
References: <06C12393-DBFA-4961-BC81-3B9C6CB96317@mbtype.com>
 <20140902045905.B477B65018C@mail-svr1.cs.utah.edu>
 <1EAAAD42-3C2C-4DCA-B343-104BCAC1EC47@mbtype.com>
Message-ID: <CAGspUn1war2j5vZf4_kYTw9Pph0i5ZL5SP78+FK7fWB-u2H1eQ@mail.gmail.com>

If you don't mind requiring that `name` be supplied first, you could
simply pass the remaining args through unchanged:

(define-syntax (image/rp stx)
  (syntax-case stx ()
    [(_ name xs ...)
     #'(begin
         (define-runtime-path id name)
         (image id xs ...))]))

It's not ideal because normally it should be fine to put the keyword
arguments before the positional one. However if this is just a private
helper macro it's probably fine.  If you wanted, you could have the
macro sanity check that `name` is not keyword? -- or more to the
point, that it _is_ identifier?. (That kind of stuff is nice to
declare with syntax-parse; it make a nice error message for you.)


Separately, I think one gotcha with your original macro will be if you
use it more than once? I think you need to provide a fresh id to each
define-runtime-path, like so:

(define-syntax (image/rp stx)
  (syntax-case stx ()
    [(_ name xs ...)
     (with-syntax ([id (generate-temporary)])
       #'(begin
           (define-runtime-path id name)
           (image id xs ...)))]))


On Sun, Sep 7, 2014 at 2:18 PM, Matthew Butterick <mb at mbtype.com> wrote:
> I made a rudimentary helper macro called `image/rp` that upgrades a relative path-string into a runtime path:
>
> (define-syntax (image/rp stx)
>   (syntax-case stx ()
>     [(_ name) #'(image/rp name 1.0)]
>     [(_ name scale) #'(begin
>                         (define-runtime-path rp name)
>                         (image rp #:scale scale))]))
>
> Of course, the problem is that it disregards the interface for `image`, which includes keyword arguments.
>
> Is there a standard idiom for matching keyword arguments with `syntax-case` ...
>
> 1) To capture all keyword arguments and pass them through unaltered?
> 2) To unpack keyword arguments, alter them, and reapply them?
>
>
>
> On Sep 1, 2014, at 9:59 PM, Matthew Flatt <mflatt at cs.utah.edu> wrote:
>
>> I think the docs are wrong. While `raco setup` does set the current
>> directory, `scribble` does not.
>>
>> It's better to use `define-runtime-path` instead of relying on the
>> current directory, since that composes better.
>>
>> At Mon, 1 Sep 2014 18:20:05 -0700, Matthew Butterick wrote:
>>> I am confused by this statement in the docs vs. the actual behavior:
>>>
>>> The docs for `image` say that "If path is a relative path, it is relative to
>>> the current directory, which is set by `raco setup` and `scribble` to the
>>> directory of the main document file." (I'm using `image` as an example but I've
>>> seen the same problem with CSS files.)
>>>
>>> Suppose I have a project set up like this:
>>>
>>> /project
>>>    /scribblings
>>>        main.scrbl
>>>        sample.gif
>>>
>>> And "main.scrbl" contains this:
>>>
>>> #lang scribble/manual
>>> @image["sample.gif"]
>>>
>>> If I do this:
>>>> cd project
>>>> cd scribblings
>>>> scribble main.html
>>>
>>> The HTML will render fine.
>>>
>>> But if I do this:
>>>
>>>> cd project
>>>> scribble --dest-name doc scribblings/main.scrbl
>>>
>>> I'll get this error:
>>>
>>> open-input-file: cannot open input file
>>>  path: /project/sample.gif
>>>
>>> Thus, the relative path "sample.gif" is being resolved relative to the
>>> directory where I invoke the `scribble` command, not the "directory of the main
>>> document file", which in both cases seems like it should be
>>> /project/scribblings (i.e., the directory containing "main.scrbl").
>>>
>>> What am I misunderstanding?
>>>
>>>
>>>
>>>
>>>
>>>
>>> [1]
>>> http://docs.racket-lang.org/scribble/base.html?q=image#%28def._%28%28lib._scribb
>>> le%2Fbase..rkt%29._image%29%29____________________
>>>  Racket Users list:
>>>  http://lists.racket-lang.org/users
>
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users

From acarrico at memebeam.org  Sun Sep  7 19:52:17 2014
From: acarrico at memebeam.org (Anthony Carrico)
Date: Sun, 07 Sep 2014 19:52:17 -0400
Subject: [racket] typed racket cps, state machines
In-Reply-To: <1401901D-66FC-4CB8-B1D7-872E592B4624@knauth.org>
References: <5409D2E8.2070903@memebeam.org>
 <02BB644D-7722-412C-8319-BD440CC03561@ccs.neu.edu>
 <540A13A9.50204@memebeam.org>
 <F698721A-1684-49F1-B30D-9FF38731D838@ccs.neu.edu>
 <540B487E.10705@memebeam.org>
 <4DB683A6-0A63-43DE-AA7A-30C6CA34816C@ccs.neu.edu>
 <F3B7C386-F66B-41D2-9EB9-98BFA6439CD9@ccs.neu.edu>
 <540BD3FB.8080308@memebeam.org>
 <1401901D-66FC-4CB8-B1D7-872E592B4624@knauth.org>
Message-ID: <540CEFB1.4020806@memebeam.org>

On 09/05/2014 03:48 PM, Anthony Carrico wrote:> On 09/05/2014 11:48 AM,
> Apparently. I was trying to figure out if this is something I don't
> know how to annotate properly, or ...

and then,

On 09/07/2014 08:20 AM, Alexander D. Knauth wrote:
> (: fn : (All (V) [V -> V]))
> (define (fn r)
>   (: loop : [-> Void])
>   (define (loop) (loop))
>   r)

Awesome! It looks like I was bumping up against a bug, or limitation in
the syntax that I was using to annotate internal definitions. I need to
go back through this thread, and see if this fixes all my examples.
Looks very promising. Thank you very much.

-- 
Anthony Carrico


-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 181 bytes
Desc: OpenPGP digital signature
URL: <http://lists.racket-lang.org/users/archive/attachments/20140907/9a047bf3/attachment-0001.sig>

From mb at mbtype.com  Sun Sep  7 19:57:30 2014
From: mb at mbtype.com (Matthew Butterick)
Date: Sun, 7 Sep 2014 16:57:30 -0700
Subject: [racket] How `scribble` resolves relative paths
In-Reply-To: <CAGspUn1war2j5vZf4_kYTw9Pph0i5ZL5SP78+FK7fWB-u2H1eQ@mail.gmail.com>
References: <06C12393-DBFA-4961-BC81-3B9C6CB96317@mbtype.com>
 <20140902045905.B477B65018C@mail-svr1.cs.utah.edu>
 <1EAAAD42-3C2C-4DCA-B343-104BCAC1EC47@mbtype.com>
 <CAGspUn1war2j5vZf4_kYTw9Pph0i5ZL5SP78+FK7fWB-u2H1eQ@mail.gmail.com>
Message-ID: <2D3A141D-8E29-484A-A60E-09AAC2A5DA7E@mbtype.com>

> If you don't mind requiring that `name` be supplied first, you could
> simply pass the remaining args through unchanged:

For my purposes that's fine, though it's still a bit of a cheat. And cheating may not be an option next time ...


> Separately, I think one gotcha with your original macro will be if you
> use it more than once? I think you need to provide a fresh id to each
> define-runtime-path, like so:

I thought that would be the case too, though it does work without `generate-temporaries`. (Looking at the macro expansion, it seems each time the macro operates, it assigns a unique subscript to each "rp" identifier so they don't collide). But I can't tell if that's just lucky behavior. `Generate-temporaries` is clearly the right way to do things. 

Thanks.



On Sep 7, 2014, at 2:11 PM, Greg Hendershott <greghendershott at gmail.com> wrote:

> If you don't mind requiring that `name` be supplied first, you could
> simply pass the remaining args through unchanged:
> 
> (define-syntax (image/rp stx)
>  (syntax-case stx ()
>    [(_ name xs ...)
>     #'(begin
>         (define-runtime-path id name)
>         (image id xs ...))]))
> 
> It's not ideal because normally it should be fine to put the keyword
> arguments before the positional one. However if this is just a private
> helper macro it's probably fine.  If you wanted, you could have the
> macro sanity check that `name` is not keyword? -- or more to the
> point, that it _is_ identifier?. (That kind of stuff is nice to
> declare with syntax-parse; it make a nice error message for you.)
> 
> 
> Separately, I think one gotcha with your original macro will be if you
> use it more than once? I think you need to provide a fresh id to each
> define-runtime-path, like so:
> 
> (define-syntax (image/rp stx)
>  (syntax-case stx ()
>    [(_ name xs ...)
>     (with-syntax ([id (generate-temporary)])
>       #'(begin
>           (define-runtime-path id name)
>           (image id xs ...)))]))
> 
> 
> On Sun, Sep 7, 2014 at 2:18 PM, Matthew Butterick <mb at mbtype.com> wrote:
>> I made a rudimentary helper macro called `image/rp` that upgrades a relative path-string into a runtime path:
>> 
>> (define-syntax (image/rp stx)
>>  (syntax-case stx ()
>>    [(_ name) #'(image/rp name 1.0)]
>>    [(_ name scale) #'(begin
>>                        (define-runtime-path rp name)
>>                        (image rp #:scale scale))]))
>> 
>> Of course, the problem is that it disregards the interface for `image`, which includes keyword arguments.
>> 
>> Is there a standard idiom for matching keyword arguments with `syntax-case` ...
>> 
>> 1) To capture all keyword arguments and pass them through unaltered?
>> 2) To unpack keyword arguments, alter them, and reapply them?
>> 
>> 
>> 
>> On Sep 1, 2014, at 9:59 PM, Matthew Flatt <mflatt at cs.utah.edu> wrote:
>> 
>>> I think the docs are wrong. While `raco setup` does set the current
>>> directory, `scribble` does not.
>>> 
>>> It's better to use `define-runtime-path` instead of relying on the
>>> current directory, since that composes better.
>>> 
>>> At Mon, 1 Sep 2014 18:20:05 -0700, Matthew Butterick wrote:
>>>> I am confused by this statement in the docs vs. the actual behavior:
>>>> 
>>>> The docs for `image` say that "If path is a relative path, it is relative to
>>>> the current directory, which is set by `raco setup` and `scribble` to the
>>>> directory of the main document file." (I'm using `image` as an example but I've
>>>> seen the same problem with CSS files.)
>>>> 
>>>> Suppose I have a project set up like this:
>>>> 
>>>> /project
>>>>   /scribblings
>>>>       main.scrbl
>>>>       sample.gif
>>>> 
>>>> And "main.scrbl" contains this:
>>>> 
>>>> #lang scribble/manual
>>>> @image["sample.gif"]
>>>> 
>>>> If I do this:
>>>>> cd project
>>>>> cd scribblings
>>>>> scribble main.html
>>>> 
>>>> The HTML will render fine.
>>>> 
>>>> But if I do this:
>>>> 
>>>>> cd project
>>>>> scribble --dest-name doc scribblings/main.scrbl
>>>> 
>>>> I'll get this error:
>>>> 
>>>> open-input-file: cannot open input file
>>>> path: /project/sample.gif
>>>> 
>>>> Thus, the relative path "sample.gif" is being resolved relative to the
>>>> directory where I invoke the `scribble` command, not the "directory of the main
>>>> document file", which in both cases seems like it should be
>>>> /project/scribblings (i.e., the directory containing "main.scrbl").
>>>> 
>>>> What am I misunderstanding?
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>>> [1]
>>>> http://docs.racket-lang.org/scribble/base.html?q=image#%28def._%28%28lib._scribb
>>>> le%2Fbase..rkt%29._image%29%29____________________
>>>> Racket Users list:
>>>> http://lists.racket-lang.org/users
>> 
>> 
>> ____________________
>>  Racket Users list:
>>  http://lists.racket-lang.org/users



From greghendershott at gmail.com  Sun Sep  7 21:02:45 2014
From: greghendershott at gmail.com (Greg Hendershott)
Date: Sun, 7 Sep 2014 21:02:45 -0400
Subject: [racket] How `scribble` resolves relative paths
In-Reply-To: <2D3A141D-8E29-484A-A60E-09AAC2A5DA7E@mbtype.com>
References: <06C12393-DBFA-4961-BC81-3B9C6CB96317@mbtype.com>
 <20140902045905.B477B65018C@mail-svr1.cs.utah.edu>
 <1EAAAD42-3C2C-4DCA-B343-104BCAC1EC47@mbtype.com>
 <CAGspUn1war2j5vZf4_kYTw9Pph0i5ZL5SP78+FK7fWB-u2H1eQ@mail.gmail.com>
 <2D3A141D-8E29-484A-A60E-09AAC2A5DA7E@mbtype.com>
Message-ID: <CAGspUn1VFUTgrrFF+HHMUPQXVhvVYyT+wn20Qecs3EmJjYw3ZQ@mail.gmail.com>

>> Separately, I think one gotcha with your original macro will be if you
>> use it more than once? I think you need to provide a fresh id to each
>> define-runtime-path, like so:
>
> I thought that would be the case too, though it does work without `generate-temporaries`. (Looking at the macro expansion, it seems each time the macro operates, it assigns a unique subscript to each "rp" identifier so they don't collide). But I can't tell if that's just lucky behavior. `Generate-temporaries` is clearly the right way to do things.
>

Hmm, it definitely errors for me. And I see it expand duplicating the
identifier.


>> If you don't mind requiring that `name` be supplied first, you could
>> simply pass the remaining args through unchanged:
>
> For my purposes that's fine, though it's still a bit of a cheat. And cheating may not be an option next time ...

Here's another attempt that cheats a bit less. ;)  I do make the image
file name a required keyword argument, `#:path`:

#lang racket/base

(require (for-syntax racket/base
                     racket/function
                     racket/syntax
                     syntax/parse)
         racket/pretty
         racket/runtime-path
         scribble/base)

(define-syntax (image/rp stx)
  (define-splicing-syntax-class arg
    (pattern (~seq kw:keyword e:expr))
    (pattern (~seq e:expr)))
  (syntax-parse stx
    [(_ arg:arg ...)
     (define all-args (syntax->list #'(arg ...)))
     (define (path-arg stx)
       (syntax-parse stx
         [(#:path path:expr) #'path]
         [_ #f]))
     (with-syntax
         ([id (generate-temporary)]
          [path (ormap path-arg all-args)]
          [((other-args ...) ...) (filter (negate path-arg) all-args)])
       (unless (syntax-e #'path)
         (raise-syntax-error 'image/rp
                             "Required keyword argument #:path is missing"))
       #'(begin
           (define-runtime-path id path)
           (image id other-args ... ...)))]))

Examples:

  (image/rp #:path "here.png")

  (image/rp #:scale 1.0 #:suffixes '() #:path "there.png")))))

;; =>

  (begin
    (define-runtime-path g1 "here.png")
    (image g1))

  (begin
    (define-runtime-path g2 "there.png")
    (image g2 #:scale 1.0 #:suffixes '()))


From f.raimondi at gmail.com  Mon Sep  8 02:04:17 2014
From: f.raimondi at gmail.com (Franco Raimondi)
Date: Mon, 8 Sep 2014 07:04:17 +0100
Subject: [racket] usb/serial port
References: <E981D1EC-AA6B-436A-A1A8-1E69DDB5B548@gmail.com>
Message-ID: <FFFD747F-C181-45CC-982B-1CCDD9A67CEB@gmail.com>

On 7 Sep 2014, at 17:54, Eduardo Costa <edu500ac at gmail.com> wrote:

> 
> I am not sure if there a library specifically written in racket for reading/writing to serial ports, however, you can read/write to character devices using standard i/o routines.
> 
> Could a member of this list elaborate on this answer?  For instance, how can I open a device from Racket? I tried:
> 
>> (define out (open-output-file "/dev/tty.usbmodem1421"))
> . . ../../Applications/Racket v6.1/share/pkgs/drracket/drracket/private/rep.rkt:1123:24: open-output-file: cannot open output file
>  path: /dev/tty.usbmodem1421
>  system error: Permission denied; errno=13
>> 


open-output-file should work, are you sure you have the right permissions? This is what I use to write/read to/from an Arduino board connected using a USB port (where port-name is /dev/tty.usbmodem*" on mac, similar to your case):

        (system "stty -f " port-name " 57600 cs8 cread clocal")

        (set! out (open-output-file port-name #:mode 'binary #:exists 'append))
        (set! in  (open-input-file  port-name #:mode 'binary))
        (file-stream-buffer-mode out 'none)

It works both on mac and linux. I have a separate thread to read the USB port, but it may not be necessary in your case.

Have a look at open-firmata (line 414) of this file: https://bitbucket.org/fraimondi/racket-firmata/src/7707268b4b52aabe336c69d347877459dc8c7ae5/firmata.rkt?at=master if you want more information. This works also under Windows.

F



From dpavlov at ipa.nw.ru  Mon Sep  8 14:17:07 2014
From: dpavlov at ipa.nw.ru (Dmitry Pavlov)
Date: Mon, 08 Sep 2014 22:17:07 +0400
Subject: [racket] Text snip bleaches canvas space that does not belong to it
Message-ID: <540DF2A3.5020707@ipa.nw.ru>

Hello,

I need a text snip on a canvas that paints itself strictly in some defined WxH bounds.
I need no interactive moving or resizing, only interactive editing.
Here is a code I have come up with, which also displays a window with the snip and a grid.
The grid facilitates the visual check of the text snip bounds.

(require racket/gui
          racket/draw)

(define my-pasteboard%
   (class pasteboard%
     (super-new)
     (define/augment (can-select? snip on?)
       #f)
     (define/augment (can-interactive-move? snip)
       #f)
     (define/augment (can-interactive-resize? snip)
       #f)))

(define frame (new frame% (label "Test")))
(define pasteboard (new my-pasteboard%))

(define my-canvas%
   (class editor-canvas%
     (super-new)
     (define/override (on-paint)
       (super on-paint)
       (define dc (send this get-dc))
       ;; Draw a 10x10 grid with cell size 20x20 pixels
       (for ((x (in-range 10))) (send dc draw-line (* x 20) 0 (* x 20) 200))
       (for ((y (in-range 10))) (send dc draw-line 0 (* y 20)  200 (* y 20)))
       )))

(define editor-canvas
   (new my-canvas%
        (editor pasteboard)
        (vertical-inset 0) (horizontal-inset 0)
        (parent frame)
        (style '(no-border no-hscroll no-vscroll))))

(send frame resize 200 200)
(send frame show #t)

(define text-obj (new text%))
(send text-obj insert "Hello" 0)
(define text-snip (new editor-snip%
                      (editor text-obj)
                      (left-margin 0) (right-margin 0) (top-margin 0) (bottom-margin 0)
                      (left-inset 0) (right-inset 0) (top-inset 0) (bottom-inset 0)
                      (with-border? #f)
                      (min-width 40)
                      (max-width 40)
                      (min-height 20)
                      (max-height 20)))

(send pasteboard insert text-snip 40 60)
(send pasteboard set-caret-owner text-snip)


When I click on the snip in the window and try to add a "!" character after
the "Hello" text, I get this:

http://imgur.com/cikwlqh

It is obvious that the text snip (or something associated with it)
expanded its supposed area of rendering by 3-4 pixels in all directions.
And if I do silly things, like pressing "Enter", things get even worse.
Can that be avoided somehow?


Best regards,

Dmitry


From skanaley at gmail.com  Mon Sep  8 17:23:59 2014
From: skanaley at gmail.com (Sean Kanaley)
Date: Mon, 8 Sep 2014 17:23:59 -0400
Subject: [racket] DrRacket 6.1 or Hardware Issue?
Message-ID: <CA+APS0CQ-3qqQvHBCwDQ4zsKR_UOwq7dgDdPF5mK7i9wue2SDw@mail.gmail.com>

After investigating a near complete lock up some more (saved by TTY window
and killing DrRacket after seeing it with 3GB+ residency), it's clearly
correlated with background expansion of an accidentally infinitely looping
macro. A few questions..

1. Is the OS supposed to break and enter an infinite swap loop? If that is
just a fact of life for a program requesting too much memory simultaneously
(and an ever increasing amount), well for example ghc just shuts itself
down claiming "too much memory requested" or something like that.

2. Is this related to my very long time ago issue where I removed a bad 1GB
from the 3rd slot, lazily leaving the 4th slot where it is, causing me to
have 3GB instead of 4, even though it's "supposed to" have 4 based on the
max slot (I have never encountered this sort of system freeze before
however)? Maybe the OS would more gracefully handle the paging?

3. Can DrRacket simply realize it's expanding the same thing and not demand
infinite memory (I don't understand the system on the implementation level,
sorry if that's off base).

The solution, aside from me not writing incorrect macros is to disable
background expansion. Relatedly, DrRacket doesn't seem able to reclaim this
space without being restarted. Closing the file and clicking the garbage
collection thing only recovers say 20MB out of 2GB if I turn off expansion
before the OS locks.

Lastly, a related possible bug I've encountered since 6.1 is the infinite
garbage collection loop, where the OS remains responsive but DrRacket has
to be restarted, but this one I have no idea where it comes from only that
it seems to happen after a while, and fairly reliably so.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140908/796d8b9e/attachment.html>

From m4burns at uwaterloo.ca  Mon Sep  8 18:00:32 2014
From: m4burns at uwaterloo.ca (Marc Burns)
Date: Mon, 8 Sep 2014 18:00:32 -0400
Subject: [racket] DrRacket 6.1 or Hardware Issue?
In-Reply-To: <CA+APS0CQ-3qqQvHBCwDQ4zsKR_UOwq7dgDdPF5mK7i9wue2SDw@mail.gmail.com>
References: <CA+APS0CQ-3qqQvHBCwDQ4zsKR_UOwq7dgDdPF5mK7i9wue2SDw@mail.gmail.com>
Message-ID: <20140908220032.GU27226@happierface.uwaterloo.ca>

1. Generally, requests to allocate memory will succeed even when RAM is
full and the system is deep into swap. Random access over a set of pages
that cannot all fit into RAM results in "thrashing": continuously
swapping pages out to the disk and reading other pages in. This can make
the system unusable, and is independent of Racket. I'm not sure how GHC
decides to terminate with the "too much memory requested" error, but I
speculate it's a self-imposed limit perhaps based on the amount of
accessible RAM.

2. I strongly doubt the issue is related to your memory configuration.
If applications run well at normal memory pressure and the system is
generally stable, the problem is likely caused by plain old swap
thrashing. You could try booting off some Linux install disk and running
memtest86+ to be sure.

3. I think implementing a time and memory limit for each background
expansion attempt is a good idea.

I don't know about the unreclaimable memory. That seems like a bug.

On Mon, Sep 08, 2014 at 05:23:59PM -0400, Sean Kanaley wrote:
> After investigating a near complete lock up some more (saved by TTY window
> and killing DrRacket after seeing it with 3GB+ residency), it's clearly
> correlated with background expansion of an accidentally infinitely looping
> macro. A few questions..
> 
> 1. Is the OS supposed to break and enter an infinite swap loop? If that is
> just a fact of life for a program requesting too much memory simultaneously
> (and an ever increasing amount), well for example ghc just shuts itself
> down claiming "too much memory requested" or something like that.
> 
> 2. Is this related to my very long time ago issue where I removed a bad 1GB
> from the 3rd slot, lazily leaving the 4th slot where it is, causing me to
> have 3GB instead of 4, even though it's "supposed to" have 4 based on the
> max slot (I have never encountered this sort of system freeze before
> however)? Maybe the OS would more gracefully handle the paging?
> 
> 3. Can DrRacket simply realize it's expanding the same thing and not demand
> infinite memory (I don't understand the system on the implementation level,
> sorry if that's off base).
> 
> The solution, aside from me not writing incorrect macros is to disable
> background expansion. Relatedly, DrRacket doesn't seem able to reclaim this
> space without being restarted. Closing the file and clicking the garbage
> collection thing only recovers say 20MB out of 2GB if I turn off expansion
> before the OS locks.
> 
> Lastly, a related possible bug I've encountered since 6.1 is the infinite
> garbage collection loop, where the OS remains responsive but DrRacket has
> to be restarted, but this one I have no idea where it comes from only that
> it seems to happen after a while, and fairly reliably so.

> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users


From matthias at ccs.neu.edu  Mon Sep  8 20:51:35 2014
From: matthias at ccs.neu.edu (Matthias Felleisen)
Date: Mon, 8 Sep 2014 20:51:35 -0400
Subject: [racket] missing solution 20.2.4 ex:fancy-contracts-poly
In-Reply-To: <CAB08-qrbvXik1gT7R_gcH+yUjCaY-bAZQF+T=LWFj7EB2DuMug@mail.gmail.com>
References: <CAB08-qrbvXik1gT7R_gcH+yUjCaY-bAZQF+T=LWFj7EB2DuMug@mail.gmail.com>
Message-ID: <D6253DDD-91FF-48E2-A330-4024EDBF1C5E@ccs.neu.edu>


On Sep 4, 2014, at 12:54 PM, Daniel Bastos wrote:

> A candidate for a solution. (I'm not sure I'm correct regarding the
> function project. The description says "a lists of lists" (which I
> translate to (listof ITEM)) and "a function from lists to Xs", so I
> wonder if the domain of this function-argument must be (listof ITEM)
> or it could be more general such as (listof Y). I decided to imagine
> that the domain of the function-argument would be items of the "list
> of lists" and so I made it into (listof ITEM).


In this context ITEM and X and Y are variables (placeholders) that stand for sets (classes) of data. 

In the following I have translated the purpose statements for these functions into prototype calls: 

> sort: (listof ITEM) (ITEM ITEM -> boolean) -> (listof ITEM)


(sort a-list-of-Xs compare-Xs) : a-list-of-Xs 
  (same Xs indeed, but rearranged) 

> map: ((listof ITEM) -> X) (listof ITEM) -> (listof X)

(map a-function-from-item-on-list-to-X a-list-of-ITEMs) : a-list-of-Xs 



> project: (listof (listof ITEM)) ((listof ITEM) -> (listof X)) -> (listof X)

(project a-list-of-list-of-ITEMs a-function-from-list-of-ITEMs-to-X) : a-list-of-Xs 


> (*) Comparison with exercise 20.2.2


Abstraction works by taking two concrete examples and replacing matching concrete things with variables. 

20.2.2 asks you to come up with contracts for functions that work on one basic kind of data. Use a distinct kind and then apply the abstraction design recipe. You will find that at least one of the above is wrong. 

-- Matthias



From matthias at ccs.neu.edu  Mon Sep  8 20:51:37 2014
From: matthias at ccs.neu.edu (Matthias Felleisen)
Date: Mon, 8 Sep 2014 20:51:37 -0400
Subject: [racket] missing solution 20.1.1 ex:sem-funcs
In-Reply-To: <CAB08-qr3dsphx=M875-g8EBNkrdFN3-BLkvTQnytrZ4i-5uUvg@mail.gmail.com>
References: <CAB08-qr3dsphx=M875-g8EBNkrdFN3-BLkvTQnytrZ4i-5uUvg@mail.gmail.com>
Message-ID: <DB47056E-CF26-45EA-97B5-7935F08A7AB2@ccs.neu.edu>


On Sep 2, 2014, at 11:45 AM, Daniel Bastos wrote:

> A candidate for a solution.
> 
> Exercise 20.1.1. Assume the Definitions window in DrScheme contains
> (define (f x) x). Identify the values among the following expressions:
> 
> (1) (cons f empty)
> (2) (f f)
> (3) (cons f (cons 10 (cons (f 10) empty)))
> 
> Explain why they are values and why the remaining expressions are not
> values.
> 
> Solution. First we consider (1). Here's the relevant part of the
> grammar.
> 
>  <lst> = empty | (cons <val> <lst>)
> 
>  <val> = <boo> | <sym> | <num> | empty | <lst>
>  | <var> (names of defined functions)
>  | <prm>
> 
> So (1) is a value because it is <lst> of the form (cons <var> empty),
> where f is <var> because it is a defined function.
> 
> Let's consider (3) before (2). (3) is a list of either <var> or <num>,
> so (3) is <val> as well.

You checked only one half of the value-grammar for lst. Check the other half, too. Report back whether you want to change the answer or not. -- Matthias



From matthias at ccs.neu.edu  Mon Sep  8 20:51:43 2014
From: matthias at ccs.neu.edu (Matthias Felleisen)
Date: Mon, 8 Sep 2014 20:51:43 -0400
Subject: [racket] missing solution 20.2.2 ex:fancy-contracts
In-Reply-To: <CAB08-qo--gMj5qs2qPjfZWcoNhtdjrREqH3moWtLMUgCY9ZtdA@mail.gmail.com>
References: <CAB08-qo--gMj5qs2qPjfZWcoNhtdjrREqH3moWtLMUgCY9ZtdA@mail.gmail.com>
Message-ID: <9B78EB6A-1772-4811-BFA7-2F8EC03CB6ED@ccs.neu.edu>


Yes. 


On Sep 3, 2014, at 9:30 AM, Daniel Bastos wrote:

> A candidate for a solution.
> 
> Exercise 20.2.2. Formulate contracts for the following functions:
> 
> 1. sort, which consumes a list of numbers and a function that consumes
> two numbers (from the list) and produces a boolean; sort produces a
> list of numbers.
> 
> 2. map, which consumes a function from numbers to numbers and a list
> of numbers; it also produces a list of numbers.
> 
> 3. project, which consumes a list of lists of symbols and a function
> from lists of symbols to symbols; it produces a list of symbols.
> 
> Solution.
> 
> sort: (listof number) (number number -> boolean) -> (listof number)
> 
> map: (number -> number) (listof number) -> (listof number)
> 
> project: (listof (listof symbol)) ((listof symbol) -> symbol) -> (listof symbol)
> ____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users



From matthias at ccs.neu.edu  Mon Sep  8 20:51:39 2014
From: matthias at ccs.neu.edu (Matthias Felleisen)
Date: Mon, 8 Sep 2014 20:51:39 -0400
Subject: [racket] missing solution 20.1.2 ex:syn-funcs
In-Reply-To: <CAB08-qqmrq3jesw_GzLbmR0K6PA8WBbtscezT4TGbooL373hZA@mail.gmail.com>
References: <CAB08-qqmrq3jesw_GzLbmR0K6PA8WBbtscezT4TGbooL373hZA@mail.gmail.com>
Message-ID: <7E897494-BFBC-499C-B189-AFB890468E62@ccs.neu.edu>


On Sep 2, 2014, at 12:05 PM, Daniel Bastos wrote:

> Exercise 20.1.2. Argue why the following sentences are legal
> definitions:
> 
> (define (f x) (x 10))
> 
> (define (f x) f)
> 
> (define (f x y) (x 'a y 'b))
> 
> Solution. The relevant part of the grammar is the following.
> 
>  <def> = (define (<var> <var> ...<var>) <exp>)
>        | (define <var> <exp>)
>        | (define-struct <var> (<var> <var> ...<var>))
> 
> (*) First definition
> 
> The LHS is a list of <var>, since we find f and x as members of the
> list and they're both variables. The RHS is a list consisting of a
> <var> and a <num>.


Small correction. Let's not call (x 10) a list. It's an application. '(x 10) would be a list and this one character is critical. 



> (*) Second definition
> 
> Same LHS as the previous, so we need only check the RHS which is a
> <var>. <var> is a valid form of <exp>, so we have
> 
> (*) Third definition
> 
> The LHS is (<var> <var> <var>), while the RHS is (<var> <sym> <var>
> <sym>), 
> 
> Therefore it's a legal <def>.

OKAY. 

From stevebyan at verizon.net  Tue Sep  9 10:38:57 2014
From: stevebyan at verizon.net (Steve Byan)
Date: Tue, 9 Sep 2014 10:38:57 -0400
Subject: [racket] Cut, copy,
	and paste keybindings not working on OS X with Racket 6.1
Message-ID: <BBFA4751-F7EC-4186-81DA-4C19C6667607@verizon.net>

I just installed Racket 6.1, and noticed that the standard cut, copy, and paste keybindings aren't working on OS X 10.9.4. The Edit->Keybindings->Show All Keybindiings dialog shows that they are set to d:x, d:c, and d:v. 

Undo, which is bound to d:z, works as expected. Cut (Cmd-x) doesn't appear to do anything; paste (Cmd-v) acts like it is being interpreted as esc;v (previous-page), and copy (Cmd-c) acts like it is being interpreted as esc;c (capitalize-word).

Best regards,
-Steve

-- 
Steve Byan <stevebyan at me.com>
Littleton, MA 01460





From stevebyan at verizon.net  Tue Sep  9 10:57:42 2014
From: stevebyan at verizon.net (Steve Byan)
Date: Tue, 9 Sep 2014 10:57:42 -0400
Subject: [racket] Cut, copy,
	and paste keybindings not working on OS X with Racket 6.1
In-Reply-To: <BBFA4751-F7EC-4186-81DA-4C19C6667607@verizon.net>
References: <BBFA4751-F7EC-4186-81DA-4C19C6667607@verizon.net>
Message-ID: <7F488959-21E4-4B31-81EC-67FD00024997@verizon.net>


On Sep 9, 2014, at 10:38 AM, Steve Byan <stevebyan at verizon.net> wrote:

> I just installed Racket 6.1, and noticed that the standard cut, copy, and paste keybindings aren't working on OS X 10.9.4. The Edit->Keybindings->Show All Keybindiings dialog shows that they are set to d:x, d:c, and d:v. 
> 
> Undo, which is bound to d:z, works as expected. Cut (Cmd-x) doesn't appear to do anything; paste (Cmd-v) acts like it is being interpreted as esc;v (previous-page), and copy (Cmd-c) acts like it is being interpreted as esc;c (capitalize-word).

Never mind, I just found the "Treat command key as meta" checkbox in the Preferences dialog. I should have read further in the manual.

Best regards,
-Steve

-- 
Steve Byan <stevebyan at me.com>
Littleton, MA 01460





From geoff at knauth.org  Tue Sep  9 11:32:06 2014
From: geoff at knauth.org (Geoffrey S. Knauth)
Date: Tue, 9 Sep 2014 11:32:06 -0400
Subject: [racket] Emacs keybindings for DrRacket on Win8 platform
Message-ID: <1CBB6D6D-58E7-43CE-9BCA-F9B192A10712@knauth.org>

I usually run DrRacket on *nix platforms.  When I run DrRacket on a Windows computer, how do I get my beloved Emacs keybindings?  I've looked, but can't find the preference.

Thanks,
Geoff



From geoff at knauth.org  Tue Sep  9 11:36:16 2014
From: geoff at knauth.org (Geoffrey S. Knauth)
Date: Tue, 9 Sep 2014 11:36:16 -0400
Subject: [racket] Emacs keybindings for DrRacket on Win8 platform
In-Reply-To: <1CBB6D6D-58E7-43CE-9BCA-F9B192A10712@knauth.org>
References: <1CBB6D6D-58E7-43CE-9BCA-F9B192A10712@knauth.org>
Message-ID: <6DC86009-7345-4BC9-8A3F-0033BBD8759D@knauth.org>

NEVER MIND.  I figured it out.  [Uncheck the Enable keybindings in menus preference.]  --Geoff

On Sep 9, 2014, at 11:32 , Geoffrey S. Knauth <geoff at knauth.org> wrote:

> I usually run DrRacket on *nix platforms.  When I run DrRacket on a Windows computer, how do I get my beloved Emacs keybindings?  I've looked, but can't find the preference.



From mangpo at eecs.berkeley.edu  Tue Sep  9 13:57:57 2014
From: mangpo at eecs.berkeley.edu (Mangpo Phitchaya Phothilimthana)
Date: Tue, 9 Sep 2014 10:57:57 -0700
Subject: [racket] kill and check status of future
Message-ID: <CALEObJOozCYwn+MVwdU24icmSQLch_s2BY7YTgd_PgPnx_V4zA@mail.gmail.com>

Hi,

I try to write a racket parallel program using future and touch. I want the
main thread to be able to check the status of future object if it has
already terminated for not, and able to kill future execution if timeout.
These features exist for subprocess but I can't find these features for
future.

Best,
Mangpo
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140909/ecd6bfb4/attachment.html>

From neil at neilvandyke.org  Tue Sep  9 16:33:24 2014
From: neil at neilvandyke.org (Neil Van Dyke)
Date: Tue, 09 Sep 2014 16:33:24 -0400
Subject: [racket] racket loader/vm detecting out-of-date cached compiled
	files
Message-ID: <540F6414.3040309@neilvandyke.org>

Is there any low-hanging fruit for making the command-line Racket detect 
more cases in which some of the compiled files are out-of-date?

Details...

I have run into a few instances of undetected out-of-date compiled 
Racket code a few times in the last year.

For example, today, after sys-admin did a code patch a production 
server, one part of the system gave this error when it was run:

     link: bad variable linkage; reference to a variable that has the 
wrong procedure or structure-type shape

My guess in this case is that either the code was not recompiled (though 
it should be when they do "make run" of the system), or that the compile 
was wrong because the file mtimes were wrong (such as might happen with 
SVN with clock skew, or with "tar").  This was with Racket 5.3.4.  
(Incidentally, this particular system has a large number of files and a 
somewhat funny dependency graph, possibly bigger than DrRacket's or any 
other core Racket code.  But my guess is that's not the cause of this 
problem.)

We also ran into out-of-date compiled files once before in production, 
in the last year.  Also involving some kind of update to the source files.

And I have run into out-of-date compiles in development of my open 
source code and personal apps, perhaps several times in recent years.  
Usually when using DrRacket (perhaps alternating between DrRacket and 
command-line "racket"), and probably when I was doing heavy syntax 
transformer work (especially when the module-file graph was messier, 
before submodules simplified that greatly).  Sometimes, it was obvious 
from things in error messages that an old compiled version of the code 
was being picked up from somewhere.

This is not a huge problem, but if there's some easy and efficient 
checks that could be added, that would be a win.  (Maybe there's already 
something like a file "stat" of source and corresponding compiled files, 
and some knowledge of the dependency graph that's consulted.  If that's 
bulletproof, then I wonder whether it would make sense to supplement the 
mtime checking with checking embedding timestamps embedded *within* the 
compiled files, consulted separately from the mtimes.  Though embedded 
timestamps might be significant work to do, when it's unclear whether 
bad mtimes are actually a problem for anyone in practice.)

Neil


From mb at mbtype.com  Tue Sep  9 22:56:35 2014
From: mb at mbtype.com (Matthew Butterick)
Date: Tue, 9 Sep 2014 19:56:35 -0700
Subject: [racket] racket loader/vm detecting out-of-date cached compiled
	files
In-Reply-To: <540F6414.3040309@neilvandyke.org>
References: <540F6414.3040309@neilvandyke.org>
Message-ID: <676D1B90-019A-441B-8479-6BF9F1AD6AD6@mbtype.com>

I also get similar "bad variable linkage" errors frequently with Racket. Deleting the "compiled" folder is always the cure. But it's not obvious from the error message why that should be so.


On Sep 9, 2014, at 1:33 PM, Neil Van Dyke <neil at neilvandyke.org> wrote:

> Is there any low-hanging fruit for making the command-line Racket detect more cases in which some of the compiled files are out-of-date?
> 
> Details...
> 
> I have run into a few instances of undetected out-of-date compiled Racket code a few times in the last year.
> 
> For example, today, after sys-admin did a code patch a production server, one part of the system gave this error when it was run:
> 
>    link: bad variable linkage; reference to a variable that has the wrong procedure or structure-type shape
> 
> My guess in this case is that either the code was not recompiled (though it should be when they do "make run" of the system), or that the compile was wrong because the file mtimes were wrong (such as might happen with SVN with clock skew, or with "tar").  This was with Racket 5.3.4.  (Incidentally, this particular system has a large number of files and a somewhat funny dependency graph, possibly bigger than DrRacket's or any other core Racket code.  But my guess is that's not the cause of this problem.)
> 
> We also ran into out-of-date compiled files once before in production, in the last year.  Also involving some kind of update to the source files.
> 
> And I have run into out-of-date compiles in development of my open source code and personal apps, perhaps several times in recent years.  Usually when using DrRacket (perhaps alternating between DrRacket and command-line "racket"), and probably when I was doing heavy syntax transformer work (especially when the module-file graph was messier, before submodules simplified that greatly).  Sometimes, it was obvious from things in error messages that an old compiled version of the code was being picked up from somewhere.
> 
> This is not a huge problem, but if there's some easy and efficient checks that could be added, that would be a win.  (Maybe there's already something like a file "stat" of source and corresponding compiled files, and some knowledge of the dependency graph that's consulted.  If that's bulletproof, then I wonder whether it would make sense to supplement the mtime checking with checking embedding timestamps embedded *within* the compiled files, consulted separately from the mtimes.  Though embedded timestamps might be significant work to do, when it's unclear whether bad mtimes are actually a problem for anyone in practice.)
> 
> Neil
> 
> ____________________
> Racket Users list:
> http://lists.racket-lang.org/users



From dbastos at toledo.com  Wed Sep 10 13:02:14 2014
From: dbastos at toledo.com (Daniel Bastos)
Date: Wed, 10 Sep 2014 14:02:14 -0300
Subject: [racket] htdp: functions of section 19 violate the grammar of
	section 8
Message-ID: <CAB08-qrfnpSGyxSMgYvHp8Hasga-eULLp-7C4bAokNU5Rmivyw@mail.gmail.com>

I have not been able to verify this statement.

  "As a matter of fact, the functions of section 19 violate the Scheme
grammar of section 8." -- First paragraph of section 20.

I think the statement is referring to functions such as filter1.

(define (filter1 rel-op alon t)
  (cond
    [(empty? alon) empty]
    [else (cond
   [(rel-op (first alon) t)
    (cons (first alon)
          (filter1 rel-op (rest alon) t))]
   [else
    (filter1 rel-op (rest alon) t)])]))

It must be because of rel-op, I figure. But rel-op is a <var>.
According to Figure 21 (the grammar of section 8), (<var> <exp>
...<exp>) is an <exp>, so I think filter1 doesn't violate the grammar
of section 8.

Where am I wrong? Thank you!

From stchang at ccs.neu.edu  Wed Sep 10 14:25:42 2014
From: stchang at ccs.neu.edu (Stephen Chang)
Date: Wed, 10 Sep 2014 14:25:42 -0400
Subject: [racket] htdp: functions of section 19 violate the grammar of
 section 8
In-Reply-To: <CAB08-qrfnpSGyxSMgYvHp8Hasga-eULLp-7C4bAokNU5Rmivyw@mail.gmail.com>
References: <CAB08-qrfnpSGyxSMgYvHp8Hasga-eULLp-7C4bAokNU5Rmivyw@mail.gmail.com>
Message-ID: <CAFfiA1Lg5NQ6geWiJufOqnoFRnw=n8cr-SOT1+3xxGbKhYmZ2Q@mail.gmail.com>

I believe it's answered in a subsequent paragraph?

"The abstract functions of section 19 violate Scheme's basic grammar
in two ways. First, the names of functions and primitive operations
are used as arguments in applications. An argument, though, is an
expression, and the class of expressions does not contain primitive
operations and function names. It does contain variables, but we
agreed that they are only those variables mentioned in variable
definitions and as function parameters. Second, parameters are used as
if they were functions, that is, the first position of applications.
But the grammar of section 8 allows only the names of functions and
primitive operations in this place."

On Wed, Sep 10, 2014 at 1:02 PM, Daniel Bastos <dbastos at toledo.com> wrote:
> I have not been able to verify this statement.
>
>   "As a matter of fact, the functions of section 19 violate the Scheme
> grammar of section 8." -- First paragraph of section 20.
>
> I think the statement is referring to functions such as filter1.
>
> (define (filter1 rel-op alon t)
>   (cond
>     [(empty? alon) empty]
>     [else (cond
>    [(rel-op (first alon) t)
>     (cons (first alon)
>           (filter1 rel-op (rest alon) t))]
>    [else
>     (filter1 rel-op (rest alon) t)])]))
>
> It must be because of rel-op, I figure. But rel-op is a <var>.
> According to Figure 21 (the grammar of section 8), (<var> <exp>
> ...<exp>) is an <exp>, so I think filter1 doesn't violate the grammar
> of section 8.
>
> Where am I wrong? Thank you!
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users

From dbastos at toledo.com  Wed Sep 10 14:57:58 2014
From: dbastos at toledo.com (Daniel Bastos)
Date: Wed, 10 Sep 2014 15:57:58 -0300
Subject: [racket] htdp: functions of section 19 violate the grammar of
 section 8
In-Reply-To: <CAFfiA1Lg5NQ6geWiJufOqnoFRnw=n8cr-SOT1+3xxGbKhYmZ2Q@mail.gmail.com>
References: <CAB08-qrfnpSGyxSMgYvHp8Hasga-eULLp-7C4bAokNU5Rmivyw@mail.gmail.com>
 <CAFfiA1Lg5NQ6geWiJufOqnoFRnw=n8cr-SOT1+3xxGbKhYmZ2Q@mail.gmail.com>
Message-ID: <CAB08-qpbv+T1pVUGbUYNk00i30iU63Keke849LKDhuxsDo86hw@mail.gmail.com>

I remember having had that question when I was at that chapter. I must
have thought I answered it, but I'm having it again and unable to
answer now. IOW, I'm confused.

I'm having difficulties with this phrase too. "First, the names of
functions and primitive operations are used as arguments in
applications."

What is an argument in application? I'm going to try to answer it. An
application is, according to the grammar of section 8,

   (<var> <exp> ...<exp>)

or

  (<prm> <exp> ...<exp>)

So its arguments are <var>, <exp>, ... and the final <exp>. Similarly,
the second application has arguments <prm>, <exp>, ..., <exp>. Is that
correct?

Thank you.

On Wed, Sep 10, 2014 at 3:25 PM, Stephen Chang <stchang at ccs.neu.edu> wrote:
> I believe it's answered in a subsequent paragraph?
>
> "The abstract functions of section 19 violate Scheme's basic grammar
> in two ways. First, the names of functions and primitive operations
> are used as arguments in applications. An argument, though, is an
> expression, and the class of expressions does not contain primitive
> operations and function names. It does contain variables, but we
> agreed that they are only those variables mentioned in variable
> definitions and as function parameters. Second, parameters are used as
> if they were functions, that is, the first position of applications.
> But the grammar of section 8 allows only the names of functions and
> primitive operations in this place."

From stchang at ccs.neu.edu  Wed Sep 10 15:10:03 2014
From: stchang at ccs.neu.edu (Stephen Chang)
Date: Wed, 10 Sep 2014 15:10:03 -0400
Subject: [racket] htdp: functions of section 19 violate the grammar of
 section 8
In-Reply-To: <CAB08-qpbv+T1pVUGbUYNk00i30iU63Keke849LKDhuxsDo86hw@mail.gmail.com>
References: <CAB08-qrfnpSGyxSMgYvHp8Hasga-eULLp-7C4bAokNU5Rmivyw@mail.gmail.com>
 <CAFfiA1Lg5NQ6geWiJufOqnoFRnw=n8cr-SOT1+3xxGbKhYmZ2Q@mail.gmail.com>
 <CAB08-qpbv+T1pVUGbUYNk00i30iU63Keke849LKDhuxsDo86hw@mail.gmail.com>
Message-ID: <CAFfiA1Ji+2E4rPURd47rBBiv17GRdBSNbu374U_VKROdGNpOog@mail.gmail.com>

The arguments are the <exp> ...<exp>. The function position is <var> or <prm>

Essentially, the grammar in sec20 extends sec8 with the ability to:
- pass functions and primitives, like +, as arguments, and
- compute the applied function in the function position instead of
always using predefined <var>s or <prm>s

Does that make sense?

On Wed, Sep 10, 2014 at 2:57 PM, Daniel Bastos <dbastos at toledo.com> wrote:
> I remember having had that question when I was at that chapter. I must
> have thought I answered it, but I'm having it again and unable to
> answer now. IOW, I'm confused.
>
> I'm having difficulties with this phrase too. "First, the names of
> functions and primitive operations are used as arguments in
> applications."
>
> What is an argument in application? I'm going to try to answer it. An
> application is, according to the grammar of section 8,
>
>    (<var> <exp> ...<exp>)
>
> or
>
>   (<prm> <exp> ...<exp>)
>
> So its arguments are <var>, <exp>, ... and the final <exp>. Similarly,
> the second application has arguments <prm>, <exp>, ..., <exp>. Is that
> correct?
>
> Thank you.
>
> On Wed, Sep 10, 2014 at 3:25 PM, Stephen Chang <stchang at ccs.neu.edu> wrote:
>> I believe it's answered in a subsequent paragraph?
>>
>> "The abstract functions of section 19 violate Scheme's basic grammar
>> in two ways. First, the names of functions and primitive operations
>> are used as arguments in applications. An argument, though, is an
>> expression, and the class of expressions does not contain primitive
>> operations and function names. It does contain variables, but we
>> agreed that they are only those variables mentioned in variable
>> definitions and as function parameters. Second, parameters are used as
>> if they were functions, that is, the first position of applications.
>> But the grammar of section 8 allows only the names of functions and
>> primitive operations in this place."

From johnbclements at gmail.com  Wed Sep 10 17:03:13 2014
From: johnbclements at gmail.com (John Clements)
Date: Wed, 10 Sep 2014 14:03:13 -0700
Subject: [racket] catalog pkg versioning: am I doing it right?
Message-ID: <CAAwfr=WH467+MHGdjHMnKriQTJsmhFaT7_VK0FuRHePyL524UA@mail.gmail.com>

It's possible to specify "version exceptions" in the new package catalog
system (i.e., on pkg.racket-lang.org). I want to use one version (old
version) for pre-6.0, and another one for >= 6.0. IIUC, the right way to do
this is to specify a mapping for 'default to my pre-6 branch, and then
explicit mappings for 6.0 and 6.1 to the new version.

My questions:

1) how are version numbers matched? I just tried an experiment, and it
looks like my version exceptions are confined to exact matches; e.g., when
I have a version exception for 6.1, and I do a raco pkg catalog-show for
6.1.0.8, I get the default one.
2) This requires me to add an exception for every new version, right?

Finally, I'm still kind of baffled by the pkg.racket-lang.org interface; it
seems that when I make changes to a package and then close and re-open the
floating window, my changes are still there, but that actually hitting
browser reload often causes the changes to be lost. Am I just waiting for a
change to propagate through a back-end database, here?

Many thanks,

John
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140910/6764cd47/attachment.html>

From jay.mccarthy at gmail.com  Wed Sep 10 17:26:54 2014
From: jay.mccarthy at gmail.com (Jay McCarthy)
Date: Wed, 10 Sep 2014 17:26:54 -0400
Subject: [racket] catalog pkg versioning: am I doing it right?
In-Reply-To: <CAAwfr=WH467+MHGdjHMnKriQTJsmhFaT7_VK0FuRHePyL524UA@mail.gmail.com>
References: <CAAwfr=WH467+MHGdjHMnKriQTJsmhFaT7_VK0FuRHePyL524UA@mail.gmail.com>
Message-ID: <CAJYbDam9TGzCxwooAYh0ognX0wY07QFRYA+kN9oChEgdyft2-g@mail.gmail.com>

On Wed, Sep 10, 2014 at 5:03 PM, John Clements <johnbclements at gmail.com> wrote:
> It's possible to specify "version exceptions" in the new package catalog
> system (i.e., on pkg.racket-lang.org). I want to use one version (old
> version) for pre-6.0, and another one for >= 6.0. IIUC, the right way to do
> this is to specify a mapping for 'default to my pre-6 branch, and then
> explicit mappings for 6.0 and 6.1 to the new version.
>
> My questions:
>
> 1) how are version numbers matched? I just tried an experiment, and it looks
> like my version exceptions are confined to exact matches; e.g., when I have
> a version exception for 6.1, and I do a raco pkg catalog-show for 6.1.0.8, I
> get the default one.
> 2) This requires me to add an exception for every new version, right?

This sounds right. The versioning system was principally designed for
dealing with keeping a small number of old Racket versions back on an
old version of the package. Since there are a small number of versions
since the package system existed, this seemed cheap.

> Finally, I'm still kind of baffled by the pkg.racket-lang.org interface; it
> seems that when I make changes to a package and then close and re-open the
> floating window, my changes are still there, but that actually hitting
> browser reload often causes the changes to be lost. Am I just waiting for a
> change to propagate through a back-end database, here?

When you make changes to a package's metadata, those changes are saved
instantly (as soon as you press enter or de-focus the text input) and
both your browser and the back-end server (modulo network errors.)
Every time the back-end server receives a change, it updates the S3
cache. This updating process takes about X seconds and then S3 takes
about Y seconds. If you reload your browser, you are throwing away
your local copy with updated changes and getting the S3 cache. If you
reload in under X+Y seconds (or your browser doesn't make the request
at all), then you won't see any changes.

Jay

> Many thanks,
>
> John
>
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>



-- 
Jay McCarthy
http://jeapostrophe.github.io

           "Wherefore, be not weary in well-doing,
      for ye are laying the foundation of a great work.
And out of small things proceedeth that which is great."
                          - D&C 64:33

From ameliastahlman at gmail.com  Wed Sep 10 20:29:53 2014
From: ameliastahlman at gmail.com (Amy Stahlman)
Date: Wed, 10 Sep 2014 19:29:53 -0500
Subject: [racket] Possible bug with init-auto-scrollbars
Message-ID: <CACfnStUtBzmpvC+1jnonbs4VA_CoMBgKi7zsKOrySe3=xosSwg@mail.gmail.com>

I've been working on a program that uses "init-auto-scrollbars" and I
noticed that when the window is resized it actually changes the size of the
virtual canvas along with the literal canvas instead of just resizing the
scrollbar and changing the size of the literal canvas.

Here's an example that shows this:














*(require racket/gui)(define frame (new frame%                  [label
"Example"]                  [width 300]                  [height
300]))(define canv (new canvas% [parent frame]    [style '(vscroll
hscroll)]            [paint-callback             (lambda (canvas dc)(send
dc draw-ellipse 0 0 500 500))]))(send frame show #t)(send canv
init-auto-scrollbars 500 500 0.0 0.0)*

Resize the window, making it large enough to show the entire ellipse;
notice how the scrollbar's relative size doesn't change, and there is blank
space at the bottom.
Is this by design or some kind of bug?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140910/9bdad02c/attachment.html>

From mflatt at cs.utah.edu  Wed Sep 10 20:46:54 2014
From: mflatt at cs.utah.edu (Matthew Flatt)
Date: Wed, 10 Sep 2014 18:46:54 -0600
Subject: [racket] Possible bug with init-auto-scrollbars
In-Reply-To: <CACfnStUtBzmpvC+1jnonbs4VA_CoMBgKi7zsKOrySe3=xosSwg@mail.gmail.com>
References: <CACfnStUtBzmpvC+1jnonbs4VA_CoMBgKi7zsKOrySe3=xosSwg@mail.gmail.com>
Message-ID: <20140911004655.A7029650194@mail-svr1.cs.utah.edu>

That's a bug in v6.1 and earlier (on platforms other than Mac OS X).

The bug has been fixed for the next version, so you could try a
snapshot from
 http://pre.racket-lang.org/

If you need to stick with a different version, I think you can work
around the bug by creating a `canvas%` subclass that calls
`init-auto-scrollbars` on each `on-resize`:

 #lang racket
 (require racket/gui)

 (define frame (new frame%
                    [label "Example"]
                    [width 300]
                    [height 300]))

 (define workaround-canvas%
   (class canvas%
     (super-new)
     (inherit init-auto-scrollbars)
     (define/override (on-size w h)
       (init-auto-scrollbars 500 500 0.0 0.0))))

 (define canv (new workaround-canvas% 
                   [parent frame]  
                   [style '(vscroll hscroll)]
                   [paint-callback 
                    (lambda (canvas dc)
                      (send dc draw-ellipse 0 0 500 500))]))

 (send frame show #t)

At Wed, 10 Sep 2014 19:29:53 -0500, Amy Stahlman wrote:
> I've been working on a program that uses "init-auto-scrollbars" and I
> noticed that when the window is resized it actually changes the size of the
> virtual canvas along with the literal canvas instead of just resizing the
> scrollbar and changing the size of the literal canvas.
> 
> Here's an example that shows this:
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> *(require racket/gui)(define frame (new frame%                  [label
> "Example"]                  [width 300]                  [height
> 300]))(define canv (new canvas% [parent frame]    [style '(vscroll
> hscroll)]            [paint-callback             (lambda (canvas dc)(send
> dc draw-ellipse 0 0 500 500))]))(send frame show #t)(send canv
> init-auto-scrollbars 500 500 0.0 0.0)*
> 
> Resize the window, making it large enough to show the entire ellipse;
> notice how the scrollbar's relative size doesn't change, and there is blank
> space at the bottom.
> Is this by design or some kind of bug?
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users

From ameliastahlman at gmail.com  Wed Sep 10 21:55:13 2014
From: ameliastahlman at gmail.com (Amy Stahlman)
Date: Wed, 10 Sep 2014 20:55:13 -0500
Subject: [racket] Possible bug with init-auto-scrollbars
In-Reply-To: <20140911004655.A7029650194@mail-svr1.cs.utah.edu>
References: <CACfnStUtBzmpvC+1jnonbs4VA_CoMBgKi7zsKOrySe3=xosSwg@mail.gmail.com>
 <20140911004655.A7029650194@mail-svr1.cs.utah.edu>
Message-ID: <CACfnStX+QAA1kxNwqLGx6_+MUqV_HFUDNx5jjSiuUbufd46Neg@mail.gmail.com>

On Wed, Sep 10, 2014 at 7:46 PM, Matthew Flatt <mflatt at cs.utah.edu> wrote:

> That's a bug in v6.1 and earlier (on platforms other than Mac OS X).
>
> The bug has been fixed for the next version, so you could try a
> snapshot from
>  http://pre.racket-lang.org/
>
> If you need to stick with a different version, I think you can work
> around the bug by creating a `canvas%` subclass that calls
> `init-auto-scrollbars` on each `on-resize`:
>

Thanks, the workaround works great for my program.
Sincerely, Campbell Stahlman

>
>  #lang racket
>  (require racket/gui)
>
>  (define frame (new frame%
>                     [label "Example"]
>                     [width 300]
>                     [height 300]))
>
>  (define workaround-canvas%
>    (class canvas%
>      (super-new)
>      (inherit init-auto-scrollbars)
>      (define/override (on-size w h)
>        (init-auto-scrollbars 500 500 0.0 0.0))))
>
>  (define canv (new workaround-canvas%
>                    [parent frame]
>                    [style '(vscroll hscroll)]
>                    [paint-callback
>                     (lambda (canvas dc)
>                       (send dc draw-ellipse 0 0 500 500))]))
>
>  (send frame show #t)
>
> At Wed, 10 Sep 2014 19:29:53 -0500, Amy Stahlman wrote:
> > I've been working on a program that uses "init-auto-scrollbars" and I
> > noticed that when the window is resized it actually changes the size of
> the
> > virtual canvas along with the literal canvas instead of just resizing the
> > scrollbar and changing the size of the literal canvas.
> >
> > Here's an example that shows this:
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > *(require racket/gui)(define frame (new frame%                  [label
> > "Example"]                  [width 300]                  [height
> > 300]))(define canv (new canvas% [parent frame]    [style '(vscroll
> > hscroll)]            [paint-callback             (lambda (canvas dc)(send
> > dc draw-ellipse 0 0 500 500))]))(send frame show #t)(send canv
> > init-auto-scrollbars 500 500 0.0 0.0)*
> >
> > Resize the window, making it large enough to show the entire ellipse;
> > notice how the scrollbar's relative size doesn't change, and there is
> blank
> > space at the bottom.
> > Is this by design or some kind of bug?
> > ____________________
> >   Racket Users list:
> >   http://lists.racket-lang.org/users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140910/673a6017/attachment.html>

From acarrico at memebeam.org  Wed Sep 10 22:50:35 2014
From: acarrico at memebeam.org (Anthony Carrico)
Date: Wed, 10 Sep 2014 22:50:35 -0400
Subject: [racket] TR predicate documentation, typo?
Message-ID: <54110DFB.3000803@memebeam.org>

In http://docs.racket-lang.org/ts-reference/type-ref.html

1.6 Other Type Constructors

"The filter specifies that when (string? x) evaluates to a true value,
the variable x can be assumed to have type String. Likewise, if the
expression evaluates to #f, the variable has type String."

I think there must be a typo here, what exactly happens if the
expression evaluates to #f?

-- 
Anthony Carrico

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 181 bytes
Desc: OpenPGP digital signature
URL: <http://lists.racket-lang.org/users/archive/attachments/20140910/9520039b/attachment.sig>

From asumu at ccs.neu.edu  Wed Sep 10 23:43:26 2014
From: asumu at ccs.neu.edu (Asumu Takikawa)
Date: Wed, 10 Sep 2014 23:43:26 -0400
Subject: [racket] TR predicate documentation, typo?
In-Reply-To: <54110DFB.3000803@memebeam.org>
References: <54110DFB.3000803@memebeam.org>
Message-ID: <20140911034326.GC21882@localhost>

On 2014-09-10 22:50:35 -0400, Anthony Carrico wrote:
> "The filter specifies that when (string? x) evaluates to a true value,
> the variable x can be assumed to have type String. Likewise, if the
> expression evaluates to #f, the variable has type String."
>
> I think there must be a typo here, what exactly happens if the
> expression evaluates to #f?

You're right, that sentence doesn't make sense (my fault, I think). It
should say that the variable is known to *not* have type String in
branches where `(string? x)` evaluated to #f.

Thanks, will fix it.

Cheers,
Asumu

From zeppieri at gmail.com  Thu Sep 11 00:56:12 2014
From: zeppieri at gmail.com (Jon Zeppieri)
Date: Thu, 11 Sep 2014 00:56:12 -0400
Subject: [racket] possibility of defining recontract-all-from-out
Message-ID: <CAKfDxxwVnF_bMUgsMoQtE3x6VVp_maKDCZmwYy8xJ2th50c_eQ@mail.gmail.com>

Rather than listing a large number of identifiers in a (recontract-out
...) spec, I'd like to re-contract everything from a certain required
private module. I don't know how to do this. The crux of the problem
is that syntax-local-module-required-identifiers can only be called
while a provide transformer is running, but recontract-out is a
provide pre-transformer. So, by the time I'm able to grab all of the
identifiers from the required module, it's too late to make use of
recontract-out (or contract-out).

Is it even possible to do this? And, if so, is there a way to do it
without duplicating much of racket/contract/private/out.rkt?

-Jon

From whukriede at gmail.com  Thu Sep 11 06:00:01 2014
From: whukriede at gmail.com (Wolfgang Hukriede)
Date: Thu, 11 Sep 2014 12:00:01 +0200
Subject: [racket] error writing to stream port (Broken pipe; errno=32)
Message-ID: <n88egviqxy6.fsf@almelo.raji.xor>

How am I supposed to catch:

error writing to stream port (Broken pipe; errno=32) ?

Thanks, Wolfgang

From whukriede at gmail.com  Thu Sep 11 06:26:00 2014
From: whukriede at gmail.com (Wolfgang Hukriede)
Date: Thu, 11 Sep 2014 12:26:00 +0200
Subject: [racket] error writing to stream port (Broken pipe; errno=32)
Message-ID: <n88ha0esbbb.fsf@almelo.raji.xor>

To expand a bit: 
the handling of the broken pipe seems okay to me, I just
don't want to see the error message.
Thanks!

From whukriede at gmail.com  Thu Sep 11 07:18:18 2014
From: whukriede at gmail.com (Wolfgang Hukriede)
Date: Thu, 11 Sep 2014 13:18:18 +0200
Subject: [racket] error writing to stream port (Broken pipe; errno=32)
Message-ID: <n8861guwglh.fsf@almelo.raji.xor>

Do I need to install a signal handler catching SIGPIPE? (Or replace
the existing signal-handler?)
Thanks!

From matthias at ccs.neu.edu  Thu Sep 11 08:01:38 2014
From: matthias at ccs.neu.edu (Matthias Felleisen)
Date: Thu, 11 Sep 2014 08:01:38 -0400
Subject: [racket] error writing to stream port (Broken pipe; errno=32)
In-Reply-To: <n8861guwglh.fsf@almelo.raji.xor>
References: <n8861guwglh.fsf@almelo.raji.xor>
Message-ID: <0EFC27ED-01FE-4B47-A588-3F159121BD0C@ccs.neu.edu>


(with-handlers ((exn:fail? (lambda (x) "let's throw away the exception from a closed port"))
  ... do stuff with ports and pipes ...)

[[ You probably need to provide more context to get truly useful answers. ]]


On Sep 11, 2014, at 7:18 AM, Wolfgang Hukriede wrote:

> Do I need to install a signal handler catching SIGPIPE? (Or replace
> the existing signal-handler?)
> Thanks!
> ____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users



From matthias at ccs.neu.edu  Thu Sep 11 10:35:05 2014
From: matthias at ccs.neu.edu (Matthias Felleisen)
Date: Thu, 11 Sep 2014 10:35:05 -0400
Subject: [racket] possibility of defining recontract-all-from-out
In-Reply-To: <CAKfDxxwVnF_bMUgsMoQtE3x6VVp_maKDCZmwYy8xJ2th50c_eQ@mail.gmail.com>
References: <CAKfDxxwVnF_bMUgsMoQtE3x6VVp_maKDCZmwYy8xJ2th50c_eQ@mail.gmail.com>
Message-ID: <BE3F6816-1307-46BF-B45C-C74BA917FC83@ccs.neu.edu>


For a project last year, I developed a suite of syntactic abstractions
that expand into contracted provides. With the core abstract you specify
a header-file-like 'thing' and the exporting module(s) can then require
this 'thing' to export identifiers. A re-export would work the same way. 

If my abstractions were robust enough, I'd put them on github but they
aren't and I don't have the time to fortify them sufficiently now. But 
the idea is slightly different from yours and may help you think along
different lines. 

-- Matthias





On Sep 11, 2014, at 12:56 AM, Jon Zeppieri <zeppieri at gmail.com> wrote:

> Rather than listing a large number of identifiers in a (recontract-out
> ...) spec, I'd like to re-contract everything from a certain required
> private module. I don't know how to do this. The crux of the problem
> is that syntax-local-module-required-identifiers can only be called
> while a provide transformer is running, but recontract-out is a
> provide pre-transformer. So, by the time I'm able to grab all of the
> identifiers from the required module, it's too late to make use of
> recontract-out (or contract-out).
> 
> Is it even possible to do this? And, if so, is there a way to do it
> without duplicating much of racket/contract/private/out.rkt?
> 
> -Jon
> ____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users


From dbastos at toledo.com  Thu Sep 11 17:34:44 2014
From: dbastos at toledo.com (Daniel Bastos)
Date: Thu, 11 Sep 2014 18:34:44 -0300
Subject: [racket] htdp: functions of section 19 violate the grammar of
 section 8
In-Reply-To: <CAFfiA1Ji+2E4rPURd47rBBiv17GRdBSNbu374U_VKROdGNpOog@mail.gmail.com>
References: <CAB08-qrfnpSGyxSMgYvHp8Hasga-eULLp-7C4bAokNU5Rmivyw@mail.gmail.com>
 <CAFfiA1Lg5NQ6geWiJufOqnoFRnw=n8cr-SOT1+3xxGbKhYmZ2Q@mail.gmail.com>
 <CAB08-qpbv+T1pVUGbUYNk00i30iU63Keke849LKDhuxsDo86hw@mail.gmail.com>
 <CAFfiA1Ji+2E4rPURd47rBBiv17GRdBSNbu374U_VKROdGNpOog@mail.gmail.com>
Message-ID: <CAB08-qo8soFpyarXTAkFnGfD4Ox9pQY2SrHi-zxdMVJeJC2fNA@mail.gmail.com>

It makes sense, yes. I mean, the overall message. But I'm trying to
pinpoint where the grammar of section 8 tells me "no, you cannot put a
function name in an argument of an application". I read again the
entire section 8, am reading section 19 again and I still don't see
it. I'll write my argument and you'll spot a false statement
somewhere.

(By the way, I know the language Beginning Student doesn't allow a
function name to be passed as an application argument. I've checked
it.)

It's clear that a primitive cannot be passed as an argument in the
grammar of section 8. Here's my proof. An application in that grammar
is of the form (<var> <exp> ...<exp>). Since primitives do not appear
in this form, it cannot be present in an application. (We're done.)

However, any expression /can/ be used as an application argument
because an application is of the form (<var> <exp> ...<exp>). What are
the possible expressions in the grammar of section 8? It begins with
<var>. When I define a user-defined function, its name is a <var>.
Therefore, any user-defined function's name is an expression and hence
can appear in an application.

But section 20 says that "[an argument [...] is an expression, and the
class of expressions does not contain primitive operations and
function names. It does contain variables, but we agreed that they are
only those variables mentioned in variable definitions and as function
parameters."

This must be where the problem is. Where did we agree to that? I went
back to section 8 to see what a <var> is. "The first category is that
of variables, which are the names of functions and values." So a <var>
names a function such as

  (define (f x) (* 2 x))

and so f should be allowed to appear as an application argument.
Doesn't that follow from the text?

Thank you!

On Wed, Sep 10, 2014 at 4:10 PM, Stephen Chang <stchang at ccs.neu.edu> wrote:
> The arguments are the <exp> ...<exp>. The function position is <var> or <prm>
>
> Essentially, the grammar in sec20 extends sec8 with the ability to:
> - pass functions and primitives, like +, as arguments, and
> - compute the applied function in the function position instead of
> always using predefined <var>s or <prm>s
>
> Does that make sense?

From zeppieri at gmail.com  Fri Sep 12 10:20:17 2014
From: zeppieri at gmail.com (Jon Zeppieri)
Date: Fri, 12 Sep 2014 10:20:17 -0400
Subject: [racket] possibility of defining recontract-all-from-out
In-Reply-To: <BE3F6816-1307-46BF-B45C-C74BA917FC83@ccs.neu.edu>
References: <CAKfDxxwVnF_bMUgsMoQtE3x6VVp_maKDCZmwYy8xJ2th50c_eQ@mail.gmail.com>
 <BE3F6816-1307-46BF-B45C-C74BA917FC83@ccs.neu.edu>
Message-ID: <CAKfDxxyivPwq2vmxGPDbx0w+3yVgefs+-09SOhWcY0eVJc_nHQ@mail.gmail.com>

Matthias,

Would you be willing to elaborate on this a bit? Specifically, how is
the header specified?

Thanks,
Jon


On Thu, Sep 11, 2014 at 10:35 AM, Matthias Felleisen
<matthias at ccs.neu.edu> wrote:
>
> For a project last year, I developed a suite of syntactic abstractions
> that expand into contracted provides. With the core abstract you specify
> a header-file-like 'thing' and the exporting module(s) can then require
> this 'thing' to export identifiers. A re-export would work the same way.
>
> If my abstractions were robust enough, I'd put them on github but they
> aren't and I don't have the time to fortify them sufficiently now. But
> the idea is slightly different from yours and may help you think along
> different lines.
>
> -- Matthias
>
>
>
>
>
> On Sep 11, 2014, at 12:56 AM, Jon Zeppieri <zeppieri at gmail.com> wrote:
>
>> Rather than listing a large number of identifiers in a (recontract-out
>> ...) spec, I'd like to re-contract everything from a certain required
>> private module. I don't know how to do this. The crux of the problem
>> is that syntax-local-module-required-identifiers can only be called
>> while a provide transformer is running, but recontract-out is a
>> provide pre-transformer. So, by the time I'm able to grab all of the
>> identifiers from the required module, it's too late to make use of
>> recontract-out (or contract-out).
>>
>> Is it even possible to do this? And, if so, is there a way to do it
>> without duplicating much of racket/contract/private/out.rkt?
>>
>> -Jon
>> ____________________
>>  Racket Users list:
>>  http://lists.racket-lang.org/users
>

From matthias at ccs.neu.edu  Fri Sep 12 10:47:20 2014
From: matthias at ccs.neu.edu (Matthias Felleisen)
Date: Fri, 12 Sep 2014 10:47:20 -0400
Subject: [racket] possibility of defining recontract-all-from-out
In-Reply-To: <CAKfDxxyivPwq2vmxGPDbx0w+3yVgefs+-09SOhWcY0eVJc_nHQ@mail.gmail.com>
References: <CAKfDxxwVnF_bMUgsMoQtE3x6VVp_maKDCZmwYy8xJ2th50c_eQ@mail.gmail.com>
 <BE3F6816-1307-46BF-B45C-C74BA917FC83@ccs.neu.edu>
 <CAKfDxxyivPwq2vmxGPDbx0w+3yVgefs+-09SOhWcY0eVJc_nHQ@mail.gmail.com>
Message-ID: <578676E7-627B-4C4B-AF19-50F57362648D@ccs.neu.edu>


Does this sketch help: 

#lang racket

;; define a header file language 
(module L racket
  ;; specification
  (provide export)
  ;; ----------------------------------------------
  ;; implementation 
  (define-syntax (export stx)
    (syntax-case stx ()
      [(_ file name ctc)
       #'(define-syntax (file stx)
           (syntax-case stx ()
             [(_) 
              ;; let's break hygiene here: 
              (let* ([x '(provide (contract-out (name ctc)))]
                     [y (datum->syntax stx x)])
                y)]))])))

;; -------------------------------------------------------
(module header-file racket
  (require (submod ".." L))
  (provide provide-stack)
  ;; -----------------------------------------------------
  (export provide-stack make (-> empty?)))

(module good-server racket 
  (require (submod ".." header-file))
  (provide-stack)
  ;; -----------------------------------------------------
  (define (make)
    '()))

(module bad-server racket 
  (require (submod ".." header-file))
  (provide-stack)
  ;; -----------------------------------------------------
  (define (make)
    "the empty stack"))

(module client racket 
  (provide
   (contract-out
    (test (-> (values string? string?)))))
  ;; -----------------------------------------------------
  (require (prefix-in good: (submod ".." good-server))
           (prefix-in bad: (submod ".." bad-server)))
  
  (define (test)
    (values 
     (with-handlers ((exn:fail:contract? (lambda (_) "passed")))
       (bad:make)
       "failed")
     (begin (good:make) "passed"))))

;; RUN BABY RUN 
(require 'client)
(test)
  
    
                
 

From zeppieri at gmail.com  Fri Sep 12 11:39:36 2014
From: zeppieri at gmail.com (Jon Zeppieri)
Date: Fri, 12 Sep 2014 11:39:36 -0400
Subject: [racket] possibility of defining recontract-all-from-out
In-Reply-To: <578676E7-627B-4C4B-AF19-50F57362648D@ccs.neu.edu>
References: <CAKfDxxwVnF_bMUgsMoQtE3x6VVp_maKDCZmwYy8xJ2th50c_eQ@mail.gmail.com>
 <BE3F6816-1307-46BF-B45C-C74BA917FC83@ccs.neu.edu>
 <CAKfDxxyivPwq2vmxGPDbx0w+3yVgefs+-09SOhWcY0eVJc_nHQ@mail.gmail.com>
 <578676E7-627B-4C4B-AF19-50F57362648D@ccs.neu.edu>
Message-ID: <CAKfDxxx5RKQbSOJ_Fd+WF9dV8uduD7fXZOtE6Z+mEktpoMaiJg@mail.gmail.com>

Yes, this is very helpful. Thank you.

-Jon


On Fri, Sep 12, 2014 at 10:47 AM, Matthias Felleisen
<matthias at ccs.neu.edu> wrote:
>
> Does this sketch help:
>
> #lang racket
>
> ;; define a header file language
> (module L racket
>   ;; specification
>   (provide export)
>   ;; ----------------------------------------------
>   ;; implementation
>   (define-syntax (export stx)
>     (syntax-case stx ()
>       [(_ file name ctc)
>        #'(define-syntax (file stx)
>            (syntax-case stx ()
>              [(_)
>               ;; let's break hygiene here:
>               (let* ([x '(provide (contract-out (name ctc)))]
>                      [y (datum->syntax stx x)])
>                 y)]))])))
>
> ;; -------------------------------------------------------
> (module header-file racket
>   (require (submod ".." L))
>   (provide provide-stack)
>   ;; -----------------------------------------------------
>   (export provide-stack make (-> empty?)))
>
> (module good-server racket
>   (require (submod ".." header-file))
>   (provide-stack)
>   ;; -----------------------------------------------------
>   (define (make)
>     '()))
>
> (module bad-server racket
>   (require (submod ".." header-file))
>   (provide-stack)
>   ;; -----------------------------------------------------
>   (define (make)
>     "the empty stack"))
>
> (module client racket
>   (provide
>    (contract-out
>     (test (-> (values string? string?)))))
>   ;; -----------------------------------------------------
>   (require (prefix-in good: (submod ".." good-server))
>            (prefix-in bad: (submod ".." bad-server)))
>
>   (define (test)
>     (values
>      (with-handlers ((exn:fail:contract? (lambda (_) "passed")))
>        (bad:make)
>        "failed")
>      (begin (good:make) "passed"))))
>
> ;; RUN BABY RUN
> (require 'client)
> (test)
>
>
>
>

From dbastos at toledo.com  Fri Sep 12 13:05:04 2014
From: dbastos at toledo.com (Daniel Bastos)
Date: Fri, 12 Sep 2014 14:05:04 -0300
Subject: [racket] missing solution 20.2.4 ex:fancy-contracts-poly
In-Reply-To: <D6253DDD-91FF-48E2-A330-4024EDBF1C5E@ccs.neu.edu>
References: <CAB08-qrbvXik1gT7R_gcH+yUjCaY-bAZQF+T=LWFj7EB2DuMug@mail.gmail.com>
 <D6253DDD-91FF-48E2-A330-4024EDBF1C5E@ccs.neu.edu>
Message-ID: <CAB08-qq1Qj7Upvf4LEHLPvR35151ro8NGx-3DF=xd-s=FoKrGw@mail.gmail.com>

On Mon, Sep 8, 2014 at 9:51 PM,
Matthias Felleisen <matthias at ccs.neu.edu> wrote:

On Sep 4, 2014, at 12:54 PM, Daniel Bastos wrote:
>
> > A candidate for a solution. (I'm not sure I'm correct regarding the
> > function project. The description says "a lists of lists" (which I
> > translate to (listof ITEM)) and "a function from lists to Xs", so I
> > wonder if the domain of this function-argument must be (listof ITEM)
> > or it could be more general such as (listof Y). I decided to imagine
> > that the domain of the function-argument would be items of the "list
> > of lists" and so I made it into (listof ITEM).
>
>
> In this context ITEM and X and Y are variables (placeholders) that stand
> for sets (classes) of data.
>
> In the following I have translated the purpose statements for these
> functions into prototype calls:
>
> > sort: (listof ITEM) (ITEM ITEM -> boolean) -> (listof ITEM)
>
>
> (sort a-list-of-Xs compare-Xs) : a-list-of-Xs
>   (same Xs indeed, but rearranged)
>
> > map: ((listof ITEM) -> X) (listof ITEM) -> (listof X)
>
> (map a-function-from-item-on-list-to-X a-list-of-ITEMs) : a-list-of-Xs
>
>
>
> > project: (listof (listof ITEM)) ((listof ITEM) -> (listof X)) -> (listof
> X)
>
> (project a-list-of-list-of-ITEMs a-function-from-list-of-ITEMs-to-X) :
> a-list-of-Xs
>
> > (*) Comparison with exercise 20.2.2
>
> Abstraction works by taking two concrete examples and replacing matching
> concrete things with variables.
>
> 20.2.2 asks you to come up with contracts for functions that work on one
> basic kind of data. Use a distinct kind and then apply the abstraction
> design recipe. You will find that at least one of the above is wrong.


Indeed. (That was map.) Thank you.

(*) Candidate for solution for exercise 20.2.4

Our approach will be to take two concrete examples and replace matching
concrete things with variables.

(-) Sort

sort1: (listof number) (number number -> boolean) -> (listof number)
sort2: (listof name) (name name -> boolean) -> (listof name)
sort: (listof ITEM) (ITEM ITEM -> boolean) -> (listof ITEM)

Type variable: ITEM.

(-) Map

map1: (number -> number) (listof number) -> (listof number)
map2: (name -> name) (listof name) -> (listof name)
map: (x -> x) (listof x) -> (listof x)

Type variable: x.

(-) Project

project1: (listof (listof symbol)) ((listof symbol) -> symbol) -> (listof
symbol)
project2: (listof (listof name)) ((listof name) -> name) -> (listof name)
project: (listof (listof x)) ((listof x) -> x) -> (listof x)

Type variable: x.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140912/66e720ee/attachment.html>

From whukriede at gmail.com  Fri Sep 12 15:57:30 2014
From: whukriede at gmail.com (Wolfgang Hukriede)
Date: Fri, 12 Sep 2014 21:57:30 +0200
Subject: [racket] error writing to stream port (Broken pipe; errno=32)
Message-ID: <n88fvfwy5lh.fsf@almelo.raji.xor>

Hi Matthias, nice to meet you! Yes, this was what I was looking for.

For the records:

(define (epipe? x)
  (and
   (exn:fail:filesystem:errno? x) 
   (equal? '(32 . posix) (exn:fail:filesystem:errno-errno x))))

(define (epipe-handler x)
  "ignore")

(with-handlers
 ((epipe? epipe-handler))
 transput...)

Many thanks!

Results are mixed though. Usage like 

$ mzscheme --no-init-file --load pipetest.scm | head

mostly works okay, but depending on how and how much is written to
standard out and on the parameters to "head" failures reproducibly
occur: in some situations the signal is not caught, yielding a "system
error: Broken pipe; errno=32" message (versions 6.1 and 5.93 on
freebsd, and 4.1 on generic linux, 6.1 precompiled racket on
archlinux), or sometimes the pipe even just hangs (6.1 on freebsd).
These were the combinations I tried.
(On 4.1 exn:fail:filesystem was used instead of non-existing
exn:fail:filesystem:errno.)

From dbastos at toledo.com  Fri Sep 12 16:40:14 2014
From: dbastos at toledo.com (Daniel Bastos)
Date: Fri, 12 Sep 2014 17:40:14 -0300
Subject: [racket] missing solution 20.1.1 ex:sem-funcs
In-Reply-To: <DB47056E-CF26-45EA-97B5-7935F08A7AB2@ccs.neu.edu>
References: <CAB08-qr3dsphx=M875-g8EBNkrdFN3-BLkvTQnytrZ4i-5uUvg@mail.gmail.com>
 <DB47056E-CF26-45EA-97B5-7935F08A7AB2@ccs.neu.edu>
Message-ID: <CAB08-qomWeikoWpv_=Y4S7i75=hMP2RybqX0F914F5b3OqzEcg@mail.gmail.com>

On Mon, Sep 8, 2014 at 9:51 PM,
Matthias Felleisen <matthias at ccs.neu.edu> wrote:

On Sep 2, 2014, at 11:45 AM, Daniel Bastos wrote:
>
> > A candidate for a solution.
> >
> > Exercise 20.1.1. Assume the Definitions window in DrScheme contains
> > (define (f x) x). Identify the values among the following expressions:
> >
> > (1) (cons f empty)
> > (2) (f f)
> > (3) (cons f (cons 10 (cons (f 10) empty)))
> >
> > Explain why they are values and why the remaining expressions are not
> > values.
> >
> > Solution. First we consider (1). Here's the relevant part of the
> > grammar.
> >
> >  <lst> = empty | (cons <val> <lst>)
> >
> >  <val> = <boo> | <sym> | <num> | empty | <lst>
> >  | <var> (names of defined functions)
> >  | <prm>
> >
> > So (1) is a value because it is <lst> of the form (cons <var> empty),
> > where f is <var> because it is a defined function.
> >
> > Let's consider (3) before (2). (3) is a list of either <var> or <num>,
> > so (3) is <val> as well.
>
> You checked only one half of the value-grammar for lst. Check the other
> half, too. Report back whether you want to change the answer or not. --
> Matthias
>

I want to change it completely. I'm not very sure about what's going on.
Perhaps the other half would be this: since <var> is a value in this
context because f is in fact a defined function, then we get (cons <val>
<lst>) because empty is a <lst> too and then (cons <val> <lst>) = <lst> due
to the definition of <lst>. Since any <lst> is a <val>, we get a value. (Is
that what you meant with 'check the other half'?)

Here's my new complete attempt at the problem. (I ended up saying they're
all values, due to the fact that we know f and f always yields values.)

Exercise 20.1.1. Assume the Definitions window in DrScheme contains
(define (f x) x). Identify the values among the following expressions:

 (1) (cons f empty)
 (2) (f f)
 (3) (cons f (cons 10 (cons (f 10) empty)))

Explain why they are values and why the remaining expressions are not
values.

Solution. A value is anything of the form

  <val> = <boo> | <sym> | <num> | empty | <lst>
  | <var> (names of defined functions)
  | <prm>

where <lst> is of the form

  <lst> = empty | (cons <val> <lst>).

Now we start with (1) and apply a series of substitutions based on the
definitions we have.

    (cons f empty)
  = (cons <var> <lst>)  ;; since f is a <var>
  = (cons <var> <lst>)  ;; since empty is a <lst>
  = (cons <val> <lst>)  ;; since <var> is a subset of <val>
  = <lst>               ;; definition of <lst>
  = <val>               ;; definition of <val>

Therefore (1) is a value.

Next we consider (2). Here's the relevant part of the grammar.

  <exp> = <var>
  | <prm>
  | (<exp> <exp> ...<exp>)

Again, we start with (2) and apply a series of substitutions.

    (f f)
  = (<var> <var>) ;; since f is a <var>
  = (<exp> <exp>) ;; since <var> is a subset of <exp>

However, (<exp> <exp>) is not by definition a value. It is an
expression which may or may not /have/ a value. We know the
definition of f, however, which tells us that it behaves like
the identity function. So the evaluation of that (<exp> <exp>)
yields a function, which is a value. We end with

  = <val>

Therefore (2) is a value.

Now (3).

    (cons f (cons 10 (cons (f 10) empty)))
  = (cons <var> (cons <num> (cons (<var> <num>) <lst>)))

Now again, the whole expressions depends on the application (<exp> <exp>),
which may or may not have a value. Since that is a call to f, the
identity function, we know it returns its argument, so

  = (cons <var> (cons <num> (cons <num> <lst>)))
  = (cons <var> (cons <num> (cons <val> <lst>))) ;; <num> is a <val>
  = (cons <var> (cons <num> <lst>))              ;; definition of <lst>
  = (cons <var> (cons <val> <lst>))              ;; <num> is a <val>
  = (cons <var> <lst>)                           ;; definition of <lst>
  = (cons <val> <lst>)                           ;; <var> is a <val>
  = <lst>                                        ;; definition of <lst>
  = <val>                                        ;; <lst> is a <val>

Therefore (3) is a value.

Thanks for your attention.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140912/22846cbd/attachment.html>

From matthias at ccs.neu.edu  Fri Sep 12 17:43:46 2014
From: matthias at ccs.neu.edu (Matthias Felleisen)
Date: Fri, 12 Sep 2014 17:43:46 -0400
Subject: [racket] missing solution 20.1.1 ex:sem-funcs
In-Reply-To: <CAB08-qomWeikoWpv_=Y4S7i75=hMP2RybqX0F914F5b3OqzEcg@mail.gmail.com>
References: <CAB08-qr3dsphx=M875-g8EBNkrdFN3-BLkvTQnytrZ4i-5uUvg@mail.gmail.com>
 <DB47056E-CF26-45EA-97B5-7935F08A7AB2@ccs.neu.edu>
 <CAB08-qomWeikoWpv_=Y4S7i75=hMP2RybqX0F914F5b3OqzEcg@mail.gmail.com>
Message-ID: <C238EB4D-A9B9-466D-AE4A-13F3EBF597EA@ccs.neu.edu>


On Sep 12, 2014, at 4:40 PM, Daniel Bastos <dbastos at toledo.com> wrote:

> Again, we start with (2) and apply a series of substitutions.
> 
>     (f f)
>   = (<var> <var>) ;; since f is a <var>
>   = (<exp> <exp>) ;; since <var> is a subset of <exp>
> 
> However, (<exp> <exp>) is not by definition a value. It is an 
> expression which may or may not /have/ a value. We know the 
> definition of f, however, which tells us that it behaves like
> the identity function. So the evaluation of that (<exp> <exp>)
> yields a function, which is a value. We end with 
> 
>   = <val>
> 
> Therefore (2) is a value

Is there a difference between the statements 

 "Daniel is a hat."

and 

 "Daniel has a hat."

or do they really mean the same thing? 

In this spirit, is there a difference between the statement 

 "(f f) is a value." 

and 

 "(f f) has a value." 

or do they really mean the same thing? 

-- Matthias




From robby at eecs.northwestern.edu  Fri Sep 12 18:16:34 2014
From: robby at eecs.northwestern.edu (Robby Findler)
Date: Fri, 12 Sep 2014 17:16:34 -0500
Subject: [racket] DrRacket 6.1 or Hardware Issue?
In-Reply-To: <CA+APS0CQ-3qqQvHBCwDQ4zsKR_UOwq7dgDdPF5mK7i9wue2SDw@mail.gmail.com>
References: <CA+APS0CQ-3qqQvHBCwDQ4zsKR_UOwq7dgDdPF5mK7i9wue2SDw@mail.gmail.com>
Message-ID: <CAL3TdOMnojmQr88JjSBQah8JZQwKBro7dgsmHBC_s-PdLaHdOQ@mail.gmail.com>

I've pushed a fix for DrRacket so that it limits online expansion to
the same limits that are in place for running programs.

Thanks,
Robby

On Mon, Sep 8, 2014 at 4:23 PM, Sean Kanaley <skanaley at gmail.com> wrote:
> After investigating a near complete lock up some more (saved by TTY window
> and killing DrRacket after seeing it with 3GB+ residency), it's clearly
> correlated with background expansion of an accidentally infinitely looping
> macro. A few questions..
>
> 1. Is the OS supposed to break and enter an infinite swap loop? If that is
> just a fact of life for a program requesting too much memory simultaneously
> (and an ever increasing amount), well for example ghc just shuts itself down
> claiming "too much memory requested" or something like that.
>
> 2. Is this related to my very long time ago issue where I removed a bad 1GB
> from the 3rd slot, lazily leaving the 4th slot where it is, causing me to
> have 3GB instead of 4, even though it's "supposed to" have 4 based on the
> max slot (I have never encountered this sort of system freeze before
> however)? Maybe the OS would more gracefully handle the paging?
>
> 3. Can DrRacket simply realize it's expanding the same thing and not demand
> infinite memory (I don't understand the system on the implementation level,
> sorry if that's off base).
>
> The solution, aside from me not writing incorrect macros is to disable
> background expansion. Relatedly, DrRacket doesn't seem able to reclaim this
> space without being restarted. Closing the file and clicking the garbage
> collection thing only recovers say 20MB out of 2GB if I turn off expansion
> before the OS locks.
>
> Lastly, a related possible bug I've encountered since 6.1 is the infinite
> garbage collection loop, where the OS remains responsive but DrRacket has to
> be restarted, but this one I have no idea where it comes from only that it
> seems to happen after a while, and fairly reliably so.
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>

From uselpa.list at gmail.com  Fri Sep 12 18:21:11 2014
From: uselpa.list at gmail.com (Patrick Useldinger)
Date: Sat, 13 Sep 2014 00:21:11 +0200
Subject: [racket] integer-sqrt
Message-ID: <541371D7.5090206@gmail.com>

Hi all,

according to the documentation, the above procedure accepts a *number* 
and returns an integer.

But (at least in Racket 6.1) there's an error message stating that the 
parameter needs to be an *integer*:

---------------------------------------
  > (integer-sqrt 151.29)
   integer-sqrt: contract violation
   expected: integer?
   given: 151.29
---------------------------------------

Is this an error in the documentation or in the implementation?

Regards,
-Patrick

From robby at eecs.northwestern.edu  Fri Sep 12 18:27:50 2014
From: robby at eecs.northwestern.edu (Robby Findler)
Date: Fri, 12 Sep 2014 17:27:50 -0500
Subject: [racket] integer-sqrt
In-Reply-To: <541371D7.5090206@gmail.com>
References: <541371D7.5090206@gmail.com>
Message-ID: <CAL3TdONqBKFyQ8eeOHyU7iQXhqRYQmFSQdZyU3TJT5vsgQG3pg@mail.gmail.com>

The docs, I believe.

Robby

On Fri, Sep 12, 2014 at 5:21 PM, Patrick Useldinger
<uselpa.list at gmail.com> wrote:
> Hi all,
>
> according to the documentation, the above procedure accepts a *number* and
> returns an integer.
>
> But (at least in Racket 6.1) there's an error message stating that the
> parameter needs to be an *integer*:
>
> ---------------------------------------
>  > (integer-sqrt 151.29)
>   integer-sqrt: contract violation
>   expected: integer?
>   given: 151.29
> ---------------------------------------
>
> Is this an error in the documentation or in the implementation?
>
> Regards,
> -Patrick
> ____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users

From mflatt at cs.utah.edu  Fri Sep 12 18:34:22 2014
From: mflatt at cs.utah.edu (Matthew Flatt)
Date: Fri, 12 Sep 2014 16:34:22 -0600
Subject: [racket] error writing to stream port (Broken pipe; errno=32)
In-Reply-To: <n88fvfwy5lh.fsf@almelo.raji.xor>
References: <n88fvfwy5lh.fsf@almelo.raji.xor>
Message-ID: <20140912223424.5BA9565018C@mail-svr1.cs.utah.edu>

To make the program concrete, suppose we have "a.rkt" as

 #lang racket/base

 (define (epipe? x)
   (and
    (exn:fail:filesystem:errno? x) 
    (equal? '(32 . posix) (exn:fail:filesystem:errno-errno x))))
 (define (epipe-handler x)
   "ignore")

 (with-handlers ([epipe? epipe-handler])
  (let loop ()
    (displayln "hi")
    (loop)))

and then we run

 racket a.rkt | more

and then we hit "q" to exit `more`.

In that case, we will get

 error writing to stream port
  system error: Broken pipe; errno=32

The problem is that `epipe-handler` returns a string that becomes the
result of the whole `with-handlers`. When Racket attempts to print that
string, it fails (outside the `with-handlers`, of course). In other
words, the exception reports the failure to print "ignore", not a
failure from `(displayln "hi")`.


In this little example, changing `epipe-handler` to

 (define (epipe-handler x)
   (void))

avoids the problem, because a void result isn't printed. Another solution is

 (define (epipe-handler x)
   (exit 0))

to immediately exit, instead of returning a value.




At Fri, 12 Sep 2014 21:57:30 +0200, Wolfgang Hukriede wrote:
> Hi Matthias, nice to meet you! Yes, this was what I was looking for.
> 
> For the records:
> 
> (define (epipe? x)
>   (and
>    (exn:fail:filesystem:errno? x) 
>    (equal? '(32 . posix) (exn:fail:filesystem:errno-errno x))))
> 
> (define (epipe-handler x)
>   "ignore")
> 
> (with-handlers
>  ((epipe? epipe-handler))
>  transput...)
> 
> Many thanks!
> 
> Results are mixed though. Usage like 
> 
> $ mzscheme --no-init-file --load pipetest.scm | head
> 
> mostly works okay, but depending on how and how much is written to
> standard out and on the parameters to "head" failures reproducibly
> occur: in some situations the signal is not caught, yielding a "system
> error: Broken pipe; errno=32" message (versions 6.1 and 5.93 on
> freebsd, and 4.1 on generic linux, 6.1 precompiled racket on
> archlinux), or sometimes the pipe even just hangs (6.1 on freebsd).
> These were the combinations I tried.
> (On 4.1 exn:fail:filesystem was used instead of non-existing
> exn:fail:filesystem:errno.)
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users

From matthias at ccs.neu.edu  Fri Sep 12 18:42:39 2014
From: matthias at ccs.neu.edu (Matthias Felleisen)
Date: Fri, 12 Sep 2014 18:42:39 -0400
Subject: [racket] integer-sqrt
In-Reply-To: <CAL3TdONqBKFyQ8eeOHyU7iQXhqRYQmFSQdZyU3TJT5vsgQG3pg@mail.gmail.com>
References: <541371D7.5090206@gmail.com>
 <CAL3TdONqBKFyQ8eeOHyU7iQXhqRYQmFSQdZyU3TJT5vsgQG3pg@mail.gmail.com>
Message-ID: <66074A93-14E0-4040-8F92-7A8860739974@ccs.neu.edu>


The docs contradict themselves here :-) 


On Sep 12, 2014, at 6:27 PM, Robby Findler <robby at eecs.northwestern.edu> wrote:

> The docs, I believe.
> 
> Robby
> 
> On Fri, Sep 12, 2014 at 5:21 PM, Patrick Useldinger
> <uselpa.list at gmail.com> wrote:
>> Hi all,
>> 
>> according to the documentation, the above procedure accepts a *number* and
>> returns an integer.
>> 
>> But (at least in Racket 6.1) there's an error message stating that the
>> parameter needs to be an *integer*:
>> 
>> ---------------------------------------
>>> (integer-sqrt 151.29)
>>  integer-sqrt: contract violation
>>  expected: integer?
>>  given: 151.29
>> ---------------------------------------
>> 
>> Is this an error in the documentation or in the implementation?
>> 
>> Regards,
>> -Patrick
>> ____________________
>> Racket Users list:
>> http://lists.racket-lang.org/users
> ____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users



From matthias at ccs.neu.edu  Fri Sep 12 19:03:16 2014
From: matthias at ccs.neu.edu (Matthias Felleisen)
Date: Fri, 12 Sep 2014 19:03:16 -0400
Subject: [racket] integer-sqrt
In-Reply-To: <541371D7.5090206@gmail.com>
References: <541371D7.5090206@gmail.com>
Message-ID: <B498CA23-F868-412D-8047-C05D16CB3807@ccs.neu.edu>


I have pushed a repair for the teaching languages. The Racket docs for git head seem to be okay. 


On Sep 12, 2014, at 6:21 PM, Patrick Useldinger <uselpa.list at gmail.com> wrote:

> Hi all,
> 
> according to the documentation, the above procedure accepts a *number* and returns an integer.
> 
> But (at least in Racket 6.1) there's an error message stating that the parameter needs to be an *integer*:
> 
> ---------------------------------------
> > (integer-sqrt 151.29)
>  integer-sqrt: contract violation
>  expected: integer?
>  given: 151.29
> ---------------------------------------
> 
> Is this an error in the documentation or in the implementation?
> 
> Regards,
> -Patrick
> ____________________
> Racket Users list:
> http://lists.racket-lang.org/users



From oev-racket at sibmail.com  Sat Sep 13 10:28:10 2014
From: oev-racket at sibmail.com (Evgeny Odegov)
Date: Sat, 13 Sep 2014 21:28:10 +0700 (NOVT)
Subject: [racket] Some problem with hash-update!
Message-ID: <52456.79.136.149.232.1410618490.squirrel@www.sibmail.com>

Hi, all!

I have a problem with the code below:

(define result (make-hash))
(define (callback dwTextID lpszInfoText lpvUser)
  (hash-update! result
                lpszInfoText
                (lambda (id-lst)
                  (define v (list dwTextID))
                  (if id-lst (append id-lst v) v))
                #f)
  #t)

(The `callback' is passed to some enumeration function via FFI. Its type
is defined as:

(define _TXT_ENUM_INFOTEXTS_PROC
  (_cprocedure (list _DWORD _string/locale _pointer)
               _bool
               #:abi 'sysv)).


It seems like problem not with FFI, but I could give more details if needed.)

Mostly, parameter `lpszInfoText' is unique, so length of `is-lst' is
rarely more than one.
But, when there are lot of entries with empty `lpszInfoText', the list
becomes big.
At some moment something goes wrong, and my program fails with error:

append: contract violation
  expected: list?
  given: (list* 80 579 580 581 582 583 584 585 586 587 588 589 590 591 592
593 594 595 596 597 598 599 600 602 603 604 605 606 607 608 609 610 612
613 614 615 616 617 618 619 621 622 623 624 625 626 627 628 629 631 632
633 634 635 636 637 638 640 641 642 6...


I add a code that prints the wrong list before `append'. Result differs
from run to run, but always belongs to one of two categories.

Examples:
* http://pastebin.com/bnNYY8EL
* http://pastebin.com/x7nDdnWn


If I change the code to below, it works without problems:

(define result (make-hash))
(define big-list (list))
(define (callback dwTextID lpszInfoText lpvUser)
  (if (equal? lpszInfoText "")
      (set! big-list (cons dwTextID big-list))
      (hash-update! result
                    lpszInfoText
                    (lambda (id-lst)
                      (define v (list dwTextID))
                      (if id-lst (append id-lst v) v))
                    #f))
  #t)


Racket version: 6.1 release, x86, 32-bit
OS: Windows XP Prof 32 bit


From oev-racket at sibmail.com  Sat Sep 13 11:45:28 2014
From: oev-racket at sibmail.com (Evgeny Odegov)
Date: Sat, 13 Sep 2014 22:45:28 +0700 (NOVT)
Subject: [racket] Some problem with hash-update!
In-Reply-To: <52456.79.136.149.232.1410618490.squirrel@www.sibmail.com>
References: <52456.79.136.149.232.1410618490.squirrel@www.sibmail.com>
Message-ID: <52709.79.136.149.232.1410623128.squirrel@www.sibmail.com>

Oh, I'm wrong. The last code is not equivalent to initial.
The equivalent code has this problem:

(define result (make-hash))
(define big-list (list))
(define (callback dwTextID lpszInfoText lpvUser)
  (if (equal? lpszInfoText "")
      (set! big-list (append big-list (list dwTextID)))
      (hash-update! result
                    lpszInfoText
                    (lambda (id-lst)
                      (define v (list dwTextID))
                      (append id-lst v))
                    '()))
  #t)


The code with `cons' instead `append' works fine:

(define result (make-hash))
(define (callback dwTextID lpszInfoText lpvUser)
  (hash-update! result
                lpszInfoText
                (lambda (id-lst)
                  (define new-id-lst (cons dwTextID id-lst))
                  (unless (list? new-id-lst)

                    (error "~v" new-id-lst))
                  new-id-lst)
                '())
  #t)





From cody at cs.brown.edu  Sun Sep 14 03:10:00 2014
From: cody at cs.brown.edu (Mello, Cody)
Date: Sun, 14 Sep 2014 03:10:00 -0400
Subject: [racket] Setting umask from within Racket
Message-ID: <CAK1kL9EUXE-YLEK2-9q_MfUdsaWcq9+FAjsPRvj8EVxSEfQSBA@mail.gmail.com>

Hello everyone,

Is there a way to set the umask of the process from within Racket? Or to
specify the mode of a newly created file? I've been looking through the
documentation, but I'm not finding anything along these lines. I was hoping
that I could do something like the following from within my program:

(parameterize ([umask #o007])
  (display-to-file 5 "test-file"))

Or like this:

(display-to-file 5 "test-file" #:perms #o660)

I could change the permissions of the file with
file-or-directory-permissions after creating the file, but what I really
want to do is create the file with those permissions from the start. Is
there a good way to accomplish this? Am I missing something in the
documentation?

Thanks!

- Cody
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140914/75786421/attachment.html>

From neil at neilvandyke.org  Sun Sep 14 04:09:53 2014
From: neil at neilvandyke.org (Neil Van Dyke)
Date: Sun, 14 Sep 2014 04:09:53 -0400
Subject: [racket] Setting umask from within Racket
In-Reply-To: <CAK1kL9EUXE-YLEK2-9q_MfUdsaWcq9+FAjsPRvj8EVxSEfQSBA@mail.gmail.com>
References: <CAK1kL9EUXE-YLEK2-9q_MfUdsaWcq9+FAjsPRvj8EVxSEfQSBA@mail.gmail.com>
Message-ID: <54154D51.8050102@neilvandyke.org>

If you don't find exactly the solution you want, possible measures:

* If the concern is security, create the files in a directory on which 
you've set sufficiently restrictive permissions.  (Even if you always 
set the file permissions immediately after opening the file for writing, 
there could be a race condition that lets the file be opened between 
create and permissions set.)

* If you are ruthless, you could set the umask for entire process, using 
the FFI for the system call.

* If you don't want to FFI, but you really-really want to second-guess 
the parent process's umask, you could kludge it in pure Racket plus host 
processes, at least on Unix-ish systems.  Have the Racket program check 
its umask, and if not the desired one, use process/shell API to create a 
new process of the same Racket program, with the same command line, but 
under the new umask.

* Consider writing your own pure Racket convenience procedures for 
opening files with desired permissions.  (Maybe you could even have it 
secretly do tricks to do the non-atomic create and chmod securely 
without messing with FFI or umask.  Some tricks of moving and copying 
files might occasionally run afoul of unusual situations in the 
filesystem, and of finer points of different filesystem semantics, but 
in practice, most people will never encounter any problems with simple 
tricks.)

Side note: The Unix o/g/a-r/w/x permissions are pragmatic and archaic, 
so (like much of Unix) don't read too much into them, but they seem to 
have survived OK.  And there seems to be less need for fancier ACLs and 
compartmentalization and such, now that we've moved most multi-user 
access control out of the filesystem and into Internet protocols and 
larger structures of same.  And into VMs.

Neil V.


From brettstahlman at gmail.com  Sun Sep 14 07:57:34 2014
From: brettstahlman at gmail.com (Brett Stahlman)
Date: Sun, 14 Sep 2014 06:57:34 -0500
Subject: [racket] Possible bug involving sync on tcp ports
Message-ID: <CAOyxqK=ZZtaPWUqM-6M17EfW4LoUEviKqchNxzB+AsfkYjmVGQ@mail.gmail.com>

Hello,

I have the following loop in a tcp client program that sends forms typed by
user at the terminal to a listening server, and receives such forms typed by
other clients (broadcast to all clients by the server).

(define server-msg-evt (wrap-evt in handle-server-msg))
(define rdy-to-send-evt (wrap-evt (standard-input-port) handle-rdy-to-send))

(let loop ()
  (sync rdy-to-send-evt server-msg-evt)
  (loop))

The problem is that the server-msg-evt isn't reliably ready when the server
flushes a message. I know the messages are being sent and flushed properly
by
the server because if I type a new form on stdin and hit enter, I see the
following sequence:

1. Client processes new stdin form in handle-rdy-to-send
2. Client loops and calls sync
3. The server-msg-evt is *immediately* ready, and client handles the
   message in handle-server-msg.
Note: The message handled in step 3 may have been sent long before, but
the client remained blocked in sync until the (unrelated) stdin event
unblocked it.

The really weird part is that the client blocks in this manner even if I
change the sync to a sync/timeout!

Any clue as to what could be going on here? I'm running both client and
server
in a DOS box on a Windows 7 machine.

Thanks,
Brett Stahlman
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140914/407b5637/attachment.html>

From mflatt at cs.utah.edu  Sun Sep 14 08:40:07 2014
From: mflatt at cs.utah.edu (Matthew Flatt)
Date: Sun, 14 Sep 2014 14:40:07 +0200
Subject: [racket] Some problem with hash-update!
In-Reply-To: <52709.79.136.149.232.1410623128.squirrel@www.sibmail.com>
References: <52456.79.136.149.232.1410618490.squirrel@www.sibmail.com>
 <52709.79.136.149.232.1410623128.squirrel@www.sibmail.com>
Message-ID: <20140914124010.BD2A765019F@mail-svr1.cs.utah.edu>

My best guess is that it's an issue with memory management, because
memory management gets a lot trickier when you call a function that can
call back into Racket. A GC can happen during the callback, and
therefore during the dynamic extent of the foreign-function call, which
can move arguments that were passed to the foreign function, for example.

Would it be possible to provide a complete example?

At Sat, 13 Sep 2014 22:45:28 +0700 (NOVT), "Evgeny Odegov" wrote:
> Oh, I'm wrong. The last code is not equivalent to initial.
> The equivalent code has this problem:
> 
> (define result (make-hash))
> (define big-list (list))
> (define (callback dwTextID lpszInfoText lpvUser)
>   (if (equal? lpszInfoText "")
>       (set! big-list (append big-list (list dwTextID)))
>       (hash-update! result
>                     lpszInfoText
>                     (lambda (id-lst)
>                       (define v (list dwTextID))
>                       (append id-lst v))
>                     '()))
>   #t)
> 
> 
> The code with `cons' instead `append' works fine:
> 
> (define result (make-hash))
> (define (callback dwTextID lpszInfoText lpvUser)
>   (hash-update! result
>                 lpszInfoText
>                 (lambda (id-lst)
>                   (define new-id-lst (cons dwTextID id-lst))
>                   (unless (list? new-id-lst)
> 
>                     (error "~v" new-id-lst))
>                   new-id-lst)
>                 '())
>   #t)
> 
> 
> 
> 
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users

From mflatt at cs.utah.edu  Sun Sep 14 08:56:52 2014
From: mflatt at cs.utah.edu (Matthew Flatt)
Date: Sun, 14 Sep 2014 14:56:52 +0200
Subject: [racket] Possible bug involving sync on tcp ports
In-Reply-To: <CAOyxqK=ZZtaPWUqM-6M17EfW4LoUEviKqchNxzB+AsfkYjmVGQ@mail.gmail.com>
References: <CAOyxqK=ZZtaPWUqM-6M17EfW4LoUEviKqchNxzB+AsfkYjmVGQ@mail.gmail.com>
Message-ID: <20140914125656.22E386501A8@mail-svr1.cs.utah.edu>

It's possible that something is wrong with `sync` and TCP ports ---
although the test suite at least includes that combination --- but can
you provide a complete program?

For example, I'm interested in what `handle-rdy-to-send` does. In the
part that you show, the function will be triggered when one byte is
available on stdin. If `handle-rdy-to-send` tries to read more bytes
than are immediately available, then it can block; i.e., the blocking
part could be in `handle-rdy-to-send`, not in `sync`. That would also
explain why using `sync/timeout` doesn't change anything.

Again, that's just an example of what might be wrong other than `sync`.
A complete program would let us sort out the possibilities.

At Sun, 14 Sep 2014 06:57:34 -0500, Brett Stahlman wrote:
> Hello,
> 
> I have the following loop in a tcp client program that sends forms typed by
> user at the terminal to a listening server, and receives such forms typed by
> other clients (broadcast to all clients by the server).
> 
> (define server-msg-evt (wrap-evt in handle-server-msg))
> (define rdy-to-send-evt (wrap-evt (standard-input-port) handle-rdy-to-send))
> 
> (let loop ()
>   (sync rdy-to-send-evt server-msg-evt)
>   (loop))
> 
> The problem is that the server-msg-evt isn't reliably ready when the server
> flushes a message. I know the messages are being sent and flushed properly
> by
> the server because if I type a new form on stdin and hit enter, I see the
> following sequence:
> 
> 1. Client processes new stdin form in handle-rdy-to-send
> 2. Client loops and calls sync
> 3. The server-msg-evt is *immediately* ready, and client handles the
>    message in handle-server-msg.
> Note: The message handled in step 3 may have been sent long before, but
> the client remained blocked in sync until the (unrelated) stdin event
> unblocked it.
> 
> The really weird part is that the client blocks in this manner even if I
> change the sync to a sync/timeout!
> 
> Any clue as to what could be going on here? I'm running both client and
> server
> in a DOS box on a Windows 7 machine.
> 
> Thanks,
> Brett Stahlman
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users

From brettstahlman at gmail.com  Sun Sep 14 09:20:26 2014
From: brettstahlman at gmail.com (Brett Stahlman)
Date: Sun, 14 Sep 2014 08:20:26 -0500
Subject: [racket] Possible bug involving sync on tcp ports
In-Reply-To: <20140914125656.22E386501A8@mail-svr1.cs.utah.edu>
References: <CAOyxqK=ZZtaPWUqM-6M17EfW4LoUEviKqchNxzB+AsfkYjmVGQ@mail.gmail.com>
 <20140914125656.22E386501A8@mail-svr1.cs.utah.edu>
Message-ID: <CAOyxqKkS==oQu7-UqcRK4Bc-YyT7JA8RM9ST+wf3MYz2m3OatA@mail.gmail.com>

On Sep 14, 2014 7:56 AM, "Matthew Flatt" <mflatt at cs.utah.edu> wrote:
>
> It's possible that something is wrong with `sync` and TCP ports ---
> although the test suite at least includes that combination --- but can
> you provide a complete program?
>
> For example, I'm interested in what `handle-rdy-to-send` does. In the
> part that you show, the function will be triggered when one byte is
> available on stdin. If `handle-rdy-to-send` tries to read more bytes
> than are immediately available, then it can block; i.e., the blocking
> part could be in `handle-rdy-to-send`, not in `sync`. That would also
> explain why using `sync/timeout` doesn't change anything.

Actually, I know it's blocking in the sync (or sync/timeout) because the
actual code is littered with immediately-flushed printf's, and there's one
right before and one right after the sync call. I will try to put together
a version to send you, though. Would you prefer it with the debug printf's
stripped out?

Thanks, Brett S.

>
> Again, that's just an example of what might be wrong other than `sync`.
> A complete program would let us sort out the possibilities.
>
> At Sun, 14 Sep 2014 06:57:34 -0500, Brett Stahlman wrote:
> > Hello,
> >
> > I have the following loop in a tcp client program that sends forms
typed by
> > user at the terminal to a listening server, and receives such forms
typed by
> > other clients (broadcast to all clients by the server).
> >
> > (define server-msg-evt (wrap-evt in handle-server-msg))
> > (define rdy-to-send-evt (wrap-evt (standard-input-port)
handle-rdy-to-send))
> >
> > (let loop ()
> >   (sync rdy-to-send-evt server-msg-evt)
> >   (loop))
> >
> > The problem is that the server-msg-evt isn't reliably ready when the
server
> > flushes a message. I know the messages are being sent and flushed
properly
> > by
> > the server because if I type a new form on stdin and hit enter, I see
the
> > following sequence:
> >
> > 1. Client processes new stdin form in handle-rdy-to-send
> > 2. Client loops and calls sync
> > 3. The server-msg-evt is *immediately* ready, and client handles the
> >    message in handle-server-msg.
> > Note: The message handled in step 3 may have been sent long before, but
> > the client remained blocked in sync until the (unrelated) stdin event
> > unblocked it.
> >
> > The really weird part is that the client blocks in this manner even if I
> > change the sync to a sync/timeout!
> >
> > Any clue as to what could be going on here? I'm running both client and
> > server
> > in a DOS box on a Windows 7 machine.
> >
> > Thanks,
> > Brett Stahlman
> > ____________________
> >   Racket Users list:
> >   http://lists.racket-lang.org/users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140914/a347c427/attachment-0001.html>

From mflatt at cs.utah.edu  Sun Sep 14 09:24:18 2014
From: mflatt at cs.utah.edu (Matthew Flatt)
Date: Sun, 14 Sep 2014 15:24:18 +0200
Subject: [racket] Possible bug involving sync on tcp ports
In-Reply-To: <CAOyxqKkS==oQu7-UqcRK4Bc-YyT7JA8RM9ST+wf3MYz2m3OatA@mail.gmail.com>
References: <CAOyxqK=ZZtaPWUqM-6M17EfW4LoUEviKqchNxzB+AsfkYjmVGQ@mail.gmail.com>
 <20140914125656.22E386501A8@mail-svr1.cs.utah.edu>
 <CAOyxqKkS==oQu7-UqcRK4Bc-YyT7JA8RM9ST+wf3MYz2m3OatA@mail.gmail.com>
Message-ID: <20140914132423.08A346501B1@mail-svr1.cs.utah.edu>

At Sun, 14 Sep 2014 08:20:26 -0500, Brett Stahlman wrote:
> I will try to put together
> a version to send you, though. Would you prefer it with the debug printf's
> stripped out?

Either way is fine.


From mflatt at cs.utah.edu  Sun Sep 14 10:23:32 2014
From: mflatt at cs.utah.edu (Matthew Flatt)
Date: Sun, 14 Sep 2014 16:23:32 +0200
Subject: [racket] Possible bug involving sync on tcp ports
In-Reply-To: <CAOyxqKkS==oQu7-UqcRK4Bc-YyT7JA8RM9ST+wf3MYz2m3OatA@mail.gmail.com>
References: <CAOyxqK=ZZtaPWUqM-6M17EfW4LoUEviKqchNxzB+AsfkYjmVGQ@mail.gmail.com>
 <20140914125656.22E386501A8@mail-svr1.cs.utah.edu>
 <CAOyxqKkS==oQu7-UqcRK4Bc-YyT7JA8RM9ST+wf3MYz2m3OatA@mail.gmail.com>
Message-ID: <20140914142336.7EE6F6501C1@mail-svr1.cs.utah.edu>

I think the problem is in `standard-input-port` from `rnrs/io/ports-6`.

If you use `current-input-port` instead of `standard-input-port`, does
the client behave correctly?

At Sun, 14 Sep 2014 08:20:26 -0500, Brett Stahlman wrote:
> On Sep 14, 2014 7:56 AM, "Matthew Flatt" <mflatt at cs.utah.edu> wrote:
> >
> > It's possible that something is wrong with `sync` and TCP ports ---
> > although the test suite at least includes that combination --- but can
> > you provide a complete program?
> >
> > For example, I'm interested in what `handle-rdy-to-send` does. In the
> > part that you show, the function will be triggered when one byte is
> > available on stdin. If `handle-rdy-to-send` tries to read more bytes
> > than are immediately available, then it can block; i.e., the blocking
> > part could be in `handle-rdy-to-send`, not in `sync`. That would also
> > explain why using `sync/timeout` doesn't change anything.
> 
> Actually, I know it's blocking in the sync (or sync/timeout) because the
> actual code is littered with immediately-flushed printf's, and there's one
> right before and one right after the sync call. I will try to put together
> a version to send you, though. Would you prefer it with the debug printf's
> stripped out?
> 
> Thanks, Brett S.
> 
> >
> > Again, that's just an example of what might be wrong other than `sync`.
> > A complete program would let us sort out the possibilities.
> >
> > At Sun, 14 Sep 2014 06:57:34 -0500, Brett Stahlman wrote:
> > > Hello,
> > >
> > > I have the following loop in a tcp client program that sends forms
> typed by
> > > user at the terminal to a listening server, and receives such forms
> typed by
> > > other clients (broadcast to all clients by the server).
> > >
> > > (define server-msg-evt (wrap-evt in handle-server-msg))
> > > (define rdy-to-send-evt (wrap-evt (standard-input-port)
> handle-rdy-to-send))
> > >
> > > (let loop ()
> > >   (sync rdy-to-send-evt server-msg-evt)
> > >   (loop))
> > >
> > > The problem is that the server-msg-evt isn't reliably ready when the
> server
> > > flushes a message. I know the messages are being sent and flushed
> properly
> > > by
> > > the server because if I type a new form on stdin and hit enter, I see
> the
> > > following sequence:
> > >
> > > 1. Client processes new stdin form in handle-rdy-to-send
> > > 2. Client loops and calls sync
> > > 3. The server-msg-evt is *immediately* ready, and client handles the
> > >    message in handle-server-msg.
> > > Note: The message handled in step 3 may have been sent long before, but
> > > the client remained blocked in sync until the (unrelated) stdin event
> > > unblocked it.
> > >
> > > The really weird part is that the client blocks in this manner even if I
> > > change the sync to a sync/timeout!
> > >
> > > Any clue as to what could be going on here? I'm running both client and
> > > server
> > > in a DOS box on a Windows 7 machine.
> > >
> > > Thanks,
> > > Brett Stahlman
> > > ____________________
> > >   Racket Users list:
> > >   http://lists.racket-lang.org/users
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users

From brettstahlman at gmail.com  Sun Sep 14 10:38:49 2014
From: brettstahlman at gmail.com (Brett Stahlman)
Date: Sun, 14 Sep 2014 09:38:49 -0500
Subject: [racket] Possible bug involving sync on tcp ports
In-Reply-To: <20140914142336.7EE6F6501C1@mail-svr1.cs.utah.edu>
References: <CAOyxqK=ZZtaPWUqM-6M17EfW4LoUEviKqchNxzB+AsfkYjmVGQ@mail.gmail.com>
 <20140914125656.22E386501A8@mail-svr1.cs.utah.edu>
 <CAOyxqKkS==oQu7-UqcRK4Bc-YyT7JA8RM9ST+wf3MYz2m3OatA@mail.gmail.com>
 <20140914142336.7EE6F6501C1@mail-svr1.cs.utah.edu>
Message-ID: <CAOyxqKk0u_EkZSbpZUm4ikDaeqWUAD=668kk2wAyw39AjjSMEA@mail.gmail.com>

On Sun, Sep 14, 2014 at 9:23 AM, Matthew Flatt <mflatt at cs.utah.edu> wrote:
> I think the problem is in `standard-input-port` from `rnrs/io/ports-6`.
>
> If you use `current-input-port` instead of `standard-input-port`, does
> the client behave correctly?

It does indeed! Can't recall why I used standard-input-port to begin
with, but I'm thinking this could account for some of the weirdness
I've been seeing. I didn't mention it in my original posts, but part
of the reason I went to such a convoluted strategy for peeking and
reading ports was that byte-ready? was blocking as well. Actually,
I've been planning to change the client/server architecture entirely -
to one in which each port reader operates in a dedicated thread, and
simply blocks until a new form can be read, at which point it writes
the data to an async channel to be processed at a higher level - but I
didn't want to proceed without resolving these issues...

Out of curiosity, what is the proper way to set the current input port
to stdin (if somehow it's been changed from the default)? (It hasn't,
in my case, but perhaps the reason I went to standard-input-port in
the first place was that there didn't appear to be any guarantee that
the current input port would be stdin...)

Thanks,
Brett S.

>
> At Sun, 14 Sep 2014 08:20:26 -0500, Brett Stahlman wrote:
>> On Sep 14, 2014 7:56 AM, "Matthew Flatt" <mflatt at cs.utah.edu> wrote:
>> >
>> > It's possible that something is wrong with `sync` and TCP ports ---
>> > although the test suite at least includes that combination --- but can
>> > you provide a complete program?
>> >
>> > For example, I'm interested in what `handle-rdy-to-send` does. In the
>> > part that you show, the function will be triggered when one byte is
>> > available on stdin. If `handle-rdy-to-send` tries to read more bytes
>> > than are immediately available, then it can block; i.e., the blocking
>> > part could be in `handle-rdy-to-send`, not in `sync`. That would also
>> > explain why using `sync/timeout` doesn't change anything.
>>
>> Actually, I know it's blocking in the sync (or sync/timeout) because the
>> actual code is littered with immediately-flushed printf's, and there's one
>> right before and one right after the sync call. I will try to put together
>> a version to send you, though. Would you prefer it with the debug printf's
>> stripped out?
>>
>> Thanks, Brett S.
>>
>> >
>> > Again, that's just an example of what might be wrong other than `sync`.
>> > A complete program would let us sort out the possibilities.
>> >
>> > At Sun, 14 Sep 2014 06:57:34 -0500, Brett Stahlman wrote:
>> > > Hello,
>> > >
>> > > I have the following loop in a tcp client program that sends forms
>> typed by
>> > > user at the terminal to a listening server, and receives such forms
>> typed by
>> > > other clients (broadcast to all clients by the server).
>> > >
>> > > (define server-msg-evt (wrap-evt in handle-server-msg))
>> > > (define rdy-to-send-evt (wrap-evt (standard-input-port)
>> handle-rdy-to-send))
>> > >
>> > > (let loop ()
>> > >   (sync rdy-to-send-evt server-msg-evt)
>> > >   (loop))
>> > >
>> > > The problem is that the server-msg-evt isn't reliably ready when the
>> server
>> > > flushes a message. I know the messages are being sent and flushed
>> properly
>> > > by
>> > > the server because if I type a new form on stdin and hit enter, I see
>> the
>> > > following sequence:
>> > >
>> > > 1. Client processes new stdin form in handle-rdy-to-send
>> > > 2. Client loops and calls sync
>> > > 3. The server-msg-evt is *immediately* ready, and client handles the
>> > >    message in handle-server-msg.
>> > > Note: The message handled in step 3 may have been sent long before, but
>> > > the client remained blocked in sync until the (unrelated) stdin event
>> > > unblocked it.
>> > >
>> > > The really weird part is that the client blocks in this manner even if I
>> > > change the sync to a sync/timeout!
>> > >
>> > > Any clue as to what could be going on here? I'm running both client and
>> > > server
>> > > in a DOS box on a Windows 7 machine.
>> > >
>> > > Thanks,
>> > > Brett Stahlman
>> > > ____________________
>> > >   Racket Users list:
>> > >   http://lists.racket-lang.org/users
>> ____________________
>>   Racket Users list:
>>   http://lists.racket-lang.org/users

From mflatt at cs.utah.edu  Sun Sep 14 10:49:56 2014
From: mflatt at cs.utah.edu (Matthew Flatt)
Date: Sun, 14 Sep 2014 16:49:56 +0200
Subject: [racket] Possible bug involving sync on tcp ports
In-Reply-To: <CAOyxqKk0u_EkZSbpZUm4ikDaeqWUAD=668kk2wAyw39AjjSMEA@mail.gmail.com>
References: <CAOyxqK=ZZtaPWUqM-6M17EfW4LoUEviKqchNxzB+AsfkYjmVGQ@mail.gmail.com>
 <20140914125656.22E386501A8@mail-svr1.cs.utah.edu>
 <CAOyxqKkS==oQu7-UqcRK4Bc-YyT7JA8RM9ST+wf3MYz2m3OatA@mail.gmail.com>
 <20140914142336.7EE6F6501C1@mail-svr1.cs.utah.edu>
 <CAOyxqKk0u_EkZSbpZUm4ikDaeqWUAD=668kk2wAyw39AjjSMEA@mail.gmail.com>
Message-ID: <20140914145002.69ACF6501C4@mail-svr1.cs.utah.edu>

I'll push a repair to make `rnrs/io/ports-6` ports work right with
`sync`.

There's not really a way to get the OS-process-level stdin and stdout
in Racket, except by using scheme_make_fd_input_port() via
`ffi/unsafe`. The intent is that the initial ports, which correspond to
stdin and stdout, can be made inaccessible to untrusted code by
changing the `current-input-port` and `current-output-port` parameters.

Usually, it's best to just use `(current-input-port)`. The initial
parameter value corresponds to stdin for a stand-alone Racket process;
in other cases, a program should normally accept whatever adjustments
are made by its environment. If you really need the OS-process-level
stdin and stdout, through, then `ffi/unsafe' can reach them.

The `standard-input-port` function in Racket's R6RS implementation,
meanwhile, returns the current input port at the time that the R6RS
library was loaded. So, it's not necessary stdin in the sense of the OS
process, but it's "stdin" in the sense of the being the designated
source for input when the library starts.

At Sun, 14 Sep 2014 09:38:49 -0500, Brett Stahlman wrote:
> On Sun, Sep 14, 2014 at 9:23 AM, Matthew Flatt <mflatt at cs.utah.edu> wrote:
> > I think the problem is in `standard-input-port` from `rnrs/io/ports-6`.
> >
> > If you use `current-input-port` instead of `standard-input-port`, does
> > the client behave correctly?
> 
> It does indeed! Can't recall why I used standard-input-port to begin
> with, but I'm thinking this could account for some of the weirdness
> I've been seeing. I didn't mention it in my original posts, but part
> of the reason I went to such a convoluted strategy for peeking and
> reading ports was that byte-ready? was blocking as well. Actually,
> I've been planning to change the client/server architecture entirely -
> to one in which each port reader operates in a dedicated thread, and
> simply blocks until a new form can be read, at which point it writes
> the data to an async channel to be processed at a higher level - but I
> didn't want to proceed without resolving these issues...
> 
> Out of curiosity, what is the proper way to set the current input port
> to stdin (if somehow it's been changed from the default)? (It hasn't,
> in my case, but perhaps the reason I went to standard-input-port in
> the first place was that there didn't appear to be any guarantee that
> the current input port would be stdin...)
> 
> Thanks,
> Brett S.
> 
> >
> > At Sun, 14 Sep 2014 08:20:26 -0500, Brett Stahlman wrote:
> >> On Sep 14, 2014 7:56 AM, "Matthew Flatt" <mflatt at cs.utah.edu> wrote:
> >> >
> >> > It's possible that something is wrong with `sync` and TCP ports ---
> >> > although the test suite at least includes that combination --- but can
> >> > you provide a complete program?
> >> >
> >> > For example, I'm interested in what `handle-rdy-to-send` does. In the
> >> > part that you show, the function will be triggered when one byte is
> >> > available on stdin. If `handle-rdy-to-send` tries to read more bytes
> >> > than are immediately available, then it can block; i.e., the blocking
> >> > part could be in `handle-rdy-to-send`, not in `sync`. That would also
> >> > explain why using `sync/timeout` doesn't change anything.
> >>
> >> Actually, I know it's blocking in the sync (or sync/timeout) because the
> >> actual code is littered with immediately-flushed printf's, and there's one
> >> right before and one right after the sync call. I will try to put together
> >> a version to send you, though. Would you prefer it with the debug printf's
> >> stripped out?
> >>
> >> Thanks, Brett S.
> >>
> >> >
> >> > Again, that's just an example of what might be wrong other than `sync`.
> >> > A complete program would let us sort out the possibilities.
> >> >
> >> > At Sun, 14 Sep 2014 06:57:34 -0500, Brett Stahlman wrote:
> >> > > Hello,
> >> > >
> >> > > I have the following loop in a tcp client program that sends forms
> >> typed by
> >> > > user at the terminal to a listening server, and receives such forms
> >> typed by
> >> > > other clients (broadcast to all clients by the server).
> >> > >
> >> > > (define server-msg-evt (wrap-evt in handle-server-msg))
> >> > > (define rdy-to-send-evt (wrap-evt (standard-input-port)
> >> handle-rdy-to-send))
> >> > >
> >> > > (let loop ()
> >> > >   (sync rdy-to-send-evt server-msg-evt)
> >> > >   (loop))
> >> > >
> >> > > The problem is that the server-msg-evt isn't reliably ready when the
> >> server
> >> > > flushes a message. I know the messages are being sent and flushed
> >> properly
> >> > > by
> >> > > the server because if I type a new form on stdin and hit enter, I see
> >> the
> >> > > following sequence:
> >> > >
> >> > > 1. Client processes new stdin form in handle-rdy-to-send
> >> > > 2. Client loops and calls sync
> >> > > 3. The server-msg-evt is *immediately* ready, and client handles the
> >> > >    message in handle-server-msg.
> >> > > Note: The message handled in step 3 may have been sent long before, but
> >> > > the client remained blocked in sync until the (unrelated) stdin event
> >> > > unblocked it.
> >> > >
> >> > > The really weird part is that the client blocks in this manner even if I
> >> > > change the sync to a sync/timeout!
> >> > >
> >> > > Any clue as to what could be going on here? I'm running both client and
> >> > > server
> >> > > in a DOS box on a Windows 7 machine.
> >> > >
> >> > > Thanks,
> >> > > Brett Stahlman
> >> > > ____________________
> >> > >   Racket Users list:
> >> > >   http://lists.racket-lang.org/users
> >> ____________________
> >>   Racket Users list:
> >>   http://lists.racket-lang.org/users
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users

From brettstahlman at gmail.com  Sun Sep 14 11:03:39 2014
From: brettstahlman at gmail.com (Brett Stahlman)
Date: Sun, 14 Sep 2014 10:03:39 -0500
Subject: [racket] Possible bug involving sync on tcp ports
In-Reply-To: <20140914145002.69ACF6501C4@mail-svr1.cs.utah.edu>
References: <CAOyxqK=ZZtaPWUqM-6M17EfW4LoUEviKqchNxzB+AsfkYjmVGQ@mail.gmail.com>
 <20140914125656.22E386501A8@mail-svr1.cs.utah.edu>
 <CAOyxqKkS==oQu7-UqcRK4Bc-YyT7JA8RM9ST+wf3MYz2m3OatA@mail.gmail.com>
 <20140914142336.7EE6F6501C1@mail-svr1.cs.utah.edu>
 <CAOyxqKk0u_EkZSbpZUm4ikDaeqWUAD=668kk2wAyw39AjjSMEA@mail.gmail.com>
 <20140914145002.69ACF6501C4@mail-svr1.cs.utah.edu>
Message-ID: <CAOyxqKmOT1gWEhTNZ+Xr023oF=5vA3f0gDMahmzOYvEjeVWBPQ@mail.gmail.com>

On Sun, Sep 14, 2014 at 9:49 AM, Matthew Flatt <mflatt at cs.utah.edu> wrote:
> I'll push a repair to make `rnrs/io/ports-6` ports work right with
> `sync`.

Excellent! Thanks.

>
> There's not really a way to get the OS-process-level stdin and stdout
> in Racket, except by using scheme_make_fd_input_port() via
> `ffi/unsafe`. The intent is that the initial ports, which correspond to
> stdin and stdout, can be made inaccessible to untrusted code by
> changing the `current-input-port` and `current-output-port` parameters.
>
> Usually, it's best to just use `(current-input-port)`. The initial
> parameter value corresponds to stdin for a stand-alone Racket process;
> in other cases, a program should normally accept whatever adjustments
> are made by its environment. If you really need the OS-process-level
> stdin and stdout, through, then `ffi/unsafe' can reach them.

Understood. That certainly works for my use-case...

Brett S.

>
> The `standard-input-port` function in Racket's R6RS implementation,
> meanwhile, returns the current input port at the time that the R6RS
> library was loaded. So, it's not necessary stdin in the sense of the OS
> process, but it's "stdin" in the sense of the being the designated
> source for input when the library starts.
>
> At Sun, 14 Sep 2014 09:38:49 -0500, Brett Stahlman wrote:
>> On Sun, Sep 14, 2014 at 9:23 AM, Matthew Flatt <mflatt at cs.utah.edu> wrote:
>> > I think the problem is in `standard-input-port` from `rnrs/io/ports-6`.
>> >
>> > If you use `current-input-port` instead of `standard-input-port`, does
>> > the client behave correctly?
>>
>> It does indeed! Can't recall why I used standard-input-port to begin
>> with, but I'm thinking this could account for some of the weirdness
>> I've been seeing. I didn't mention it in my original posts, but part
>> of the reason I went to such a convoluted strategy for peeking and
>> reading ports was that byte-ready? was blocking as well. Actually,
>> I've been planning to change the client/server architecture entirely -
>> to one in which each port reader operates in a dedicated thread, and
>> simply blocks until a new form can be read, at which point it writes
>> the data to an async channel to be processed at a higher level - but I
>> didn't want to proceed without resolving these issues...
>>
>> Out of curiosity, what is the proper way to set the current input port
>> to stdin (if somehow it's been changed from the default)? (It hasn't,
>> in my case, but perhaps the reason I went to standard-input-port in
>> the first place was that there didn't appear to be any guarantee that
>> the current input port would be stdin...)
>>
>> Thanks,
>> Brett S.
>>
>> >
>> > At Sun, 14 Sep 2014 08:20:26 -0500, Brett Stahlman wrote:
>> >> On Sep 14, 2014 7:56 AM, "Matthew Flatt" <mflatt at cs.utah.edu> wrote:
>> >> >
>> >> > It's possible that something is wrong with `sync` and TCP ports ---
>> >> > although the test suite at least includes that combination --- but can
>> >> > you provide a complete program?
>> >> >
>> >> > For example, I'm interested in what `handle-rdy-to-send` does. In the
>> >> > part that you show, the function will be triggered when one byte is
>> >> > available on stdin. If `handle-rdy-to-send` tries to read more bytes
>> >> > than are immediately available, then it can block; i.e., the blocking
>> >> > part could be in `handle-rdy-to-send`, not in `sync`. That would also
>> >> > explain why using `sync/timeout` doesn't change anything.
>> >>
>> >> Actually, I know it's blocking in the sync (or sync/timeout) because the
>> >> actual code is littered with immediately-flushed printf's, and there's one
>> >> right before and one right after the sync call. I will try to put together
>> >> a version to send you, though. Would you prefer it with the debug printf's
>> >> stripped out?
>> >>
>> >> Thanks, Brett S.
>> >>
>> >> >
>> >> > Again, that's just an example of what might be wrong other than `sync`.
>> >> > A complete program would let us sort out the possibilities.
>> >> >
>> >> > At Sun, 14 Sep 2014 06:57:34 -0500, Brett Stahlman wrote:
>> >> > > Hello,
>> >> > >
>> >> > > I have the following loop in a tcp client program that sends forms
>> >> typed by
>> >> > > user at the terminal to a listening server, and receives such forms
>> >> typed by
>> >> > > other clients (broadcast to all clients by the server).
>> >> > >
>> >> > > (define server-msg-evt (wrap-evt in handle-server-msg))
>> >> > > (define rdy-to-send-evt (wrap-evt (standard-input-port)
>> >> handle-rdy-to-send))
>> >> > >
>> >> > > (let loop ()
>> >> > >   (sync rdy-to-send-evt server-msg-evt)
>> >> > >   (loop))
>> >> > >
>> >> > > The problem is that the server-msg-evt isn't reliably ready when the
>> >> server
>> >> > > flushes a message. I know the messages are being sent and flushed
>> >> properly
>> >> > > by
>> >> > > the server because if I type a new form on stdin and hit enter, I see
>> >> the
>> >> > > following sequence:
>> >> > >
>> >> > > 1. Client processes new stdin form in handle-rdy-to-send
>> >> > > 2. Client loops and calls sync
>> >> > > 3. The server-msg-evt is *immediately* ready, and client handles the
>> >> > >    message in handle-server-msg.
>> >> > > Note: The message handled in step 3 may have been sent long before, but
>> >> > > the client remained blocked in sync until the (unrelated) stdin event
>> >> > > unblocked it.
>> >> > >
>> >> > > The really weird part is that the client blocks in this manner even if I
>> >> > > change the sync to a sync/timeout!
>> >> > >
>> >> > > Any clue as to what could be going on here? I'm running both client and
>> >> > > server
>> >> > > in a DOS box on a Windows 7 machine.
>> >> > >
>> >> > > Thanks,
>> >> > > Brett Stahlman
>> >> > > ____________________
>> >> > >   Racket Users list:
>> >> > >   http://lists.racket-lang.org/users
>> >> ____________________
>> >>   Racket Users list:
>> >>   http://lists.racket-lang.org/users
>> ____________________
>>   Racket Users list:
>>   http://lists.racket-lang.org/users

From oev-racket at sibmail.com  Mon Sep 15 00:07:52 2014
From: oev-racket at sibmail.com (Evgeny Odegov)
Date: Mon, 15 Sep 2014 11:07:52 +0700 (NOVT)
Subject: [racket] Some problem with hash-update!
In-Reply-To: <20140914124010.BD2A765019F@mail-svr1.cs.utah.edu>
References: <52456.79.136.149.232.1410618490.squirrel@www.sibmail.com>
 <52709.79.136.149.232.1410623128.squirrel@www.sibmail.com>
 <20140914124010.BD2A765019F@mail-svr1.cs.utah.edu>
Message-ID: <58297.91.221.60.217.1410754072.squirrel@www.sibmail.com>

More complete example:
http://pasterack.org/pastes/25901

I guess I've found reason of the problem.

I think the problem was in that line:
(define lpdwItems (make-bytes (ctype-sizeof _pointer) 0))

I've changed it to:
(define lpdwItems (make-bytes (ctype-sizeof _DWORD) 0))

There is no visible problems now, but I doubt that applying `make-bytes'
here is right.

Should I use `malloc' in such cases?


Anyway, sorry for the noise.


> My best guess is that it's an issue with memory management, because
> memory management gets a lot trickier when you call a function that can
> call back into Racket. A GC can happen during the callback, and
> therefore during the dynamic extent of the foreign-function call, which
> can move arguments that were passed to the foreign function, for example.
>
> Would it be possible to provide a complete example?
>
> At Sat, 13 Sep 2014 22:45:28 +0700 (NOVT), "Evgeny Odegov" wrote:
>> Oh, I'm wrong. The last code is not equivalent to initial.
>> The equivalent code has this problem:
>>
>> (define result (make-hash))
>> (define big-list (list))
>> (define (callback dwTextID lpszInfoText lpvUser)
>>   (if (equal? lpszInfoText "")
>>       (set! big-list (append big-list (list dwTextID)))
>>       (hash-update! result
>>                     lpszInfoText
>>                     (lambda (id-lst)
>>                       (define v (list dwTextID))
>>                       (append id-lst v))
>>                     '()))
>>   #t)
>>
>>
>> The code with `cons' instead `append' works fine:
>>
>> (define result (make-hash))
>> (define (callback dwTextID lpszInfoText lpvUser)
>>   (hash-update! result
>>                 lpszInfoText
>>                 (lambda (id-lst)
>>                   (define new-id-lst (cons dwTextID id-lst))
>>                   (unless (list? new-id-lst)
>>
>>                     (error "~v" new-id-lst))
>>                   new-id-lst)
>>                 '())
>>   #t)
>>
>>
>>
>>
>> ____________________
>>   Racket Users list:
>>   http://lists.racket-lang.org/users
>



From mflatt at cs.utah.edu  Mon Sep 15 01:58:08 2014
From: mflatt at cs.utah.edu (Matthew Flatt)
Date: Mon, 15 Sep 2014 07:58:08 +0200
Subject: [racket] Some problem with hash-update!
In-Reply-To: <58297.91.221.60.217.1410754072.squirrel@www.sibmail.com>
References: <52456.79.136.149.232.1410618490.squirrel@www.sibmail.com>
 <52709.79.136.149.232.1410623128.squirrel@www.sibmail.com>
 <20140914124010.BD2A765019F@mail-svr1.cs.utah.edu>
 <58297.91.221.60.217.1410754072.squirrel@www.sibmail.com>
Message-ID: <20140915055811.D05D0650188@mail-svr1.cs.utah.edu>

I think that changing `_pointer` to `_DWORD` can't be the real answer,
since `_pointer` is at least as large as `_DWORD`.

But I agree that you should use `malloc`, and you should also supply
'atomic-interior:

 (malloc _DWORD 'atomic-interior)

The 'atomic-interior flag ensures that the allocated memory does not
move during an invocation of the callback.

Along similar lines, I think you need to add

 #:malloc-mode 'atomic-interior

to the declaration of `_CMN_ERROR`.

You may need to copy the `filter` argument into non-moving memory, but
that's probably not an issue if it's always "".

At Mon, 15 Sep 2014 11:07:52 +0700 (NOVT), "Evgeny Odegov" wrote:
> More complete example:
> http://pasterack.org/pastes/25901
> 
> I guess I've found reason of the problem.
> 
> I think the problem was in that line:
> (define lpdwItems (make-bytes (ctype-sizeof _pointer) 0))
> 
> I've changed it to:
> (define lpdwItems (make-bytes (ctype-sizeof _DWORD) 0))
> 
> There is no visible problems now, but I doubt that applying `make-bytes'
> here is right.
> 
> Should I use `malloc' in such cases?
> 
> 
> Anyway, sorry for the noise.
> 
> 
> > My best guess is that it's an issue with memory management, because
> > memory management gets a lot trickier when you call a function that can
> > call back into Racket. A GC can happen during the callback, and
> > therefore during the dynamic extent of the foreign-function call, which
> > can move arguments that were passed to the foreign function, for example.
> >
> > Would it be possible to provide a complete example?
> >
> > At Sat, 13 Sep 2014 22:45:28 +0700 (NOVT), "Evgeny Odegov" wrote:
> >> Oh, I'm wrong. The last code is not equivalent to initial.
> >> The equivalent code has this problem:
> >>
> >> (define result (make-hash))
> >> (define big-list (list))
> >> (define (callback dwTextID lpszInfoText lpvUser)
> >>   (if (equal? lpszInfoText "")
> >>       (set! big-list (append big-list (list dwTextID)))
> >>       (hash-update! result
> >>                     lpszInfoText
> >>                     (lambda (id-lst)
> >>                       (define v (list dwTextID))
> >>                       (append id-lst v))
> >>                     '()))
> >>   #t)
> >>
> >>
> >> The code with `cons' instead `append' works fine:
> >>
> >> (define result (make-hash))
> >> (define (callback dwTextID lpszInfoText lpvUser)
> >>   (hash-update! result
> >>                 lpszInfoText
> >>                 (lambda (id-lst)
> >>                   (define new-id-lst (cons dwTextID id-lst))
> >>                   (unless (list? new-id-lst)
> >>
> >>                     (error "~v" new-id-lst))
> >>                   new-id-lst)
> >>                 '())
> >>   #t)
> >>
> >>
> >>
> >>
> >> ____________________
> >>   Racket Users list:
> >>   http://lists.racket-lang.org/users
> >

From oev-racket at sibmail.com  Mon Sep 15 04:04:12 2014
From: oev-racket at sibmail.com (Evgeny Odegov)
Date: Mon, 15 Sep 2014 15:04:12 +0700 (NOVT)
Subject: [racket] Some problem with hash-update!
In-Reply-To: <20140915055811.D05D0650188@mail-svr1.cs.utah.edu>
References: <52456.79.136.149.232.1410618490.squirrel@www.sibmail.com>   
 <52709.79.136.149.232.1410623128.squirrel@www.sibmail.com>   
 <20140914124010.BD2A765019F@mail-svr1.cs.utah.edu>
 <58297.91.221.60.217.1410754072.squirrel@www.sibmail.com>
 <20140915055811.D05D0650188@mail-svr1.cs.utah.edu>
Message-ID: <62737.91.221.60.217.1410768252.squirrel@www.sibmail.com>

> I think that changing `_pointer` to `_DWORD` can't be the real answer,
> since `_pointer` is at least as large as `_DWORD`.

Facepalm. Of course, you're right! Somehow, I've thought that _pointer
 is twice shorter :)

> But I agree that you should use `malloc`, and you should also supply
> 'atomic-interior:
>
>  (malloc _DWORD 'atomic-interior)
>
> The 'atomic-interior flag ensures that the allocated memory does not
> move during an invocation of the callback.
>

It seems like I've got a wrong impression about #:atomic? parameter of
`_fun` form. I thought it's intended to prevent such issues.


> Along similar lines, I think you need to add
>
>  #:malloc-mode 'atomic-interior
>
> to the declaration of `_CMN_ERROR`.
>

Thank you for the notes!  So, as I see, `malloc' with 'atomic-inferior mode
must be used, when a parameter is changed by foreign function.


> You may need to copy the `filter` argument into non-moving memory, but
> that's probably not an issue if it's always "".

The parameter `lpszFilter` is not changed by `TXTEnumInfoText`.
Is `_string/locale` enough in such case?

What could you say about `_string/locale/len'?
Is there something that I have not took into account?


>
> At Mon, 15 Sep 2014 11:07:52 +0700 (NOVT), "Evgeny Odegov" wrote:
>> More complete example:
>> http://pasterack.org/pastes/25901
>>
>> I guess I've found reason of the problem.
>>
>> I think the problem was in that line:
>> (define lpdwItems (make-bytes (ctype-sizeof _pointer) 0))
>>
>> I've changed it to:
>> (define lpdwItems (make-bytes (ctype-sizeof _DWORD) 0))
>>
>> There is no visible problems now, but I doubt that applying `make-bytes'
>> here is right.
>>
>> Should I use `malloc' in such cases?
>>
>>
>> Anyway, sorry for the noise.
>>
>>
>> > My best guess is that it's an issue with memory management, because
>> > memory management gets a lot trickier when you call a function that
>> can
>> > call back into Racket. A GC can happen during the callback, and
>> > therefore during the dynamic extent of the foreign-function call,
>> which
>> > can move arguments that were passed to the foreign function, for
>> example.
>> >
>> > Would it be possible to provide a complete example?
>> >
>> > At Sat, 13 Sep 2014 22:45:28 +0700 (NOVT), "Evgeny Odegov" wrote:
>> >> Oh, I'm wrong. The last code is not equivalent to initial.
>> >> The equivalent code has this problem:
>> >>
>> >> (define result (make-hash))
>> >> (define big-list (list))
>> >> (define (callback dwTextID lpszInfoText lpvUser)
>> >>   (if (equal? lpszInfoText "")
>> >>       (set! big-list (append big-list (list dwTextID)))
>> >>       (hash-update! result
>> >>                     lpszInfoText
>> >>                     (lambda (id-lst)
>> >>                       (define v (list dwTextID))
>> >>                       (append id-lst v))
>> >>                     '()))
>> >>   #t)
>> >>
>> >>
>> >> The code with `cons' instead `append' works fine:
>> >>
>> >> (define result (make-hash))
>> >> (define (callback dwTextID lpszInfoText lpvUser)
>> >>   (hash-update! result
>> >>                 lpszInfoText
>> >>                 (lambda (id-lst)
>> >>                   (define new-id-lst (cons dwTextID id-lst))
>> >>                   (unless (list? new-id-lst)
>> >>
>> >>                     (error "~v" new-id-lst))
>> >>                   new-id-lst)
>> >>                 '())
>> >>   #t)
>> >>
>> >>
>> >>
>> >>
>> >> ____________________
>> >>   Racket Users list:
>> >>   http://lists.racket-lang.org/users
>> >
>



From dbastos at toledo.com  Mon Sep 15 09:20:30 2014
From: dbastos at toledo.com (Daniel Bastos)
Date: Mon, 15 Sep 2014 10:20:30 -0300
Subject: [racket] missing solution 20.1.1 ex:sem-funcs
In-Reply-To: <C238EB4D-A9B9-466D-AE4A-13F3EBF597EA@ccs.neu.edu>
References: <CAB08-qr3dsphx=M875-g8EBNkrdFN3-BLkvTQnytrZ4i-5uUvg@mail.gmail.com>
 <DB47056E-CF26-45EA-97B5-7935F08A7AB2@ccs.neu.edu>
 <CAB08-qomWeikoWpv_=Y4S7i75=hMP2RybqX0F914F5b3OqzEcg@mail.gmail.com>
 <C238EB4D-A9B9-466D-AE4A-13F3EBF597EA@ccs.neu.edu>
Message-ID: <CAB08-qoHtRDn65t7OMVbowHtJafMfegBDTGrKjjoSLzeGPOVYQ@mail.gmail.com>

On Fri, Sep 12, 2014 at 6:43 PM,
Matthias Felleisen <matthias at ccs.neu.edu> wrote:


> On Sep 12, 2014, at 4:40 PM, Daniel Bastos <dbastos at toledo.com> wrote:
>
> > Again, we start with (2) and apply a series of substitutions.
> >
> >     (f f)
> >   = (<var> <var>) ;; since f is a <var>
> >   = (<exp> <exp>) ;; since <var> is a subset of <exp>
> >
> > However, (<exp> <exp>) is not by definition a value. It is an
> > expression which may or may not /have/ a value. We know the
> > definition of f, however, which tells us that it behaves like
> > the identity function. So the evaluation of that (<exp> <exp>)
> > yields a function, which is a value. We end with
> >
> >   = <val>
> >
> > Therefore (2) is a value
>
> Is there a difference between the statements
>
>  "Daniel is a hat."
>
> and
>
>  "Daniel has a hat."
>
> or do they really mean the same thing?
>

There's a difference.


> In this spirit, is there a difference between the statement
>
>  "(f f) is a value."
>
> and
>
>  "(f f) has a value."
>
> or do they really mean the same thing?
>

There's a difference. The analogy is good. (I do remember having read this
analogy in the book, though I can't seem to find it right now.)

Here's my final rewrite. Feel free to scrutinize it. (BTW, thanks for your
attention.)

Exercise 20.1.1. Assume the Definitions window in DrScheme contains
(define (f x) x). Identify the values among the following expressions:

 (1) (cons f empty)
 (2) (f f)
 (3) (cons f (cons 10 (cons (f 10) empty)))

Explain why they are values and why the remaining expressions are not
values.

Solution. A value is anything of the form

  <val> = <boo> | <sym> | <num> | empty | <lst>
  | <var> (names of defined functions)
  | <prm>

where <lst> is of the form

  <lst> = empty | (cons <val> <lst>).

Now we start with (1) and apply a series of substitutions based on the
definitions we have.

    (cons f empty)
  = (cons <var> <lst>)  ;; since f is a <var>
  = (cons <var> <lst>)  ;; since empty is a <lst>
  = (cons <val> <lst>)  ;; since <var> is a subset of <val>
  = <lst>               ;; definition of <lst>
  = <val>               ;; definition of <val>

Therefore (1) is a value.

Next we consider (2). Here's a relevant part of the grammar for (2).

  <exp> = <var>
  | <prm>
  | (<exp> <exp> ...<exp>)

Start with (2)...

    (f f)
  = (<var> <var>) ;; since f is a <var>
  = (<exp> <exp>) ;; since <var> is a subset of <exp>

Now (<exp> <exp>) is not by definition a value. Expressions may have
values, but they're not values. (Daniel may have a hat, but Daniel is
not a hat.)

Therefore (2) is not a value.

Now (3).

    (cons f (cons 10 (cons (f 10) empty)))
  = (cons <var> (cons <num> (cons (<var> <num>) <lst>)))

The whole expression depends on the application (<exp> <exp>),
which may or may not have a value. Since we know f, we know it has a value,
but it /is not/ a value.

Therefore (3) is not a value.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140915/ab02c125/attachment-0001.html>

From dbastos at toledo.com  Mon Sep 15 09:26:28 2014
From: dbastos at toledo.com (Daniel Bastos)
Date: Mon, 15 Sep 2014 10:26:28 -0300
Subject: [racket] missing solution 20.1.2 ex:syn-funcs
In-Reply-To: <7E897494-BFBC-499C-B189-AFB890468E62@ccs.neu.edu>
References: <CAB08-qqmrq3jesw_GzLbmR0K6PA8WBbtscezT4TGbooL373hZA@mail.gmail.com>
 <7E897494-BFBC-499C-B189-AFB890468E62@ccs.neu.edu>
Message-ID: <CAB08-qpU2gBdMYZk0dS3_=4r3iwgg5-ett4DN2gvCNm3eDXvPA@mail.gmail.com>

On Mon, Sep 8, 2014 at 9:51 PM,
Matthias Felleisen <matthias at ccs.neu.edu> wrote:

On Sep 2, 2014, at 12:05 PM, Daniel Bastos wrote:
>
> > Exercise 20.1.2. Argue why the following sentences are legal
> > definitions:
> >
> > (define (f x) (x 10))
> >
> > (define (f x) f)
> >
> > (define (f x y) (x 'a y 'b))
> >
> > Solution. The relevant part of the grammar is the following.
> >
> >  <def> = (define (<var> <var> ...<var>) <exp>)
> >        | (define <var> <exp>)
> >        | (define-struct <var> (<var> <var> ...<var>))
> >
> > (*) First definition
> >
> > The LHS is a list of <var>, since we find f and x as members of the
> > list and they're both variables. The RHS is a list consisting of a
> > <var> and a <num>.
>
> Small correction. Let's not call (x 10) a list. It's an application. '(x
> 10) would be a list and this one character is critical.
>

Indeed. Here's a full rewrite for completeness. (Thanks!)

Exercise 20.1.2. Argue why the following sentences are legal
definitions:

(define (f x) (x 10))

(define (f x) f)

(define (f x y) (x 'a y 'b))

Solution. The relevant part of the grammar is the following.

  <def> = (define (<var> <var> ...<var>) <exp>)
        | (define <var> <exp>)
        | (define-struct <var> (<var> <var> ...<var>))

(*) First definition

The LHS is a list of <var>, since we find f and x as members of the
list and they're both variables. The RHS is an application of the form
(<var> <num>), which makes up an <exp>, satisfying the first form
of <def>. In other words,

  (define (<var> <var>) (<var> <num)) =
  (define (<var> <var>) (<exp> <exp>) =
  (define (<var> <var>) <exp>).

Therefore it's a legal <def>.

(*) Second definition

Same LHS as the previous, so we need only check the RHS which is a
<var>. <var> is a valid form of <exp>, so we have

  (define (<var> <var>) <var>) =
  (define (<var> <var>) <exp>).

Therefore it's a legal <def>.

(*) Third definition

The LHS is (<var> <var> <var>), while the RHS is (<var> <sym> <var>
<sym>), so we have

  (define (<var> <var> <var>) (<var> <sym> <var> <sym>)) =
  (define (<var> <var> <var>) (<exp> <exp> <exp> <exp>)) =
  (define (<var> <var> <var>) <exp>).

Therefore it's a legal <def>.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140915/6d6bc76e/attachment.html>

From matthias at ccs.neu.edu  Mon Sep 15 09:49:18 2014
From: matthias at ccs.neu.edu (Matthias Felleisen)
Date: Mon, 15 Sep 2014 09:49:18 -0400
Subject: [racket] missing solution 20.1.1 ex:sem-funcs
In-Reply-To: <CAB08-qoHtRDn65t7OMVbowHtJafMfegBDTGrKjjoSLzeGPOVYQ@mail.gmail.com>
References: <CAB08-qr3dsphx=M875-g8EBNkrdFN3-BLkvTQnytrZ4i-5uUvg@mail.gmail.com>
 <DB47056E-CF26-45EA-97B5-7935F08A7AB2@ccs.neu.edu>
 <CAB08-qomWeikoWpv_=Y4S7i75=hMP2RybqX0F914F5b3OqzEcg@mail.gmail.com>
 <C238EB4D-A9B9-466D-AE4A-13F3EBF597EA@ccs.neu.edu>
 <CAB08-qoHtRDn65t7OMVbowHtJafMfegBDTGrKjjoSLzeGPOVYQ@mail.gmail.com>
Message-ID: <BC5F3276-ABEF-492A-85D3-8EAAF3DA58C7@ccs.neu.edu>


Good. 


On Sep 15, 2014, at 9:20 AM, Daniel Bastos <dbastos at toledo.com> wrote:

> On Fri, Sep 12, 2014 at 6:43 PM, 
> Matthias Felleisen <matthias at ccs.neu.edu> wrote:
> 
> 
> On Sep 12, 2014, at 4:40 PM, Daniel Bastos <dbastos at toledo.com> wrote:
> 
> > Again, we start with (2) and apply a series of substitutions.
> >
> >     (f f)
> >   = (<var> <var>) ;; since f is a <var>
> >   = (<exp> <exp>) ;; since <var> is a subset of <exp>
> >
> > However, (<exp> <exp>) is not by definition a value. It is an
> > expression which may or may not /have/ a value. We know the
> > definition of f, however, which tells us that it behaves like
> > the identity function. So the evaluation of that (<exp> <exp>)
> > yields a function, which is a value. We end with
> >
> >   = <val>
> >
> > Therefore (2) is a value
> 
> Is there a difference between the statements
> 
>  "Daniel is a hat."
> 
> and
> 
>  "Daniel has a hat."
> 
> or do they really mean the same thing?
> 
> There's a difference.
>  
> In this spirit, is there a difference between the statement
> 
>  "(f f) is a value."
> 
> and
> 
>  "(f f) has a value."
> 
> or do they really mean the same thing?
> 
> There's a difference. The analogy is good. (I do remember having read this analogy in the book, though I can't seem to find it right now.)
> 
> Here's my final rewrite. Feel free to scrutinize it. (BTW, thanks for your attention.)
> 
> Exercise 20.1.1. Assume the Definitions window in DrScheme contains
> (define (f x) x). Identify the values among the following expressions:
> 
>  (1) (cons f empty)
>  (2) (f f)
>  (3) (cons f (cons 10 (cons (f 10) empty)))
> 
> Explain why they are values and why the remaining expressions are not
> values.
> 
> Solution. A value is anything of the form
> 
>   <val> = <boo> | <sym> | <num> | empty | <lst>
>   | <var> (names of defined functions)
>   | <prm>
> 
> where <lst> is of the form
> 
>   <lst> = empty | (cons <val> <lst>).
> 
> Now we start with (1) and apply a series of substitutions based on the
> definitions we have.
> 
>     (cons f empty)
>   = (cons <var> <lst>)  ;; since f is a <var>
>   = (cons <var> <lst>)  ;; since empty is a <lst>
>   = (cons <val> <lst>)  ;; since <var> is a subset of <val>
>   = <lst>               ;; definition of <lst>
>   = <val>               ;; definition of <val>
> 
> Therefore (1) is a value.
> 
> Next we consider (2). Here's a relevant part of the grammar for (2).
> 
>   <exp> = <var>
>   | <prm>
>   | (<exp> <exp> ...<exp>)
> 
> Start with (2)...
> 
>     (f f)
>   = (<var> <var>) ;; since f is a <var>
>   = (<exp> <exp>) ;; since <var> is a subset of <exp>
> 
> Now (<exp> <exp>) is not by definition a value. Expressions may have
> values, but they're not values. (Daniel may have a hat, but Daniel is
> not a hat.)
> 
> Therefore (2) is not a value.
> 
> Now (3).
> 
>     (cons f (cons 10 (cons (f 10) empty)))
>   = (cons <var> (cons <num> (cons (<var> <num>) <lst>)))
> 
> The whole expression depends on the application (<exp> <exp>),
> which may or may not have a value. Since we know f, we know it has a value,
> but it /is not/ a value.
> 
> Therefore (3) is not a value.



From rudolphg1 at citadel.edu  Mon Sep 15 13:50:10 2014
From: rudolphg1 at citadel.edu (George Rudolph)
Date: Mon, 15 Sep 2014 17:50:10 +0000
Subject: [racket] How best to repeat a function call n times?
Message-ID: <60998febf5a54dfab298d3604b0d415b@BLUPR01MB129.prod.exchangelabs.com>


All,
I am sure this question has been answered before, but I can't find it.
Suppose I want to call some function, either built-in or user-defined, n number of times, and measure the total time it takes. Is there a more efficient or elegant way
of doing this than passing the expression to a tail-recursive counting loop such as the following?

(define (repeater sexpr count)
  (sexpr)
  (cond
    [(> count 0) (repeater sexpr (- count 1)) ])
)



George Rudolph
Associate Professor of Computer Science
Department of Mathematics and Computer Science
225 Thompson Hall
The Citadel
171 Moultrie Street
Charleston, SC 29409
843.953.5032

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140915/938e5bc6/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image001.gif
Type: image/gif
Size: 92 bytes
Desc: image001.gif
URL: <http://lists.racket-lang.org/users/archive/attachments/20140915/938e5bc6/attachment.gif>

From robby at eecs.northwestern.edu  Mon Sep 15 13:53:26 2014
From: robby at eecs.northwestern.edu (Robby Findler)
Date: Mon, 15 Sep 2014 12:53:26 -0500
Subject: [racket] How best to repeat a function call n times?
In-Reply-To: <60998febf5a54dfab298d3604b0d415b@BLUPR01MB129.prod.exchangelabs.com>
References: <60998febf5a54dfab298d3604b0d415b@BLUPR01MB129.prod.exchangelabs.com>
Message-ID: <CAL3TdOOLFK2ViBB3D-VABP74hmJaiGvm8xu49q6dEHdjHy=h2Q@mail.gmail.com>

Not more efficient, but perhaps more readable:

(define (repeater f count)
  (for ([x (in-range count)])
    (f)))


On Mon, Sep 15, 2014 at 12:50 PM, George Rudolph <rudolphg1 at citadel.edu>
wrote:

>   All,
>
> I am sure this question has been answered before, but I can?t find it.
>
> Suppose I want to call some function, either built-in or user-defined, n
> number of times, and measure the total time it takes. Is there a more
> efficient or elegant way
>
> of doing this than passing the expression to a tail-recursive counting
> loop such as the following?
>
>
>
> (define (repeater sexpr count)
>
>   (sexpr)
>
>   (cond
>
>     [(> count 0) (repeater sexpr (- count 1)) ])
>
> )
>
>
>
>
>
>
>
> George Rudolph
>
> Associate Professor of Computer Science
>
> Department of Mathematics and Computer Science
>
> 225 Thompson Hall
>
> The Citadel
>
> 171 Moultrie Street
>
> Charleston, SC 29409
>
> 843.953.5032
>
>
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140915/9adf9a49/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image001.gif
Type: image/gif
Size: 92 bytes
Desc: not available
URL: <http://lists.racket-lang.org/users/archive/attachments/20140915/9adf9a49/attachment.gif>

From matthias at ccs.neu.edu  Mon Sep 15 13:57:04 2014
From: matthias at ccs.neu.edu (Matthias Felleisen)
Date: Mon, 15 Sep 2014 13:57:04 -0400
Subject: [racket] How best to repeat a function call n times?
In-Reply-To: <60998febf5a54dfab298d3604b0d415b@BLUPR01MB129.prod.exchangelabs.com>
References: <60998febf5a54dfab298d3604b0d415b@BLUPR01MB129.prod.exchangelabs.com>
Message-ID: <FEB45042-D5B2-4DBB-9A17-0E9AE0DD8C94@ccs.neu.edu>



Welcome to Racket v6.1.0.8.
> (define (repeater f count) (for ((i (in-range count))) (f)))
> (repeater (lambda () (displayln "hello world")) 10)
hello world
hello world
hello world
hello world
hello world
hello world
hello world
hello world
hello world
hello world



On Sep 15, 2014, at 1:50 PM, George Rudolph <rudolphg1 at citadel.edu> wrote:

> All,
> I am sure this question has been answered before, but I can?t find it.
> Suppose I want to call some function, either built-in or user-defined, n number of times, and measure the total time it takes. Is there a more efficient or elegant way
> of doing this than passing the expression to a tail-recursive counting loop such as the following?
>  
> (define (repeater sexpr count)
>   (sexpr)
>   (cond
>     [(> count 0) (repeater sexpr (- count 1)) ])
> ) 
>  
>  
>  
> George Rudolph
> Associate Professor of Computer Science
> Department of Mathematics and Computer Science
> 225 Thompson Hall
> The Citadel
> 171 Moultrie Street
> Charleston, SC 29409
> 843.953.5032
>  
> ____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users



From mb at mbtype.com  Mon Sep 15 16:38:00 2014
From: mb at mbtype.com (Matthew Butterick)
Date: Mon, 15 Sep 2014 13:38:00 -0700
Subject: [racket] Can a #reader be used with (module ...) ?
Message-ID: <54174E28.4000001@mbtype.com>

I'm aware that `#lang name ...` is equivalent to `(module id name ...)`. 
But `#lang` automatically picks up a reader; `module` does not.

Is there a way to  attach a reader to the `module` form (the idea being 
that then, within the `module` expression, you could write using the 
native syntax of the #lang)?

Studying the example in reader extensions (Guide ? 17.2), I found that 
this works, but I haven't figured out if it can be generalized:

#lang racket/base
(module inner racket/base
   '(#reader"five.rkt"23456))
(require 'inner)

For instance, my current reader works in the REPL, like so:

'#reader(submod pollen/main reader)Hello World

But this doesn't work, because the reader, which operates in text mode, 
absorbs the closing parenthesis:

#lang racket/base
(module inner racket/base
   '#reader(submod pollen/main reader)Hello World
   )

From dbastos at toledo.com  Mon Sep 15 16:48:08 2014
From: dbastos at toledo.com (Daniel Bastos)
Date: Mon, 15 Sep 2014 17:48:08 -0300
Subject: [racket] on reversing a list by way of sort
Message-ID: <CAB08-qqQyfysyAr_uzcQ4rajfGJSVOj_06T8SC5pfXhA2apSNQ@mail.gmail.com>

Dear Racketeers, I was studying the exercise 20.2.4 of HtDP when I came up
with this way of reversing lists. (Every element is a least element. Or
greatest.)

(define (f x y)
  true)

(define (rev ls)
  (sort ls f))

Welcome to DrRacket, version 6.0.1 [3m].
Language: Intermediate Student; memory limit: 256 MB.
> (rev (list 1 3 4))
(list 4 3 1)
> (rev '(a b c d))
(list 'd 'c 'b 'a)
>

Now I'm feeling like an expert. (But my friends still call me Danewbie.)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140915/82a84e1d/attachment.html>

From dvanhorn at cs.umd.edu  Mon Sep 15 16:53:50 2014
From: dvanhorn at cs.umd.edu (David Van Horn)
Date: Mon, 15 Sep 2014 16:53:50 -0400
Subject: [racket] on reversing a list by way of sort
In-Reply-To: <CAB08-qqQyfysyAr_uzcQ4rajfGJSVOj_06T8SC5pfXhA2apSNQ@mail.gmail.com>
References: <CAB08-qqQyfysyAr_uzcQ4rajfGJSVOj_06T8SC5pfXhA2apSNQ@mail.gmail.com>
Message-ID: <541751DE.5050407@cs.umd.edu>

I don't think you made enough examples.

On 9/15/14, 4:48 PM, Daniel Bastos wrote:
> Dear Racketeers, I was studying the exercise 20.2.4 of HtDP when I
> came up with this way of reversing lists. (Every element is a least
> element. Or greatest.)
> 
> (define (f x y) true)
> 
> (define (rev ls) (sort ls f))
> 
> Welcome to DrRacket, version 6.0.1 [3m]. Language: Intermediate
> Student; memory limit: 256 MB.
>> (rev (list 1 3 4))
> (list 4 3 1)
>> (rev '(a b c d))
> (list 'd 'c 'b 'a)
>> 
> 
> Now I'm feeling like an expert. (But my friends still call me
> Danewbie.)
> 
> 
> ____________________ Racket Users list: 
> http://lists.racket-lang.org/users
> 


From dvanhorn at cs.umd.edu  Mon Sep 15 17:05:01 2014
From: dvanhorn at cs.umd.edu (David Van Horn)
Date: Mon, 15 Sep 2014 17:05:01 -0400
Subject: [racket] on reversing a list by way of sort
In-Reply-To: <541751DE.5050407@cs.umd.edu>
References: <CAB08-qqQyfysyAr_uzcQ4rajfGJSVOj_06T8SC5pfXhA2apSNQ@mail.gmail.com>
 <541751DE.5050407@cs.umd.edu>
Message-ID: <5417547D.3060309@cs.umd.edu>

On 9/15/14, 4:53 PM, David Van Horn wrote:
> I don't think you made enough examples.

Nope - my bad.  Cute.  Awful, but cute (and there could be correct
implementations of sort that would break your rev).

David



From robby at eecs.northwestern.edu  Mon Sep 15 17:16:02 2014
From: robby at eecs.northwestern.edu (Robby Findler)
Date: Mon, 15 Sep 2014 16:16:02 -0500
Subject: [racket] on reversing a list by way of sort
In-Reply-To: <5417547D.3060309@cs.umd.edu>
References: <CAB08-qqQyfysyAr_uzcQ4rajfGJSVOj_06T8SC5pfXhA2apSNQ@mail.gmail.com>
 <541751DE.5050407@cs.umd.edu> <5417547D.3060309@cs.umd.edu>
Message-ID: <CAL3TdON=6kCKdkqGbvRq-be5K9ewKOvMg1tWsUk79N5UGt971g@mail.gmail.com>

Fritz Henglein has studied the problem "what is a sorting function?" that
touches on this kind of thing.

http://www.sciencedirect.com/science/article/pii/S1567832608001094


Robby

On Monday, September 15, 2014, David Van Horn <dvanhorn at cs.umd.edu> wrote:

> On 9/15/14, 4:53 PM, David Van Horn wrote:
> > I don't think you made enough examples.
>
> Nope - my bad.  Cute.  Awful, but cute (and there could be correct
> implementations of sort that would break your rev).
>
> David
>
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140915/4c7626b2/attachment.html>

From mb at mbtype.com  Mon Sep 15 21:57:51 2014
From: mb at mbtype.com (Matthew Butterick)
Date: Mon, 15 Sep 2014 18:57:51 -0700
Subject: [racket] on reversing a list by way of sort
In-Reply-To: <CAL3TdON=6kCKdkqGbvRq-be5K9ewKOvMg1tWsUk79N5UGt971g@mail.gmail.com>
References: <CAB08-qqQyfysyAr_uzcQ4rajfGJSVOj_06T8SC5pfXhA2apSNQ@mail.gmail.com>
 <541751DE.5050407@cs.umd.edu> <5417547D.3060309@cs.umd.edu>
 <CAL3TdON=6kCKdkqGbvRq-be5K9ewKOvMg1tWsUk79N5UGt971g@mail.gmail.com>
Message-ID: <5417991F.8020807@mbtype.com>

Mike Bostock's visual demonstration of why naive sorting functions are 
false friends. As he and Henglein point out, transitivity of the 
comparator is essential.

http://bost.ocks.org/mike/shuffle/compare.html

On 09/15/14 2:16 PM, Robby Findler wrote:
> Fritz Henglein has studied the problem "what is a sorting function?" 
> that touches on this kind of thing.
>
> http://www.sciencedirect.com/science/article/pii/S1567832608001094
>
>
> Robby
>
> On Monday, September 15, 2014, David Van Horn <dvanhorn at cs.umd.edu 
> <mailto:dvanhorn at cs.umd.edu>> wrote:
>
>     On 9/15/14, 4:53 PM, David Van Horn wrote:
>     > I don't think you made enough examples.
>
>     Nope - my bad.  Cute.  Awful, but cute (and there could be correct
>     implementations of sort that would break your rev).
>
>     David
>
>
>     ____________________
>       Racket Users list:
>     http://lists.racket-lang.org/users
>
>
>
> ____________________
>    Racket Users list:
>    http://lists.racket-lang.org/users


From oev-racket at sibmail.com  Mon Sep 15 23:43:24 2014
From: oev-racket at sibmail.com (Evgeny Odegov)
Date: Tue, 16 Sep 2014 10:43:24 +0700 (NOVT)
Subject: [racket] Some problem with hash-update!
In-Reply-To: <20140915153217.CF4C165018C@mail-svr1.cs.utah.edu>
References: <52456.79.136.149.232.1410618490.squirrel@www.sibmail.com>   
 <52709.79.136.149.232.1410623128.squirrel@www.sibmail.com>   
 <20140914124010.BD2A765019F@mail-svr1.cs.utah.edu>
 <58297.91.221.60.217.1410754072.squirrel@www.sibmail.com>
 <20140915055811.D05D0650188@mail-svr1.cs.utah.edu>
 <62737.91.221.60.217.1410768252.squirrel@www.sibmail.com>
 <20140915153217.CF4C165018C@mail-svr1.cs.utah.edu>
Message-ID: <52591.91.221.60.217.1410839004.squirrel@www.sibmail.com>

> Using `#:atomic? #t` disables Racket thread switches, but it doesn't
> disable GC.

> It's not so much about modification as referencing the memory ---
> either reading or writing --- across a potential GC.

Thank you! It's clear now.

> Now that I look closely, I think you need the `bytes-copy!` code that
> you have commented out. Otherwise, initializing a array will read past
> the end of the allocated byte string.
>
> For example,
>
>  (define-cstruct _foo ([a _int]
>                        [b (_string/locale/len 10000000)]))
>
>  (make-foo 0 "boo")
>
> crashes for me, because 10000000 is big enough that it reads past
> allocated regions. That example works if I uncomment your `bytes-copy!`
> use, which makes it return a byte string that is 10000000 bytes long.

Good example. I commented those lines to experiment and forgot about them.

Thank you for your help!


From asumu at ccs.neu.edu  Mon Sep 15 23:50:49 2014
From: asumu at ccs.neu.edu (Asumu Takikawa)
Date: Mon, 15 Sep 2014 23:50:49 -0400
Subject: [racket] on reversing a list by way of sort
In-Reply-To: <5417991F.8020807@mbtype.com>
References: <CAB08-qqQyfysyAr_uzcQ4rajfGJSVOj_06T8SC5pfXhA2apSNQ@mail.gmail.com>
 <541751DE.5050407@cs.umd.edu> <5417547D.3060309@cs.umd.edu>
 <CAL3TdON=6kCKdkqGbvRq-be5K9ewKOvMg1tWsUk79N5UGt971g@mail.gmail.com>
 <5417991F.8020807@mbtype.com>
Message-ID: <20140916035049.GS21882@localhost>

On 2014-09-15 18:57:51 -0700, Matthew Butterick wrote:
> Mike Bostock's visual demonstration of why naive sorting functions are false
> friends. As he and Henglein point out, transitivity of the comparator is
> essential.
>
> http://bost.ocks.org/mike/shuffle/compare.html

Thanks, that's really interesting. Also means that our implementation of
`shuffle` in Racket stdlib is not great:

  (define (shuffle l)
    (sort l < #:key (?(_) (random)) #:cache-keys? #t))

(more prose on the algorithm here: http://bost.ocks.org/mike/algorithms/#shuffling)

Cheers,
Asumu

From eric.n.dobson at gmail.com  Tue Sep 16 00:24:53 2014
From: eric.n.dobson at gmail.com (Eric Dobson)
Date: Mon, 15 Sep 2014 21:24:53 -0700
Subject: [racket] on reversing a list by way of sort
In-Reply-To: <20140916035049.GS21882@localhost>
References: <CAB08-qqQyfysyAr_uzcQ4rajfGJSVOj_06T8SC5pfXhA2apSNQ@mail.gmail.com>
 <541751DE.5050407@cs.umd.edu> <5417547D.3060309@cs.umd.edu>
 <CAL3TdON=6kCKdkqGbvRq-be5K9ewKOvMg1tWsUk79N5UGt971g@mail.gmail.com>
 <5417991F.8020807@mbtype.com> <20140916035049.GS21882@localhost>
Message-ID: <CAAEHQ5sBnqxaZPjVbU9emCdprsghNxKZtsVutxDyWM_yWpqnaA@mail.gmail.com>

On Mon, Sep 15, 2014 at 8:50 PM, Asumu Takikawa <asumu at ccs.neu.edu> wrote:
> On 2014-09-15 18:57:51 -0700, Matthew Butterick wrote:
>> Mike Bostock's visual demonstration of why naive sorting functions are false
>> friends. As he and Henglein point out, transitivity of the comparator is
>> essential.
>>
>> http://bost.ocks.org/mike/shuffle/compare.html
>
> Thanks, that's really interesting. Also means that our implementation of
> `shuffle` in Racket stdlib is not great:
>
>   (define (shuffle l)
>     (sort l < #:key (?(_) (random)) #:cache-keys? #t))
>
> (more prose on the algorithm here: http://bost.ocks.org/mike/algorithms/#shuffling)

This is different than linked algorithm. This one assigns a random
number from [0,1) to each element where the linked algorithm assigns a
random pairwise ordering to each comparison.

I don't think there is anything wrong from a correctness standpoint on
the stdlib's algorithm.

>
> Cheers,
> Asumu
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users


From asumu at ccs.neu.edu  Tue Sep 16 00:34:55 2014
From: asumu at ccs.neu.edu (Asumu Takikawa)
Date: Tue, 16 Sep 2014 00:34:55 -0400
Subject: [racket] on reversing a list by way of sort
In-Reply-To: <CAAEHQ5sBnqxaZPjVbU9emCdprsghNxKZtsVutxDyWM_yWpqnaA@mail.gmail.com>
References: <CAB08-qqQyfysyAr_uzcQ4rajfGJSVOj_06T8SC5pfXhA2apSNQ@mail.gmail.com>
 <541751DE.5050407@cs.umd.edu> <5417547D.3060309@cs.umd.edu>
 <CAL3TdON=6kCKdkqGbvRq-be5K9ewKOvMg1tWsUk79N5UGt971g@mail.gmail.com>
 <5417991F.8020807@mbtype.com> <20140916035049.GS21882@localhost>
 <CAAEHQ5sBnqxaZPjVbU9emCdprsghNxKZtsVutxDyWM_yWpqnaA@mail.gmail.com>
Message-ID: <20140916043455.GT21882@localhost>

On 2014-09-15 21:24:53 -0700, Eric Dobson wrote:
> This is different than linked algorithm. This one assigns a random
> number from [0,1) to each element where the linked algorithm assigns a
> random pairwise ordering to each comparison.
>
> I don't think there is anything wrong from a correctness standpoint on
> the stdlib's algorithm.

Ah, you're right. I actually tried visualizing Racket's shuffle using
picts and didn't notice any streaks so I was puzzled. That explains it.

Cheers,
Asumu

From mb at mbtype.com  Tue Sep 16 00:40:19 2014
From: mb at mbtype.com (Matthew Butterick)
Date: Mon, 15 Sep 2014 21:40:19 -0700
Subject: [racket] on reversing a list by way of sort
In-Reply-To: <20140916035049.GS21882@localhost>
References: <CAB08-qqQyfysyAr_uzcQ4rajfGJSVOj_06T8SC5pfXhA2apSNQ@mail.gmail.com>
 <541751DE.5050407@cs.umd.edu> <5417547D.3060309@cs.umd.edu>
 <CAL3TdON=6kCKdkqGbvRq-be5K9ewKOvMg1tWsUk79N5UGt971g@mail.gmail.com>
 <5417991F.8020807@mbtype.com> <20140916035049.GS21882@localhost>
Message-ID: <5417BF33.9070407@mbtype.com>

Haha, yes I too checked Racket `shuffle` after I read that. But 
`shuffle` is mathematically sound. (Was I really surprised?) It doesn't 
use a random comparator. Rather, it assigns random *keys* to the 
elements and then uses a standard `<` comparator, which preserves 
transitivity (net of the nanoscopic chance of `random` returning the 
same key twice).

I also verified this experimentally using Bostock's bias-correlation 
technique. Here's Racket `shuffle` on a 10-card deck, shuffled a million 
times (truly random shuffling will produce values close to zero):

+0.0037     -0.0037     +0.0107     -0.0087     +0.0049     +0.0432     
-0.0306     +0.0025     -0.0319     +0.0099
-0.0589     +0.0066     +0.0295     -0.0337     -0.0308     +0.0288     
-0.0106     +0.0375     +0.0202     +0.0114
-0.0191     +0.0381     -0.0308     -0.017     +0.0608     -0.0609     
+0.0183     +0.0037     -0.0037     +0.0106
-0.0025     -0.0232     -0.0135     -0.0166     -0.0127     -0.0104     
+0.0444     -0.0087     +0.0275     +0.0157
+0.0537     -0.0114     -0.0138     +0.0029     +0.0281     -0.0264     
-0.0369     +0.0237     -0.0051     -0.0148
+0.0075     +0.0254     -0.0128     +0.048     +0.0309     -0.0053     
-0.0055     -0.046     -0.0183     -0.0239
+0.0211     +0.005     -0.0162     +0.0326     -0.0019     -0.0062     
-0.0423     -0.0036     +0.0501     -0.0386
+0.027     +0.0228     -0.0327     -0.001     -0.0256     +0.013 +0.0005 
     +0.0203     -0.0157     -0.0086
-0.0039     -0.059     +0.0769     -0.0294     -0.0045     +0.0046     
+0.0121     -0.0035     -0.0344     +0.0411
-0.0286     -0.0006     +0.0027     +0.0229     -0.0492     +0.0196     
+0.0506     -0.0259     +0.0113     -0.0028

And here's the evil random-comparator shuffle, which you write thus: 
(sort l (?(a b) (< (random) 0.5)) #:cache-keys? #t)

+5.5262     +5.3953     +0.2165     -4.1699     -6.8645     +5.3694     
+5.3138     +0.2238     -4.1352     -6.8754
+5.3552     +5.5478     +0.2443     -4.1483     -6.8783     +5.2901     
+5.3356     +0.3061     -4.1741     -6.8784
+4.0834     +4.0231     +1.1727     -3.128     -6.0845     +4.0658     
+4.0908     +0.9624     -3.0706     -6.1151
+2.2696     +2.3085     +1.3081     -1.1473     -4.5383     +2.236     
+2.1868     +1.2539     -1.3606     -4.5167
+0.2824     +0.2988     +1.0478     +0.3734     -1.7924     +0.264     
+0.2866     +0.9759     +0.2977     -2.0342
-1.1146     -1.1627     +0.4308     +1.2617     +0.452     -0.885     
-1.1186     +0.4888     +1.2433     +0.4043
-1.6973     -1.7075     -0.0871     +1.4717     +1.8591     -1.6837     
-1.4007     -0.087     +1.4763     +1.8562
-2.9075     -2.945     +0.3826     +2.2916     +3.0588     -2.9184     
-2.9246     +0.5772     +2.3114     +3.0739
-4.9106     -4.8946     -0.9329     +4.7021     +5.8784     -4.8757     
-4.8899     -0.942     +4.9596     +5.9056
-6.8868     -6.8637     -3.7828     +2.493     +14.9097     -6.8625     
-6.8798     -3.7591     +2.4522     +15.1798

QED


On 09/15/14 8:50 PM, Asumu Takikawa wrote:
> On 2014-09-15 18:57:51 -0700, Matthew Butterick wrote:
>> Mike Bostock's visual demonstration of why naive sorting functions are false
>> friends. As he and Henglein point out, transitivity of the comparator is
>> essential.
>>
>> http://bost.ocks.org/mike/shuffle/compare.html
> Thanks, that's really interesting. Also means that our implementation of
> `shuffle` in Racket stdlib is not great:
>
>    (define (shuffle l)
>      (sort l < #:key (?(_) (random)) #:cache-keys? #t))
>
> (more prose on the algorithm here: http://bost.ocks.org/mike/algorithms/#shuffling)
>
> Cheers,
> Asumu


From mflatt at cs.utah.edu  Tue Sep 16 01:01:09 2014
From: mflatt at cs.utah.edu (Matthew Flatt)
Date: Tue, 16 Sep 2014 07:01:09 +0200
Subject: [racket] Can a #reader be used with (module ...) ?
In-Reply-To: <54174E28.4000001@mbtype.com>
References: <54174E28.4000001@mbtype.com>
Message-ID: <20140916050114.0F7CB650192@mail-svr1.cs.utah.edu>

There's not currently anything like `#reader` that would hide a closing
parenthesis from the reader.

I think you could make a new reader that recurs to `read` with an input
port that stops consuming input from the original just before a closing
parenthesis followed by an EOF --- maybe by peeking at all the input to
determine a length to pass to `call-with-limited-input-port`. Assuming
that the new reader is in `with-limited-port`, then

 #lang racket/base
 (module inner racket/base
    '#reader with-limited-port
     #reader (submod pollen/main reader)
    Hello World
    )

could work. Or you could build that port-limiting approach into a new
reader in a `limited-reader` submodule that always chains to the
sibling `reader` submodule:

 #lang racket/base
 (module inner racket/base
    '#reader(submod pollen/main limited-reader)Hello World
    )

At Mon, 15 Sep 2014 13:38:00 -0700, Matthew Butterick wrote:
> I'm aware that `#lang name ...` is equivalent to `(module id name ...)`. 
> But `#lang` automatically picks up a reader; `module` does not.
> 
> Is there a way to  attach a reader to the `module` form (the idea being 
> that then, within the `module` expression, you could write using the 
> native syntax of the #lang)?
> 
> Studying the example in reader extensions (Guide ? 17.2), I found that 
> this works, but I haven't figured out if it can be generalized:
> 
> #lang racket/base
> (module inner racket/base
>    '(#reader"five.rkt"23456))
> (require 'inner)
> 
> For instance, my current reader works in the REPL, like so:
> 
> '#reader(submod pollen/main reader)Hello World
> 
> But this doesn't work, because the reader, which operates in text mode, 
> absorbs the closing parenthesis:
> 
> #lang racket/base
> (module inner racket/base
>    '#reader(submod pollen/main reader)Hello World
>    )
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users


From sperber at deinprogramm.de  Tue Sep 16 04:01:00 2014
From: sperber at deinprogramm.de (Michael Sperber)
Date: Tue, 16 Sep 2014 10:01:00 +0200
Subject: [racket] BOB 2015 - 2nd Call for Contributions (Deadline Sep 30)
Message-ID: <y9ly4tkgfk3.fsf@deinprogramm.de>


Submissions on/using/reporting on Racket welcome!

                         BOB Conference 2015
                               Berlin
                             23.1.2015
                       http://bobkonf.de/2015/
                       CALL FOR CONTRIBUTIONS
               English: http://bobkonf.de/2015/cfp.html
                German: http://bobkonf.de/2015/cfp.html
                    Deadline: September 30, 2014

You drive advanced software engineering methods, implement ambitious
architectures and are open to cutting-edge innovation? Attend this
conference, meet people that share your goals, and get to know the
best software tools and technologies available today. We strive to
offer you a day full of new experiences and impressions that you can
use to immediately improve your daily life as a software developer.

If you share our vision and want to contribute, submit a proposal for 
a talk or tutorial!

We are looking for talks about best-of-breed software technology,
e.g.:

- functional programming
- reactive programming
- micro-service architectures
- persistent data structures and databases
- ?  everything really that isn?t mainstream, but you think should 
  be.

Presenters should provide the audience with information that is
practically useful for software developers.  This could take the form of
e.g.:

- experience reports
- introductory talks on technical background
- demos and how-tos

We accept proposals for presentations of 45 minutes (40 minutes talk 
+ 5 minutes questions), as well as 90 minute tutorials for beginners.
The language of presentation should be either English or German.
It should include (in your presentation language of choice):

- an abstract of max. 1500 characters.
- a short bio/cv
- contact information (including at least email)
- a list of 3-5 concrete ideas of how your work can be applied in a developer's daily life
- additional material (websites, blogs, slides, videos of past
  presentations, ?)

You can submit your proposal using the following form:

https://docs.google.com/spreadsheet/viewform?formkey=dHJ0TjR1cEhUWmdBZFVITGVRVWN5VEE6MA

- direct questions to bobkonf at active minus group dot de
- proposal deadline: September 30, 2014
- notification: October 15, 2014
- program: October 2014, 2014

NOTE: The conference fee will be waived for presenters, but travel 
expenses will not be covered.

Program Committee
- Matthias Fischmann, zerobuzz UG
- Matthias Neubauer, SICK AG
- Michael Sperber, Active Group
- Stefan Wehr, factis research

Wissenschaftlicher Beirat
- Annette Bieniusa, TU Kaiserslautern
- Torsten Grust, Uni T?bingen
- Peter Thiemann, Uni Freiburg



From eli at barzilay.org  Tue Sep 16 06:12:54 2014
From: eli at barzilay.org (Eli Barzilay)
Date: Tue, 16 Sep 2014 06:12:54 -0400
Subject: [racket] on reversing a list by way of sort
In-Reply-To: <5417BF33.9070407@mbtype.com>
References: <CAB08-qqQyfysyAr_uzcQ4rajfGJSVOj_06T8SC5pfXhA2apSNQ@mail.gmail.com>
 <541751DE.5050407@cs.umd.edu> <5417547D.3060309@cs.umd.edu>
 <CAL3TdON=6kCKdkqGbvRq-be5K9ewKOvMg1tWsUk79N5UGt971g@mail.gmail.com>
 <5417991F.8020807@mbtype.com> <20140916035049.GS21882@localhost>
 <5417BF33.9070407@mbtype.com>
Message-ID: <CALO-guvVYvycZF8iv9hjn+jBYy_5nSM6moLf0iQLBNafkrnhFw@mail.gmail.com>

On Tue, Sep 16, 2014 at 12:34 AM, Asumu Takikawa <asumu at ccs.neu.edu> wrote:
>
> Ah, you're right. I actually tried visualizing Racket's shuffle using
> picts and didn't notice any streaks so I was puzzled. That explains it.

The equivalent of Racket's shuffle on that page is the "sort (random
order)", except that their version depends on a continuous range of
integers (which the visualization always does).


On Tue, Sep 16, 2014 at 12:40 AM, Matthew Butterick <mb at mbtype.com> wrote:
>
> And here's the evil random-comparator shuffle, which you write thus: (sort l
> (?(a b) (< (random) 0.5)) #:cache-keys? #t)

(There's no need for caching the keys here.)

-- 
          ((lambda (x) (x x)) (lambda (x) (x x)))          Eli Barzilay:
                    http://barzilay.org/                   Maze is Life!


From robby at eecs.northwestern.edu  Tue Sep 16 08:22:19 2014
From: robby at eecs.northwestern.edu (Robby Findler)
Date: Tue, 16 Sep 2014 07:22:19 -0500
Subject: [racket] on reversing a list by way of sort
In-Reply-To: <5417BF33.9070407@mbtype.com>
References: <CAB08-qqQyfysyAr_uzcQ4rajfGJSVOj_06T8SC5pfXhA2apSNQ@mail.gmail.com>
 <541751DE.5050407@cs.umd.edu> <5417547D.3060309@cs.umd.edu>
 <CAL3TdON=6kCKdkqGbvRq-be5K9ewKOvMg1tWsUk79N5UGt971g@mail.gmail.com>
 <5417991F.8020807@mbtype.com> <20140916035049.GS21882@localhost>
 <5417BF33.9070407@mbtype.com>
Message-ID: <CAL3TdONJYHmKJ8H3AW7ZYhtFfKPAO=npAO=y3z_=zkqay-xZiA@mail.gmail.com>

But is fisher-yates is linear time, right? (And probably more
efficient in practice on larger lists, if not smaller.)

Robby

On Mon, Sep 15, 2014 at 11:40 PM, Matthew Butterick <mb at mbtype.com> wrote:
> Haha, yes I too checked Racket `shuffle` after I read that. But `shuffle` is
> mathematically sound. (Was I really surprised?) It doesn't use a random
> comparator. Rather, it assigns random *keys* to the elements and then uses a
> standard `<` comparator, which preserves transitivity (net of the nanoscopic
> chance of `random` returning the same key twice).
>
> I also verified this experimentally using Bostock's bias-correlation
> technique. Here's Racket `shuffle` on a 10-card deck, shuffled a million
> times (truly random shuffling will produce values close to zero):
>
> +0.0037     -0.0037     +0.0107     -0.0087     +0.0049     +0.0432
> -0.0306     +0.0025     -0.0319     +0.0099
> -0.0589     +0.0066     +0.0295     -0.0337     -0.0308     +0.0288
> -0.0106     +0.0375     +0.0202     +0.0114
> -0.0191     +0.0381     -0.0308     -0.017     +0.0608     -0.0609
> +0.0183     +0.0037     -0.0037     +0.0106
> -0.0025     -0.0232     -0.0135     -0.0166     -0.0127     -0.0104
> +0.0444     -0.0087     +0.0275     +0.0157
> +0.0537     -0.0114     -0.0138     +0.0029     +0.0281     -0.0264
> -0.0369     +0.0237     -0.0051     -0.0148
> +0.0075     +0.0254     -0.0128     +0.048     +0.0309     -0.0053
> -0.0055     -0.046     -0.0183     -0.0239
> +0.0211     +0.005     -0.0162     +0.0326     -0.0019     -0.0062
> -0.0423     -0.0036     +0.0501     -0.0386
> +0.027     +0.0228     -0.0327     -0.001     -0.0256     +0.013 +0.0005
> +0.0203     -0.0157     -0.0086
> -0.0039     -0.059     +0.0769     -0.0294     -0.0045     +0.0046
> +0.0121     -0.0035     -0.0344     +0.0411
> -0.0286     -0.0006     +0.0027     +0.0229     -0.0492     +0.0196
> +0.0506     -0.0259     +0.0113     -0.0028
>
> And here's the evil random-comparator shuffle, which you write thus: (sort l
> (?(a b) (< (random) 0.5)) #:cache-keys? #t)
>
> +5.5262     +5.3953     +0.2165     -4.1699     -6.8645     +5.3694
> +5.3138     +0.2238     -4.1352     -6.8754
> +5.3552     +5.5478     +0.2443     -4.1483     -6.8783     +5.2901
> +5.3356     +0.3061     -4.1741     -6.8784
> +4.0834     +4.0231     +1.1727     -3.128     -6.0845     +4.0658
> +4.0908     +0.9624     -3.0706     -6.1151
> +2.2696     +2.3085     +1.3081     -1.1473     -4.5383     +2.236
> +2.1868     +1.2539     -1.3606     -4.5167
> +0.2824     +0.2988     +1.0478     +0.3734     -1.7924     +0.264
> +0.2866     +0.9759     +0.2977     -2.0342
> -1.1146     -1.1627     +0.4308     +1.2617     +0.452     -0.885
> -1.1186     +0.4888     +1.2433     +0.4043
> -1.6973     -1.7075     -0.0871     +1.4717     +1.8591     -1.6837
> -1.4007     -0.087     +1.4763     +1.8562
> -2.9075     -2.945     +0.3826     +2.2916     +3.0588     -2.9184
> -2.9246     +0.5772     +2.3114     +3.0739
> -4.9106     -4.8946     -0.9329     +4.7021     +5.8784     -4.8757
> -4.8899     -0.942     +4.9596     +5.9056
> -6.8868     -6.8637     -3.7828     +2.493     +14.9097     -6.8625
> -6.8798     -3.7591     +2.4522     +15.1798
>
> QED
>
>
>
> On 09/15/14 8:50 PM, Asumu Takikawa wrote:
>>
>> On 2014-09-15 18:57:51 -0700, Matthew Butterick wrote:
>>>
>>> Mike Bostock's visual demonstration of why naive sorting functions are
>>> false
>>> friends. As he and Henglein point out, transitivity of the comparator is
>>> essential.
>>>
>>> http://bost.ocks.org/mike/shuffle/compare.html
>>
>> Thanks, that's really interesting. Also means that our implementation of
>> `shuffle` in Racket stdlib is not great:
>>
>>    (define (shuffle l)
>>      (sort l < #:key (?(_) (random)) #:cache-keys? #t))
>>
>> (more prose on the algorithm here:
>> http://bost.ocks.org/mike/algorithms/#shuffling)
>>
>> Cheers,
>> Asumu
>
>
> ____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users


From geoff at knauth.org  Tue Sep 16 09:22:42 2014
From: geoff at knauth.org (Geoffrey S. Knauth)
Date: Tue, 16 Sep 2014 09:22:42 -0400
Subject: [racket] pregexp to detect Japanese characters
Message-ID: <1410873762.3653302.168100665.5BFF00B3@webmail.messagingengine.com>

I'm writing a function to detect Japanese characters in a
string. I found this page:



[1]http://www.localizingjapan.com/blog/2012/01/20/regular-expre
ssions-for-japanese-text/



So, for example, the example Perl regexp [\x{3041}-\x{3096}]
would detect Hiragana characters (as would \p{Hiragana}).  How
do I express such a Unicode range with Racket regexps?



I looked at the docs below and it wasn't obvious to me how to
do it.  In other languages there might be, for example, a
\xnnnn or \uxxxx construct.



[2]http://docs.racket-lang.org/reference/regexp.html#%28elem._%
28rxex._30%29%29



--

Geoffrey S. Knauth | http://knauth.org/gsk

References

1. http://www.localizingjapan.com/blog/2012/01/20/regular-expressions-for-japanese-text/
2. http://docs.racket-lang.org/reference/regexp.html#%28elem._%28rxex._30%29%29
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140916/76caf3c8/attachment.html>

From robby at eecs.northwestern.edu  Tue Sep 16 09:55:26 2014
From: robby at eecs.northwestern.edu (Robby Findler)
Date: Tue, 16 Sep 2014 08:55:26 -0500
Subject: [racket] OOPSLA/SPLASH early registration deadline approaching
Message-ID: <CAL3TdOPE0SF3jbpSaKC7CUzS6J9BRJqDvbUTu0fYTzcQtyfnFg@mail.gmail.com>

This is a reminder that the SPLASH Early Registration closes this week
on Friday 19 September.

http://2014.splashcon.org/attending/registration

Also please don?t forget to reserve your hotel room at the Portland
Marriott Waterfront. [1]

  ACM Conference on
  Systems, Programming, Languages, and Applications:
  Software for Humanity (SPLASH'14)
  Portland, Oregon, USA

  20-24 October, 2014
  http://www.splashcon.org
  http://twitter.com/splashcon
  http://www.facebook.com/SPLASHCon

  Sponsored by ACM SIGPLAN
  In cooperation with ACM SIGAda

The ACM SIGPLAN conference on Systems, Programming, Languages and
Applications: Software for Humanity (SPLASH) embraces all aspects
of software construction and delivery to make it the premier
conference at the intersection of programming, languages, and
software engineering.

[1] Why stay at the conference hotel?
http://2014.splashcon.org/attending/accommodation


From geoff at knauth.org  Tue Sep 16 10:13:38 2014
From: geoff at knauth.org (Geoffrey S. Knauth)
Date: Tue, 16 Sep 2014 10:13:38 -0400
Subject: [racket] pregexp to detect Japanese characters
In-Reply-To: <1410873762.3653302.168100665.5BFF00B3@webmail.messagingengine.com>
References: <1410873762.3653302.168100665.5BFF00B3@webmail.messagingengine.com>
Message-ID: <10A07DD2-53D6-4C8E-9CAE-89AA320306A0@knauth.org>

NEVER MIND.  I figured it out:

; detects japanese characters
; contains-japanese-characters? str -> bool
(define (contains-japanese-characters? s)
 (or (regexp-match #rx"[\u3041-\u3096]" s)  ; Hiragana
     (regexp-match #rx"[\u30A0-\u30FF]" s)  ; Katakana (Full Width)
     (regexp-match #rx"[\u3400-\u4DB5\u4E00-\u9FCB\uF900-\uFA6A]" s)  ; Kanji
     (regexp-match #rx"[\u2E80-\u2FD5]" s)  ; Kanji Radicals
     (regexp-match #rx"[\uFF5F-\uFF9F]" s)  ; Katakana and Punctuation (Half Width)
     (regexp-match #rx"[\u3000-\u303F]" s)  ; Japanese Symbols and Punctuation
     (regexp-match #rx"[\u31F0-\u31FF\u3220-\u3243\u3280-\u337F]" s)  ; Misc. Japanese Symbols/Chars
     (regexp-match #rx"[\uFF01-\uFF5E]" s)))  ; Alphanumeric and Punctuation (Full Width)

On Sep 16, 2014, at 09:22 , Geoffrey S. Knauth <geoff at knauth.org> wrote:

> I'm writing a function to detect Japanese characters in a string. I found this page:
>  
> http://www.localizingjapan.com/blog/2012/01/20/regular-expressions-for-japanese-text/
>  
> So, for example, the example Perl regexp [\x{3041}-\x{3096}] would detect Hiragana characters (as would \p{Hiragana}).  How do I express such a Unicode range with Racket regexps?
>  
> I looked at the docs below and it wasn't obvious to me how to do it.  In other languages there might be, for example, a \xnnnn or \uxxxx construct.
>  
> http://docs.racket-lang.org/reference/regexp.html#%28elem._%28rxex._30%29%29
>  
> --
> Geoffrey S. Knauth | http://knauth.org/gsk
>  
> ____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140916/1cba68f9/attachment-0001.html>

From dpavlov at ipa.nw.ru  Tue Sep 16 17:12:38 2014
From: dpavlov at ipa.nw.ru (dpavlov at ipa.nw.ru)
Date: Wed, 17 Sep 2014 01:12:38 +0400
Subject: [racket] Spreadsheet editor
Message-ID: <6bc0c4456cc37557ac97f5e8ea2bed2f@ipa.nw.ru>

Hello,

I have made a reusable version of my spreadsheet editor.
Its capabilities are limited to the needs of my app, so
not much:

- Fixed number of columns with arbitrary labels, which
   can be changed while running. The number of columns
   does not scale well: 1000 is OK, while 100 000 is not.
   The width of the column is adjusted dynamically so that
   the column label is visible and all the visible cells
   are not cut. Otherwise, column width is not changeable.

- Arbitrary number of rows with arbitrary labels.
   As many rows as you want. Rows can be added or removed
   while running.

- Clickable buttons for columns and rows with arbitrary
   callbacks.

- Editable (text-only, one-line) cells with arbitrary
   callbacks for displaying the cells or updating their
   contents.

- Scrolling bars (well, sliders), also arrow keys and
   PgUp/PgDn should work.


Here is the editor along with a demo:

https://github.com/kugelblitz/spreadsheet-editor


I see two issues currently:

* During editing, text snip creates a white plate that
   extends outside the supposed borders of the snip.
   See [1] for details.

* Racket GUI does not provide scroll bars that one can
   use independently from canvas or panel. I currently
   use sliders instead of scroll bars, but with slider,
   it is impossible to go up/down by exactly one point.
   Also, it is not possible to update the limits of the
   slider, so I have to re-create it when I add or remove
   a row from the table.

I would love to hear any comments regarding those issues,
or any other issues, or anything that draws your
attention in the source code.


[1] 
http://lists.racket-lang.org/users/archive/2014-September/064137.html

From jab at math.brown.edu  Tue Sep 16 19:06:59 2014
From: jab at math.brown.edu (jab at math.brown.edu)
Date: Tue, 16 Sep 2014 19:06:59 -0400
Subject: [racket] canonical index of Racket courses? [was: Summer
 programs learning Racket for a student]
In-Reply-To: <CANwREeXB1EWu2SjHBHuzhJKba4dHJFhdGHdg7SXafJqJ=v-nsw@mail.gmail.com>
References: <CANwREeXB1EWu2SjHBHuzhJKba4dHJFhdGHdg7SXafJqJ=v-nsw@mail.gmail.com>
Message-ID: <CANwREeXVeAtEirzyXOGQedt8uU4EK--qYU--ZkNRGKWQ5hn8Ag@mail.gmail.com>

Dear Racket Users,


On Wed, Apr 16, 2014 at 4:45 PM, <jab at math.brown.edu> wrote:

> I recently asked here about in-person Racket courses available this summer
> which I could recommend to a student I've been tutoring. (Thanks again to
> everyone who let me know about their courses so far.)
>
> To make this info easier to find, I'm wondering if the racket-lang.org
> maintainers would be interested in publishing a canonical index of Racket
> courses (both online and in-person) linked off the "Community" or
> "Learning" sections on the homepage, and inviting the community to add to
> it.
>
> (Similarly, linking to a canonical index of Racket meetup groups (? la
> http://clojure.meetup.com/ and
> https://wiki.python.org/moin/LocalUserGroups)  would be great too. Google
> turned up http://racket.meetup.com/ but 2 of the 3 results there are for
> racket sports.)
>
> Thanks, and looking forward to knowing how to become better plugged into
> the Racket community!
>


Didn't get any reply to this last April, so just thought I'd try again. In
case anyone who'd want to reply just missed it! Wish I could join those of
you in St. Louis this week.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140916/b7329d77/attachment.html>

From matthias at ccs.neu.edu  Tue Sep 16 20:43:07 2014
From: matthias at ccs.neu.edu (Matthias Felleisen)
Date: Tue, 16 Sep 2014 20:43:07 -0400
Subject: [racket] canonical index of Racket courses? [was: Summer
	programs learning Racket for a student]
In-Reply-To: <CANwREeXVeAtEirzyXOGQedt8uU4EK--qYU--ZkNRGKWQ5hn8Ag@mail.gmail.com>
References: <CANwREeXB1EWu2SjHBHuzhJKba4dHJFhdGHdg7SXafJqJ=v-nsw@mail.gmail.com>
 <CANwREeXVeAtEirzyXOGQedt8uU4EK--qYU--ZkNRGKWQ5hn8Ag@mail.gmail.com>
Message-ID: <CADBC08E-D847-4AAE-87BE-D62368E91936@ccs.neu.edu>


You're right, we failed to follow up on the general part of your email. 

Do you imagine listing college courses such as Brown's 17, which uses DrRacket and the teaching languages? Or do you want Coursera courses that everyone can access? Do you want it all? now? :-) 

And yes, we should probably create a wiki like thing for such an effort or perhaps something like packages.racket-lang.org. I'll bring it up as we meet in St Louis 

-- Matthias





On Sep 16, 2014, at 7:06 PM, jab at math.brown.edu wrote:

> Dear Racket Users,
> 
> 
> On Wed, Apr 16, 2014 at 4:45 PM, <jab at math.brown.edu> wrote:
> I recently asked here about in-person Racket courses available this summer which I could recommend to a student I've been tutoring. (Thanks again to everyone who let me know about their courses so far.)
> 
> To make this info easier to find, I'm wondering if the racket-lang.org maintainers would be interested in publishing a canonical index of Racket courses (both online and in-person) linked off the "Community" or "Learning" sections on the homepage, and inviting the community to add to it.
> 
> (Similarly, linking to a canonical index of Racket meetup groups (? la http://clojure.meetup.com/ and https://wiki.python.org/moin/LocalUserGroups)  would be great too. Google turned up http://racket.meetup.com/ but 2 of the 3 results there are for racket sports.)
> 
> Thanks, and looking forward to knowing how to become better plugged into the Racket community!
> 
> 
> Didn't get any reply to this last April, so just thought I'd try again. In case anyone who'd want to reply just missed it! Wish I could join those of you in St. Louis this week.
> ____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140916/89fbd185/attachment.html>

From eli at barzilay.org  Tue Sep 16 21:22:27 2014
From: eli at barzilay.org (Eli Barzilay)
Date: Tue, 16 Sep 2014 21:22:27 -0400
Subject: [racket] on reversing a list by way of sort
In-Reply-To: <CAL3TdONJYHmKJ8H3AW7ZYhtFfKPAO=npAO=y3z_=zkqay-xZiA@mail.gmail.com>
References: <CAB08-qqQyfysyAr_uzcQ4rajfGJSVOj_06T8SC5pfXhA2apSNQ@mail.gmail.com>
 <541751DE.5050407@cs.umd.edu> <5417547D.3060309@cs.umd.edu>
 <CAL3TdON=6kCKdkqGbvRq-be5K9ewKOvMg1tWsUk79N5UGt971g@mail.gmail.com>
 <5417991F.8020807@mbtype.com> <20140916035049.GS21882@localhost>
 <5417BF33.9070407@mbtype.com>
 <CAL3TdONJYHmKJ8H3AW7ZYhtFfKPAO=npAO=y3z_=zkqay-xZiA@mail.gmail.com>
Message-ID: <CALO-gus=khdeVX4LY01KkmAX69tcqZDhMc1vbZQyRB4ujE1cgA@mail.gmail.com>

On Tue, Sep 16, 2014 at 8:22 AM, Robby Findler
<robby at eecs.northwestern.edu> wrote:
> But is fisher-yates is linear time, right? (And probably more
> efficient in practice on larger lists, if not smaller.)

Yes, but there are more considerations that I had when I did the simple
version.  In case anyone wants to tackle this, what I vaguely remember
is:

* FY is faster, but will require much more than a one-liner, since the
  list needs to be copied into a vector, then shuffled in-place, then
  copied back.  I'm not saying that this will make it slower -- since
  sort is doing a vector round-trip anyway; it'll just make it much
  heavier than something that is not needed too often, and probably very
  rarely with big lists when the differene would matter.

* Looking at the wikipedia page to refresh myself, there is also an
  "inside-out" algorithm that they say is better for creating a shuffled
  copy of the input vector.  The nice thing in that is that it scans the
  source vector in sequence, which means that it could work well to use
  it, taking items from the list and putting them into an array that is
  then converted back into a list.

* I vaguely remember something about the problem of using integer random
  numbers, which are usually the result of modulo (the WP page talks
  about it too).  I think that there was some concern about that
  possibly leading to some bias in the result somehow.  As a result,
  since I don't know enough about such issues, I used (random) which has
  the additional penalty of allocating floats.  IOW, the nice thing
  about the current one-liner is that I didn't need to go too deep into
  considering such issues, allowing me to keep my ignorance while being
  relatively confident in the result not suffering from any bias.

  In any case, this is also a point to consider if anyone wants to
  implement FY.

-- 
          ((lambda (x) (x x)) (lambda (x) (x x)))          Eli Barzilay:
                    http://barzilay.org/                   Maze is Life!

From daniel.a.prager at gmail.com  Tue Sep 16 23:08:22 2014
From: daniel.a.prager at gmail.com (Daniel Prager)
Date: Tue, 16 Sep 2014 22:08:22 -0500
Subject: [racket] on reversing a list by way of sort
In-Reply-To: <CALO-gus=khdeVX4LY01KkmAX69tcqZDhMc1vbZQyRB4ujE1cgA@mail.gmail.com>
References: <CAB08-qqQyfysyAr_uzcQ4rajfGJSVOj_06T8SC5pfXhA2apSNQ@mail.gmail.com>
 <541751DE.5050407@cs.umd.edu> <5417547D.3060309@cs.umd.edu>
 <CAL3TdON=6kCKdkqGbvRq-be5K9ewKOvMg1tWsUk79N5UGt971g@mail.gmail.com>
 <5417991F.8020807@mbtype.com> <20140916035049.GS21882@localhost>
 <5417BF33.9070407@mbtype.com>
 <CAL3TdONJYHmKJ8H3AW7ZYhtFfKPAO=npAO=y3z_=zkqay-xZiA@mail.gmail.com>
 <CALO-gus=khdeVX4LY01KkmAX69tcqZDhMc1vbZQyRB4ujE1cgA@mail.gmail.com>
Message-ID: <CAFKxZVX2rSUROs_iTHOqGur9xJ6BZFkkmsC64V=dAA=drYJZxQ@mail.gmail.com>

Here's a version of "inside-out" Fisher-Yates that should fit the bill:

(define (fy-shuffle lst)
  (define v (list->vector lst))
  (for/list ([n (in-range (length lst) 0 -1)])
    (let* ([i (sub1 n)]
           [j (random n)]
           [v_j (vector-ref v j)])
      (when (not (= i j))
        (vector-set! v j (vector-ref v i)))
      v_j)))

Dan


On Tue, Sep 16, 2014 at 8:22 PM, Eli Barzilay <eli at barzilay.org> wrote:

> On Tue, Sep 16, 2014 at 8:22 AM, Robby Findler
> <robby at eecs.northwestern.edu> wrote:
> > But is fisher-yates is linear time, right? (And probably more
> > efficient in practice on larger lists, if not smaller.)
>
> Yes, but there are more considerations that I had when I did the simple
> version.  In case anyone wants to tackle this, what I vaguely remember
> is:
>
> * FY is faster, but will require much more than a one-liner, since the
>   list needs to be copied into a vector, then shuffled in-place, then
>   copied back.  I'm not saying that this will make it slower -- since
>   sort is doing a vector round-trip anyway; it'll just make it much
>   heavier than something that is not needed too often, and probably very
>   rarely with big lists when the differene would matter.
>
> * Looking at the wikipedia page to refresh myself, there is also an
>   "inside-out" algorithm that they say is better for creating a shuffled
>   copy of the input vector.  The nice thing in that is that it scans the
>   source vector in sequence, which means that it could work well to use
>   it, taking items from the list and putting them into an array that is
>   then converted back into a list.
>
> * I vaguely remember something about the problem of using integer random
>   numbers, which are usually the result of modulo (the WP page talks
>   about it too).  I think that there was some concern about that
>   possibly leading to some bias in the result somehow.  As a result,
>   since I don't know enough about such issues, I used (random) which has
>   the additional penalty of allocating floats.  IOW, the nice thing
>   about the current one-liner is that I didn't need to go too deep into
>   considering such issues, allowing me to keep my ignorance while being
>   relatively confident in the result not suffering from any bias.
>
>   In any case, this is also a point to consider if anyone wants to
>   implement FY.
>
> --
>           ((lambda (x) (x x)) (lambda (x) (x x)))          Eli Barzilay:
>                     http://barzilay.org/                   Maze is Life!
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>



-- 
*Daniel Prager*
Agile/Lean Coaching, Software Development and Leadership
Startup: www.youpatch.com
Twitter: @agilejitsu <https://twitter.com/agilejitsu>
Blog: agile-jitsu.blogspot.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140916/131326b4/attachment-0001.html>

From neil at neilvandyke.org  Tue Sep 16 23:31:29 2014
From: neil at neilvandyke.org (Neil Van Dyke)
Date: Tue, 16 Sep 2014 23:31:29 -0400
Subject: [racket] new Jr. Racket Developer job
Message-ID: <54190091.80406@neilvandyke.org>

I'm helping a colleague hire a smart junior-level Racket developer.

The job spec is still being ironed-out, but a top candidate would be 
someone smart and responsible, and who has some experience with all of 
Racket, Java, and Python.

More important than any particular languages/tools is the demonstrated 
capacity to grasp more than one programming paradigm. So, if you've done 
idiomatic programming in each of, say, Racket, SIMULA, and ALGOL, you'd 
still be a good candidate.  Already knowing some Racket really helps you 
stand out, though.

This is for a real full-time salaried job, with the usual benefits. 
Telecommuting is possible, and the office is in the Eastern NY state 
area (not NYC area).

The company is not a startup nor a dotcom, and it is not all 
Google-exciting (no snack rooms, no plastic ball pools, no concierges).  
This is an established, stable, small-to-mid company, in a group that 
works on sober things like safety-critical process improvement.

An official job posting will be made later, once details are figured 
out.  But I'm testing the waters initially just on this Racket email 
list.  So, if you're still reading this post to this Racket email list, 
and you're interested in this junior-level Racket developer position, 
then please feel free to email your resume to me now (in PDF or ASCII 
format), and I'll pass it along.

BTW, the job is "junior" only in terms of professional Racket 
development experience, and in the challenge level of work needed right 
now.  It's very much open to all ages, and diversity is also welcome in 
general at the company.

Please *don't* re-post this elsewhere at this time; my email beep thanks 
you.

Neil V.


From eli at barzilay.org  Wed Sep 17 00:01:45 2014
From: eli at barzilay.org (Eli Barzilay)
Date: Wed, 17 Sep 2014 00:01:45 -0400
Subject: [racket] on reversing a list by way of sort
In-Reply-To: <CAFKxZVX2rSUROs_iTHOqGur9xJ6BZFkkmsC64V=dAA=drYJZxQ@mail.gmail.com>
References: <CAB08-qqQyfysyAr_uzcQ4rajfGJSVOj_06T8SC5pfXhA2apSNQ@mail.gmail.com>
 <541751DE.5050407@cs.umd.edu> <5417547D.3060309@cs.umd.edu>
 <CAL3TdON=6kCKdkqGbvRq-be5K9ewKOvMg1tWsUk79N5UGt971g@mail.gmail.com>
 <5417991F.8020807@mbtype.com> <20140916035049.GS21882@localhost>
 <5417BF33.9070407@mbtype.com>
 <CAL3TdONJYHmKJ8H3AW7ZYhtFfKPAO=npAO=y3z_=zkqay-xZiA@mail.gmail.com>
 <CALO-gus=khdeVX4LY01KkmAX69tcqZDhMc1vbZQyRB4ujE1cgA@mail.gmail.com>
 <CAFKxZVX2rSUROs_iTHOqGur9xJ6BZFkkmsC64V=dAA=drYJZxQ@mail.gmail.com>
Message-ID: <CALO-guuAm2x6EppHpWD12LifKYTaty3uRajC4ERu8H9deu4uDw@mail.gmail.com>

In Tue, Sep 16, 2014 at 11:08 PM, Daniel Prager
<daniel.a.prager at gmail.com> wrote:
> Here's a version of "inside-out" Fisher-Yates that should fit the bill:
>
> (define (fy-shuffle lst)
>   (define v (list->vector lst))
>   (for/list ([n (in-range (length lst) 0 -1)])
>     (let* ([i (sub1 n)]
>            [j (random n)]
>            [v_j (vector-ref v j)])
>       (when (not (= i j))
>         (vector-set! v j (vector-ref v i)))
>       v_j)))

I was thinking more of a direct translation, which (I think) is easier
to see that it's correct:

    (define (shuffle l)
      (define a (make-vector (length l)))
      (for ([x (in-list l)] [i (in-naturals)])
        (define j (random (add1 i)))
        (unless (= j i) (vector-set! a i (vector-ref a j)))
        (vector-set! a j x))
      (vector->list a))

But the issues from my last email are still there.

-- 
          ((lambda (x) (x x)) (lambda (x) (x x)))          Eli Barzilay:
                    http://barzilay.org/                   Maze is Life!

From daniel.a.prager at gmail.com  Wed Sep 17 01:53:59 2014
From: daniel.a.prager at gmail.com (Daniel Prager)
Date: Wed, 17 Sep 2014 00:53:59 -0500
Subject: [racket] on reversing a list by way of sort
In-Reply-To: <CALO-guuAm2x6EppHpWD12LifKYTaty3uRajC4ERu8H9deu4uDw@mail.gmail.com>
References: <CAB08-qqQyfysyAr_uzcQ4rajfGJSVOj_06T8SC5pfXhA2apSNQ@mail.gmail.com>
 <541751DE.5050407@cs.umd.edu> <5417547D.3060309@cs.umd.edu>
 <CAL3TdON=6kCKdkqGbvRq-be5K9ewKOvMg1tWsUk79N5UGt971g@mail.gmail.com>
 <5417991F.8020807@mbtype.com> <20140916035049.GS21882@localhost>
 <5417BF33.9070407@mbtype.com>
 <CAL3TdONJYHmKJ8H3AW7ZYhtFfKPAO=npAO=y3z_=zkqay-xZiA@mail.gmail.com>
 <CALO-gus=khdeVX4LY01KkmAX69tcqZDhMc1vbZQyRB4ujE1cgA@mail.gmail.com>
 <CAFKxZVX2rSUROs_iTHOqGur9xJ6BZFkkmsC64V=dAA=drYJZxQ@mail.gmail.com>
 <CALO-guuAm2x6EppHpWD12LifKYTaty3uRajC4ERu8H9deu4uDw@mail.gmail.com>
Message-ID: <CAFKxZVW+2HqUK_xNQXuGTXSvwnZxerfvWFgebhF1iS3DxHKS5Q@mail.gmail.com>

Hi Eli

I tried to squeeze out a little more efficiency, hopefully not at the
expense of too much loss of clarity, but I'm not overly fussed which exact
version of Fisher-Yates is (hopefully ;-) adopted.

[Sorry for drawing you further in.]

My take on your 3 points:

   1. Fisher-Yates is only a few lines, so although not a one-liner, it
   seems reasonable to use  it for the better space and time performance.
   2. I agree that an inside-out version is more apt.
   3. On my reading the final point is an issue if (random n) has problems
   like modulo bias, but if that's the case surely it is (random n) that needs
   fixing.


Dan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140917/906aacbd/attachment.html>

From mukeshtiwari.iiitm at gmail.com  Wed Sep 17 02:04:19 2014
From: mukeshtiwari.iiitm at gmail.com (mukesh tiwari)
Date: Wed, 17 Sep 2014 11:34:19 +0530
Subject: [racket] Racket guide
Message-ID: <CAFHZvE8gy7SGOUevg1bXpnM_qpD4K3cSwwK+VPjVA15GdHcbZw@mail.gmail.com>

Dear Racket Users,
I am learning racket and I found little bit misleading statement on guide
[1]. It says that "*Both the my-length and my-map functions run in O(n)
time for a list of length n.*". Technically this statement is correct but
when I read below

"

*For a list with n elements, evaluation will stack up n (+ 1 ...)
additions, and then finally add them up when the list is exhausted.You can
avoid piling up additions by adding along the way. To accumulate a length
this way, we need a function that takes both a list and the length of the
list seen so far; the code below uses a local function iter that
accumulates the length in an argument len*".

so just curious, shouldn't it be "*Both the my-length and my-map functions
run in O(n) space for a list of length n.*" ?


-Mukesh Tiwari



[1]
http://docs.racket-lang.org/guide/Lists__Iteration__and_Recursion.html#%28part._tail-recursion%29
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140917/fc31824b/attachment.html>

From eli at barzilay.org  Wed Sep 17 02:44:40 2014
From: eli at barzilay.org (Eli Barzilay)
Date: Wed, 17 Sep 2014 02:44:40 -0400
Subject: [racket] on reversing a list by way of sort
In-Reply-To: <CAFKxZVW+2HqUK_xNQXuGTXSvwnZxerfvWFgebhF1iS3DxHKS5Q@mail.gmail.com>
References: <CAB08-qqQyfysyAr_uzcQ4rajfGJSVOj_06T8SC5pfXhA2apSNQ@mail.gmail.com>
 <541751DE.5050407@cs.umd.edu> <5417547D.3060309@cs.umd.edu>
 <CAL3TdON=6kCKdkqGbvRq-be5K9ewKOvMg1tWsUk79N5UGt971g@mail.gmail.com>
 <5417991F.8020807@mbtype.com> <20140916035049.GS21882@localhost>
 <5417BF33.9070407@mbtype.com>
 <CAL3TdONJYHmKJ8H3AW7ZYhtFfKPAO=npAO=y3z_=zkqay-xZiA@mail.gmail.com>
 <CALO-gus=khdeVX4LY01KkmAX69tcqZDhMc1vbZQyRB4ujE1cgA@mail.gmail.com>
 <CAFKxZVX2rSUROs_iTHOqGur9xJ6BZFkkmsC64V=dAA=drYJZxQ@mail.gmail.com>
 <CALO-guuAm2x6EppHpWD12LifKYTaty3uRajC4ERu8H9deu4uDw@mail.gmail.com>
 <CAFKxZVW+2HqUK_xNQXuGTXSvwnZxerfvWFgebhF1iS3DxHKS5Q@mail.gmail.com>
Message-ID: <CALO-guudP0RkbkbjPLeDPWRb9ryYwG5AAQQR52fhEOpVK-_Nog@mail.gmail.com>

On Wed, Sep 17, 2014 at 1:53 AM, Daniel Prager
<daniel.a.prager at gmail.com> wrote:
>
> [Sorry for drawing you further in.]

:)  (Indeed, my main point is that all that random-number stuff is
foreign to me...)


> My take on your 3 points:
>
> Fisher-Yates is only a few lines, so although not a one-liner, it
> seems reasonable to use it for the better space and time performance.

(It's just time -- the space is the roughly same for both, since sorting
will create an intermediate vector too.)

> I agree that an inside-out version is more apt.
> On my reading the final point is an issue if (random n) has problems
> like modulo bias, but if that's the case surely it is (random n) that
> needs fixing.

It goes into all kinds of arguments that are exactly in that area that
I'm trying to avoid -- but IIUC, it's hard to avoid module-biases, and
the FY use of `random' is particularly nasty since it uses all modulos
in the range of the number of items.  See also the point that the WP
page makes about the size of the random number generator state, and how
it limits the number of different permutations that can be generated.
It looks to me like Racket's state is made of 6 fixnums => 2^192 states,
which still doesn't cover all of the permutations of 52 cards.  I have
no idea if this is correct, if some bias reduces the number of
permutations, or whether all of this is something to worry about...

-- 
          ((lambda (x) (x x)) (lambda (x) (x x)))          Eli Barzilay:
                    http://barzilay.org/                   Maze is Life!

From daniel.a.prager at gmail.com  Wed Sep 17 08:29:46 2014
From: daniel.a.prager at gmail.com (Daniel Prager)
Date: Wed, 17 Sep 2014 07:29:46 -0500
Subject: [racket] on reversing a list by way of sort
In-Reply-To: <CALO-guudP0RkbkbjPLeDPWRb9ryYwG5AAQQR52fhEOpVK-_Nog@mail.gmail.com>
References: <CAB08-qqQyfysyAr_uzcQ4rajfGJSVOj_06T8SC5pfXhA2apSNQ@mail.gmail.com>
 <541751DE.5050407@cs.umd.edu> <5417547D.3060309@cs.umd.edu>
 <CAL3TdON=6kCKdkqGbvRq-be5K9ewKOvMg1tWsUk79N5UGt971g@mail.gmail.com>
 <5417991F.8020807@mbtype.com> <20140916035049.GS21882@localhost>
 <5417BF33.9070407@mbtype.com>
 <CAL3TdONJYHmKJ8H3AW7ZYhtFfKPAO=npAO=y3z_=zkqay-xZiA@mail.gmail.com>
 <CALO-gus=khdeVX4LY01KkmAX69tcqZDhMc1vbZQyRB4ujE1cgA@mail.gmail.com>
 <CAFKxZVX2rSUROs_iTHOqGur9xJ6BZFkkmsC64V=dAA=drYJZxQ@mail.gmail.com>
 <CALO-guuAm2x6EppHpWD12LifKYTaty3uRajC4ERu8H9deu4uDw@mail.gmail.com>
 <CAFKxZVW+2HqUK_xNQXuGTXSvwnZxerfvWFgebhF1iS3DxHKS5Q@mail.gmail.com>
 <CALO-guudP0RkbkbjPLeDPWRb9ryYwG5AAQQR52fhEOpVK-_Nog@mail.gmail.com>
Message-ID: <CAFKxZVX2n=pNhHYAbU2V_9A3rBE58e_g5LoVTtUxbn9srv+0uQ@mail.gmail.com>

Modulo bias is easy enough to correct by checking and occasional
re-sampling. It's explained nicely in the following article:
http://zuttobenkyou.wordpress.com/2012/10/18/generating-random-numbers-without-modulo-bias/

Let's *deliberately* generate some modulo bias in Racket by simulating a
low largest random number:

(define (biased-random RAND-MAX n)
  (modulo (random (add1 RAND-MAX)) 3))


(define (check-bias method [n 10000])
  (define sample (for/list ([i n]) (method 4 3)))

  (for ([i (in-range 3)])
    (displayln
     (~a i " " (~a (real->decimal-string
                (* 100.0 (/ (count (? (j) (= i j)) sample) n))
                1) "%")))))

> (check-bias biased-random)
0 38.3%
1 41.0%
2 20.7%

Remove the generated bias by rejecting some "high" values:


(define (unbiased-random RAND-MAX n)
  (define excess (add1 (modulo RAND-MAX n)))
  (define limit (- RAND-MAX excess))
  (define (in-limit m)
    (if (<= m limit) m (in-limit (random (add1 RAND-MAX)))))
  (modulo (in-limit (random (add1 RAND-MAX))) n))


> (check-bias unbiased-random)
0 33.9%
1 33.8%
2 32.4%

Conclusion: We should be ok regarding modulo bias provided (random n) uses
a rejection strategy along these lines. If not, it's quite fixable.

Dan

On Wed, Sep 17, 2014 at 1:44 AM, Eli Barzilay <eli at barzilay.org> wrote:

> On Wed, Sep 17, 2014 at 1:53 AM, Daniel Prager
> <daniel.a.prager at gmail.com> wrote:
> >
> > [Sorry for drawing you further in.]
>
> :)  (Indeed, my main point is that all that random-number stuff is
> foreign to me...)
>
>
> > My take on your 3 points:
> >
> > Fisher-Yates is only a few lines, so although not a one-liner, it
> > seems reasonable to use it for the better space and time performance.
>
> (It's just time -- the space is the roughly same for both, since sorting
> will create an intermediate vector too.)
>
> > I agree that an inside-out version is more apt.
> > On my reading the final point is an issue if (random n) has problems
> > like modulo bias, but if that's the case surely it is (random n) that
> > needs fixing.
>
> It goes into all kinds of arguments that are exactly in that area that
> I'm trying to avoid -- but IIUC, it's hard to avoid module-biases, and
> the FY use of `random' is particularly nasty since it uses all modulos
> in the range of the number of items.  See also the point that the WP
> page makes about the size of the random number generator state, and how
> it limits the number of different permutations that can be generated.
> It looks to me like Racket's state is made of 6 fixnums => 2^192 states,
> which still doesn't cover all of the permutations of 52 cards.  I have
> no idea if this is correct, if some bias reduces the number of
> permutations, or whether all of this is something to worry about...
>
> --
>           ((lambda (x) (x x)) (lambda (x) (x x)))          Eli Barzilay:
>                     http://barzilay.org/                   Maze is Life!
>



-- 
*Daniel Prager*
Agile/Lean Coaching, Software Development and Leadership
Startup: www.youpatch.com
Twitter: @agilejitsu <https://twitter.com/agilejitsu>
Blog: agile-jitsu.blogspot.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140917/2eb4d239/attachment-0001.html>

From robby at eecs.northwestern.edu  Wed Sep 17 10:14:46 2014
From: robby at eecs.northwestern.edu (Robby Findler)
Date: Wed, 17 Sep 2014 09:14:46 -0500
Subject: [racket] on reversing a list by way of sort
In-Reply-To: <CAFKxZVX2n=pNhHYAbU2V_9A3rBE58e_g5LoVTtUxbn9srv+0uQ@mail.gmail.com>
References: <CAB08-qqQyfysyAr_uzcQ4rajfGJSVOj_06T8SC5pfXhA2apSNQ@mail.gmail.com>
 <541751DE.5050407@cs.umd.edu> <5417547D.3060309@cs.umd.edu>
 <CAL3TdON=6kCKdkqGbvRq-be5K9ewKOvMg1tWsUk79N5UGt971g@mail.gmail.com>
 <5417991F.8020807@mbtype.com> <20140916035049.GS21882@localhost>
 <5417BF33.9070407@mbtype.com>
 <CAL3TdONJYHmKJ8H3AW7ZYhtFfKPAO=npAO=y3z_=zkqay-xZiA@mail.gmail.com>
 <CALO-gus=khdeVX4LY01KkmAX69tcqZDhMc1vbZQyRB4ujE1cgA@mail.gmail.com>
 <CAFKxZVX2rSUROs_iTHOqGur9xJ6BZFkkmsC64V=dAA=drYJZxQ@mail.gmail.com>
 <CALO-guuAm2x6EppHpWD12LifKYTaty3uRajC4ERu8H9deu4uDw@mail.gmail.com>
 <CAFKxZVW+2HqUK_xNQXuGTXSvwnZxerfvWFgebhF1iS3DxHKS5Q@mail.gmail.com>
 <CALO-guudP0RkbkbjPLeDPWRb9ryYwG5AAQQR52fhEOpVK-_Nog@mail.gmail.com>
 <CAFKxZVX2n=pNhHYAbU2V_9A3rBE58e_g5LoVTtUxbn9srv+0uQ@mail.gmail.com>
Message-ID: <CAL3TdOMkAFWTD0BoJ8=6_QkgqRM4GAis_xP3cGtRYjE_v18m9g@mail.gmail.com>

There was a bunch of work done to improve the random number generation
algorithms used internally a while back, but if you wanted to have a
look at them again, that would be most welcome.

Here is a at least part of the code:

https://github.com/plt/racket/blob/master/racket/src/racket/src/random.inc

Robby

On Wed, Sep 17, 2014 at 7:29 AM, Daniel Prager
<daniel.a.prager at gmail.com> wrote:
> Modulo bias is easy enough to correct by checking and occasional
> re-sampling. It's explained nicely in the following article:
> http://zuttobenkyou.wordpress.com/2012/10/18/generating-random-numbers-without-modulo-bias/
>
> Let's *deliberately* generate some modulo bias in Racket by simulating a low
> largest random number:
>
> (define (biased-random RAND-MAX n)
>   (modulo (random (add1 RAND-MAX)) 3))
>
>
> (define (check-bias method [n 10000])
>   (define sample (for/list ([i n]) (method 4 3)))
>
>   (for ([i (in-range 3)])
>     (displayln
>      (~a i " " (~a (real->decimal-string
>                 (* 100.0 (/ (count (? (j) (= i j)) sample) n))
>                 1) "%")))))
>
>> (check-bias biased-random)
> 0 38.3%
> 1 41.0%
> 2 20.7%
>
> Remove the generated bias by rejecting some "high" values:
>
>
> (define (unbiased-random RAND-MAX n)
>   (define excess (add1 (modulo RAND-MAX n)))
>   (define limit (- RAND-MAX excess))
>   (define (in-limit m)
>     (if (<= m limit) m (in-limit (random (add1 RAND-MAX)))))
>   (modulo (in-limit (random (add1 RAND-MAX))) n))
>
>
>> (check-bias unbiased-random)
> 0 33.9%
> 1 33.8%
> 2 32.4%
>
> Conclusion: We should be ok regarding modulo bias provided (random n) uses a
> rejection strategy along these lines. If not, it's quite fixable.
>
> Dan
>
> On Wed, Sep 17, 2014 at 1:44 AM, Eli Barzilay <eli at barzilay.org> wrote:
>>
>> On Wed, Sep 17, 2014 at 1:53 AM, Daniel Prager
>> <daniel.a.prager at gmail.com> wrote:
>> >
>> > [Sorry for drawing you further in.]
>>
>> :)  (Indeed, my main point is that all that random-number stuff is
>> foreign to me...)
>>
>>
>> > My take on your 3 points:
>> >
>> > Fisher-Yates is only a few lines, so although not a one-liner, it
>> > seems reasonable to use it for the better space and time performance.
>>
>> (It's just time -- the space is the roughly same for both, since sorting
>> will create an intermediate vector too.)
>>
>> > I agree that an inside-out version is more apt.
>> > On my reading the final point is an issue if (random n) has problems
>> > like modulo bias, but if that's the case surely it is (random n) that
>> > needs fixing.
>>
>> It goes into all kinds of arguments that are exactly in that area that
>> I'm trying to avoid -- but IIUC, it's hard to avoid module-biases, and
>> the FY use of `random' is particularly nasty since it uses all modulos
>> in the range of the number of items.  See also the point that the WP
>> page makes about the size of the random number generator state, and how
>> it limits the number of different permutations that can be generated.
>> It looks to me like Racket's state is made of 6 fixnums => 2^192 states,
>> which still doesn't cover all of the permutations of 52 cards.  I have
>> no idea if this is correct, if some bias reduces the number of
>> permutations, or whether all of this is something to worry about...
>>
>> --
>>           ((lambda (x) (x x)) (lambda (x) (x x)))          Eli Barzilay:
>>                     http://barzilay.org/                   Maze is Life!
>
>
>
>
> --
> Daniel Prager
> Agile/Lean Coaching, Software Development and Leadership
> Startup: www.youpatch.com
> Twitter: @agilejitsu
> Blog: agile-jitsu.blogspot.com
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>


From samth at cs.indiana.edu  Wed Sep 17 10:51:19 2014
From: samth at cs.indiana.edu (Sam Tobin-Hochstadt)
Date: Wed, 17 Sep 2014 10:51:19 -0400
Subject: [racket] canonical index of Racket courses? [was: Summer
 programs learning Racket for a student]
In-Reply-To: <CADBC08E-D847-4AAE-87BE-D62368E91936@ccs.neu.edu>
References: <CANwREeXB1EWu2SjHBHuzhJKba4dHJFhdGHdg7SXafJqJ=v-nsw@mail.gmail.com>
 <CANwREeXVeAtEirzyXOGQedt8uU4EK--qYU--ZkNRGKWQ5hn8Ag@mail.gmail.com>
 <CADBC08E-D847-4AAE-87BE-D62368E91936@ccs.neu.edu>
Message-ID: <CAK=HD+Yj==FD+N01Dr_hAzGfm+jieKO9k=F=2VnRpg0opgmV5Q@mail.gmail.com>

I've now created a wiki page for this, with some initial content:
https://github.com/plt/racket/wiki/Courses-using-Racket

If you teach (or know of) a course using Racket/HtDP/etc, please add it there.

Sam

On Tue, Sep 16, 2014 at 8:43 PM, Matthias Felleisen
<matthias at ccs.neu.edu> wrote:
>
> You're right, we failed to follow up on the general part of your email.
>
> Do you imagine listing college courses such as Brown's 17, which uses
> DrRacket and the teaching languages? Or do you want Coursera courses that
> everyone can access? Do you want it all? now? :-)
>
> And yes, we should probably create a wiki like thing for such an effort or
> perhaps something like packages.racket-lang.org. I'll bring it up as we meet
> in St Louis
>
> -- Matthias
>
>
>
>
>
> On Sep 16, 2014, at 7:06 PM, jab at math.brown.edu wrote:
>
> Dear Racket Users,
>
>
> On Wed, Apr 16, 2014 at 4:45 PM, <jab at math.brown.edu> wrote:
>>
>> I recently asked here about in-person Racket courses available this summer
>> which I could recommend to a student I've been tutoring. (Thanks again to
>> everyone who let me know about their courses so far.)
>>
>> To make this info easier to find, I'm wondering if the racket-lang.org
>> maintainers would be interested in publishing a canonical index of Racket
>> courses (both online and in-person) linked off the "Community" or "Learning"
>> sections on the homepage, and inviting the community to add to it.
>>
>> (Similarly, linking to a canonical index of Racket meetup groups (? la
>> http://clojure.meetup.com/ and https://wiki.python.org/moin/LocalUserGroups)
>> would be great too. Google turned up http://racket.meetup.com/ but 2 of the
>> 3 results there are for racket sports.)
>>
>> Thanks, and looking forward to knowing how to become better plugged into
>> the Racket community!
>
>
>
> Didn't get any reply to this last April, so just thought I'd try again. In
> case anyone who'd want to reply just missed it! Wish I could join those of
> you in St. Louis this week.
> ____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users
>
>
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>


From kalimehtar at mail.ru  Wed Sep 17 12:59:15 2014
From: kalimehtar at mail.ru (=?UTF-8?B?Um9tYW4gS2xvY2hrb3Y=?=)
Date: Wed, 17 Sep 2014 20:59:15 +0400
Subject: [racket] =?utf-8?q?Exception_decorator=2E_is_it_possible=3F?=
Message-ID: <1410973155.471904896@f186.i.mail.ru>

 I'm writeing a parser.
So I want to see in parser exception what happend and when (not only where).

(define (parse-item item)
? ?(with-handlers ([exn? (lambda (e) (raise (struct-copy exn e [message (format "At item ~a\n~a" item (exn-message e))])))])
? ? ? ? ?....))

But this way I lose all exception types. Is there a way to change exception message, saving its type, except long copy/paste like
(with-handlers ([exn:fail:contract:arity? (raise (struct-copy exn:fail:contract:arity ...))]
? ? ? ? ? ? ? ? ? ? ? ? ?[exn:fil:contract? (raise?(struct-copy exn:fail:contract ...))]
? ? ? ? ? ? ? ? ? ? ? ? ?...) 
? ?...)

?

-- 
Roman Klochkov
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140917/de704e03/attachment.html>

From jensaxel at soegaard.net  Wed Sep 17 13:37:47 2014
From: jensaxel at soegaard.net (=?UTF-8?Q?Jens_Axel_S=C3=B8gaard?=)
Date: Wed, 17 Sep 2014 19:37:47 +0200
Subject: [racket] Spreadsheet editor
In-Reply-To: <6bc0c4456cc37557ac97f5e8ea2bed2f@ipa.nw.ru>
References: <6bc0c4456cc37557ac97f5e8ea2bed2f@ipa.nw.ru>
Message-ID: <CABefVgw-3v=0M_mq-Rjuh9CXwF-boiwBvtgA9Lo3id6FDzHdpw@mail.gmail.com>

Nice work!

> I see two issues currently:
>
> * During editing, text snip creates a white plate that
>   extends outside the supposed borders of the snip.
>   See [1] for details.
>
> * Racket GUI does not provide scroll bars that one can
>   use independently from canvas or panel. I currently
>   use sliders instead of scroll bars, but with slider,
>   it is impossible to go up/down by exactly one point.
>   Also, it is not possible to update the limits of the
>   slider, so I have to re-create it when I add or remove
>   a row from the table.
>
> I would love to hear any comments regarding those issues,
> or any other issues, or anything that draws your
> attention in the source code.

The first issue is very apparent. I am

--
Jens Axel S?gaard


From m4burns at uwaterloo.ca  Wed Sep 17 14:15:17 2014
From: m4burns at uwaterloo.ca (Marc Burns)
Date: Wed, 17 Sep 2014 14:15:17 -0400
Subject: [racket] Exception decorator. is it possible?
In-Reply-To: <1410973155.471904896@f186.i.mail.ru>
References: <1410973155.471904896@f186.i.mail.ru>
Message-ID: <20140917181517.GA18713@happierface>

It's not generally possible to edit the message contained in an
exception. However, you can add information to the program's stack using
continuation marks. Here's an example:

(define my-cont-mark-key (gensym))

(define (parse-item item)
  (with-continuation-mark my-cont-mark-key item
    ...))

(with-handlers
  ([exn? (lambda(e)
           ((error-display-handler)
            (format "Parser exception at item ~a:\n~a"
                    (first
                      (continuation-mark-set->list
                        (exn-continuation-marks e)
                        my-cont-mark-key))
                    (exn-message e))
            e))])
  (parse-item start-item))

The exception captures the continuation marks installed when it is
thrown. You can extract the information you added to the stack with 
these marks in the exception handler.

On Wed, Sep 17, 2014 at 08:59:15PM +0400, Roman Klochkov wrote:
>  I'm writeing a parser.
> So I want to see in parser exception what happend and when (not only where).
> 
> (define (parse-item item)
> ?? ??(with-handlers ([exn? (lambda (e) (raise (struct-copy exn e [message (format "At item ~a\n~a" item (exn-message e))])))])
> ?? ?? ?? ?? ??....))
> 
> But this way I lose all exception types. Is there a way to change exception message, saving its type, except long copy/paste like
> (with-handlers ([exn:fail:contract:arity? (raise (struct-copy exn:fail:contract:arity ...))]
> ?? ?? ?? ?? ?? ?? ?? ?? ?? ?? ?? ?? ??[exn:fil:contract? (raise??(struct-copy exn:fail:contract ...))]
> ?? ?? ?? ?? ?? ?? ?? ?? ?? ?? ?? ?? ??...) 
> ?? ??...)
> 
> ?
> 
> -- 
> Roman Klochkov

> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users


From dvanhorn at cs.umd.edu  Wed Sep 17 16:41:52 2014
From: dvanhorn at cs.umd.edu (David Van Horn)
Date: Wed, 17 Sep 2014 16:41:52 -0400
Subject: [racket] frtime
Message-ID: <5419F210.3090606@cs.umd.edu>

This program is less flashy than it used to be:

   #lang frtime
   seconds

Is FrTime still supported?  Is there a way to get the old behavior of
being, well, reactive?

David

From dpavlov at ipa.nw.ru  Wed Sep 17 16:48:47 2014
From: dpavlov at ipa.nw.ru (dpavlov at ipa.nw.ru)
Date: Thu, 18 Sep 2014 00:48:47 +0400
Subject: [racket] Spreadsheet editor
In-Reply-To: <CABefVgw-3v=0M_mq-Rjuh9CXwF-boiwBvtgA9Lo3id6FDzHdpw@mail.gmail.com>
References: <6bc0c4456cc37557ac97f5e8ea2bed2f@ipa.nw.ru>
 <CABefVgw-3v=0M_mq-Rjuh9CXwF-boiwBvtgA9Lo3id6FDzHdpw@mail.gmail.com>
Message-ID: <7e6737e7c1eb72dc9801e548b44ab112@ipa.nw.ru>

Jens,

> Nice work!

Thanks!

> The first issue is very apparent. I am

You most probably sent an unfinished message, did you notice that?


Regards,

Dmitry


From jensaxel at soegaard.net  Wed Sep 17 16:51:07 2014
From: jensaxel at soegaard.net (=?UTF-8?Q?Jens_Axel_S=C3=B8gaard?=)
Date: Wed, 17 Sep 2014 22:51:07 +0200
Subject: [racket] Spreadsheet editor
In-Reply-To: <7e6737e7c1eb72dc9801e548b44ab112@ipa.nw.ru>
References: <6bc0c4456cc37557ac97f5e8ea2bed2f@ipa.nw.ru>
 <CABefVgw-3v=0M_mq-Rjuh9CXwF-boiwBvtgA9Lo3id6FDzHdpw@mail.gmail.com>
 <7e6737e7c1eb72dc9801e548b44ab112@ipa.nw.ru>
Message-ID: <CABefVgz8iNFnsfWPR2kns05q8Sh3LAgNmCORZo++81wg1RU53Q@mail.gmail.com>

Hi,

Gmail had a temporary problem.

It should have said, that I don't know of way to fix it.

/Jens Axel


2014-09-17 22:48 GMT+02:00  <dpavlov at ipa.nw.ru>:
> Jens,
>
>> Nice work!
>
>
> Thanks!
>
>> The first issue is very apparent. I am
>
>
> You most probably sent an unfinished message, did you notice that?
>
>
> Regards,
>
> Dmitry
>



-- 
--
Jens Axel S?gaard


From jay.mccarthy at gmail.com  Wed Sep 17 17:06:27 2014
From: jay.mccarthy at gmail.com (Jay McCarthy)
Date: Wed, 17 Sep 2014 17:06:27 -0400
Subject: [racket] frtime
In-Reply-To: <5419F210.3090606@cs.umd.edu>
References: <5419F210.3090606@cs.umd.edu>
Message-ID: <CAJYbDakfB_O9zqN5STKpRxa5rx0zr09HEyxKhRmqFm1i2jSv9Q@mail.gmail.com>

That program never did what you were hoping for.

The program that we demo had you choose "FrTime" from the "Choose a
Language..." dialog.

The difference is that the language level can include its snip code
inside DrRacket and get its GUI to update. #langs cannot do that at
all. Robby and I have talked many times about how to allow it, but it
is unclear how to do it safely.

As an aside, the real demo is currently broken for a different reason,
that I think is related to the undefined change. I'll look into it.

Jay

On Wed, Sep 17, 2014 at 4:41 PM, David Van Horn <dvanhorn at cs.umd.edu> wrote:
> This program is less flashy than it used to be:
>
>    #lang frtime
>    seconds
>
> Is FrTime still supported?  Is there a way to get the old behavior of
> being, well, reactive?
>
> David
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users



-- 
Jay McCarthy
http://jeapostrophe.github.io

           "Wherefore, be not weary in well-doing,
      for ye are laying the foundation of a great work.
And out of small things proceedeth that which is great."
                          - D&C 64:33

From matthias at ccs.neu.edu  Wed Sep 17 14:49:41 2014
From: matthias at ccs.neu.edu (Matthias Felleisen)
Date: Wed, 17 Sep 2014 14:49:41 -0400
Subject: [racket] Racket guide
In-Reply-To: <CAFHZvE8gy7SGOUevg1bXpnM_qpD4K3cSwwK+VPjVA15GdHcbZw@mail.gmail.com>
References: <CAFHZvE8gy7SGOUevg1bXpnM_qpD4K3cSwwK+VPjVA15GdHcbZw@mail.gmail.com>
Message-ID: <E0AE9207-4A98-4579-966F-91A0AFF3EF38@ccs.neu.edu>


On Sep 17, 2014, at 2:04 AM, mukesh tiwari wrote:

> Dear Racket Users,
> I am learning racket and I found little bit misleading statement on guide [1]. It says that "Both the my-length and my-map functions run in O(n) time for a list of length n.". Technically this statement is correct but when I read below
> 
> "For a list with n elements, evaluation will stack up n (+ 1 ...) additions, and then finally add them up when the list is exhausted.
> 
> You can avoid piling up additions by adding along the way. To accumulate a length this way, we need a function that takes both a list and the length of the list seen so far; the code below uses a local function iter that accumulates the length in an argument len".
> 
> so just curious, shouldn't it be "Both the my-length and my-map functions run in O(n) space for a list of length n." ?
> 

That's correct for straightforward versions of these functions. -- Matthias

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140917/657aa9d9/attachment.html>

From daniel.a.prager at gmail.com  Wed Sep 17 23:38:33 2014
From: daniel.a.prager at gmail.com (Daniel Prager)
Date: Wed, 17 Sep 2014 22:38:33 -0500
Subject: [racket] on reversing a list by way of sort
In-Reply-To: <CAL3TdOMkAFWTD0BoJ8=6_QkgqRM4GAis_xP3cGtRYjE_v18m9g@mail.gmail.com>
References: <CAB08-qqQyfysyAr_uzcQ4rajfGJSVOj_06T8SC5pfXhA2apSNQ@mail.gmail.com>
 <541751DE.5050407@cs.umd.edu> <5417547D.3060309@cs.umd.edu>
 <CAL3TdON=6kCKdkqGbvRq-be5K9ewKOvMg1tWsUk79N5UGt971g@mail.gmail.com>
 <5417991F.8020807@mbtype.com> <20140916035049.GS21882@localhost>
 <5417BF33.9070407@mbtype.com>
 <CAL3TdONJYHmKJ8H3AW7ZYhtFfKPAO=npAO=y3z_=zkqay-xZiA@mail.gmail.com>
 <CALO-gus=khdeVX4LY01KkmAX69tcqZDhMc1vbZQyRB4ujE1cgA@mail.gmail.com>
 <CAFKxZVX2rSUROs_iTHOqGur9xJ6BZFkkmsC64V=dAA=drYJZxQ@mail.gmail.com>
 <CALO-guuAm2x6EppHpWD12LifKYTaty3uRajC4ERu8H9deu4uDw@mail.gmail.com>
 <CAFKxZVW+2HqUK_xNQXuGTXSvwnZxerfvWFgebhF1iS3DxHKS5Q@mail.gmail.com>
 <CALO-guudP0RkbkbjPLeDPWRb9ryYwG5AAQQR52fhEOpVK-_Nog@mail.gmail.com>
 <CAFKxZVX2n=pNhHYAbU2V_9A3rBE58e_g5LoVTtUxbn9srv+0uQ@mail.gmail.com>
 <CAL3TdOMkAFWTD0BoJ8=6_QkgqRM4GAis_xP3cGtRYjE_v18m9g@mail.gmail.com>
Message-ID: <CAFKxZVVkfrytcaAKCc++iaZWYAp1GN6EB=AUAqBygsikYfJPgA@mail.gmail.com>

I've done some testing of the built-in (random n) and I haven't been able
to detect evidence of modulo bias. Of course absence of evidence isn't
evidence of absence!

More generally though, this is an interesting case (at least to me) of the
trade-offs between efficiency, clarity, concision, cost of learning about a
specialised area as well as arguments from theory vs empirical assessment.

Dan

On Wed, Sep 17, 2014 at 9:14 AM, Robby Findler <robby at eecs.northwestern.edu>
wrote:

> There was a bunch of work done to improve the random number generation
> algorithms used internally a while back, but if you wanted to have a
> look at them again, that would be most welcome.
>
> Here is a at least part of the code:
>
> https://github.com/plt/racket/blob/master/racket/src/racket/src/random.inc
>
> Robby
>
> On Wed, Sep 17, 2014 at 7:29 AM, Daniel Prager
> <daniel.a.prager at gmail.com> wrote:
> > Modulo bias is easy enough to correct by checking and occasional
> > re-sampling. It's explained nicely in the following article:
> >
> http://zuttobenkyou.wordpress.com/2012/10/18/generating-random-numbers-without-modulo-bias/
> >
> > Let's *deliberately* generate some modulo bias in Racket by simulating a
> low
> > largest random number:
> >
> > (define (biased-random RAND-MAX n)
> >   (modulo (random (add1 RAND-MAX)) 3))
> >
> >
> > (define (check-bias method [n 10000])
> >   (define sample (for/list ([i n]) (method 4 3)))
> >
> >   (for ([i (in-range 3)])
> >     (displayln
> >      (~a i " " (~a (real->decimal-string
> >                 (* 100.0 (/ (count (? (j) (= i j)) sample) n))
> >                 1) "%")))))
> >
> >> (check-bias biased-random)
> > 0 38.3%
> > 1 41.0%
> > 2 20.7%
> >
> > Remove the generated bias by rejecting some "high" values:
> >
> >
> > (define (unbiased-random RAND-MAX n)
> >   (define excess (add1 (modulo RAND-MAX n)))
> >   (define limit (- RAND-MAX excess))
> >   (define (in-limit m)
> >     (if (<= m limit) m (in-limit (random (add1 RAND-MAX)))))
> >   (modulo (in-limit (random (add1 RAND-MAX))) n))
> >
> >
> >> (check-bias unbiased-random)
> > 0 33.9%
> > 1 33.8%
> > 2 32.4%
> >
> > Conclusion: We should be ok regarding modulo bias provided (random n)
> uses a
> > rejection strategy along these lines. If not, it's quite fixable.
> >
> > Dan
> >
> > On Wed, Sep 17, 2014 at 1:44 AM, Eli Barzilay <eli at barzilay.org> wrote:
> >>
> >> On Wed, Sep 17, 2014 at 1:53 AM, Daniel Prager
> >> <daniel.a.prager at gmail.com> wrote:
> >> >
> >> > [Sorry for drawing you further in.]
> >>
> >> :)  (Indeed, my main point is that all that random-number stuff is
> >> foreign to me...)
> >>
> >>
> >> > My take on your 3 points:
> >> >
> >> > Fisher-Yates is only a few lines, so although not a one-liner, it
> >> > seems reasonable to use it for the better space and time performance.
> >>
> >> (It's just time -- the space is the roughly same for both, since sorting
> >> will create an intermediate vector too.)
> >>
> >> > I agree that an inside-out version is more apt.
> >> > On my reading the final point is an issue if (random n) has problems
> >> > like modulo bias, but if that's the case surely it is (random n) that
> >> > needs fixing.
> >>
> >> It goes into all kinds of arguments that are exactly in that area that
> >> I'm trying to avoid -- but IIUC, it's hard to avoid module-biases, and
> >> the FY use of `random' is particularly nasty since it uses all modulos
> >> in the range of the number of items.  See also the point that the WP
> >> page makes about the size of the random number generator state, and how
> >> it limits the number of different permutations that can be generated.
> >> It looks to me like Racket's state is made of 6 fixnums => 2^192 states,
> >> which still doesn't cover all of the permutations of 52 cards.  I have
> >> no idea if this is correct, if some bias reduces the number of
> >> permutations, or whether all of this is something to worry about...
> >>
> >> --
> >>           ((lambda (x) (x x)) (lambda (x) (x x)))          Eli Barzilay:
> >>                     http://barzilay.org/                   Maze is
> Life!
> >
> >
> >
> >
> > --
> > Daniel Prager
> > Agile/Lean Coaching, Software Development and Leadership
> > Startup: www.youpatch.com
> > Twitter: @agilejitsu
> > Blog: agile-jitsu.blogspot.com
> >
> > ____________________
> >   Racket Users list:
> >   http://lists.racket-lang.org/users
> >
>



-- 
*Daniel Prager*
Agile/Lean Coaching, Software Development and Leadership
Startup: www.youpatch.com
Twitter: @agilejitsu <https://twitter.com/agilejitsu>
Blog: agile-jitsu.blogspot.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140917/2c30d3b8/attachment.html>

From kalimehtar at mail.ru  Thu Sep 18 01:19:40 2014
From: kalimehtar at mail.ru (=?UTF-8?B?Um9tYW4gS2xvY2hrb3Y=?=)
Date: Thu, 18 Sep 2014 09:19:40 +0400
Subject: [racket] =?utf-8?q?Exception_decorator=2E_is_it_possible=3F?=
In-Reply-To: <20140917181517.GA18713@happierface>
References: <1410973155.471904896@f186.i.mail.ru>
 <20140917181517.GA18713@happierface>
Message-ID: <1411017579.96049017@f396.i.mail.ru>

 Thank you. Yet one question^ is it possible to control, what will be in stacktrace.

Because, I have (macro generated)?
(class base object%
? ?(define/public (write stream)
? ? ? ?(write-value l1 stream field1)
? ? ? ?(write-value l1 stream field2)
? ? ? ?(write-value l1 stream field3)
? ? ? ?...))

Where l1 -- function, that takes its param and it to bytestream.

(class derived base
? (define/override (write stream)
? ? ?(super write stream)))

Then I have in stacktrace?(super write stream), and then instantly body of l1. But no string "(write-value l1 stream field2)" where I could see, what field caouse error.
Is it possible to add it to stacktrace? Or how Racket decides, what expressions will be in stacktrace?


Wed, 17 Sep 2014 14:15:17 -0400 ?? Marc Burns <m4burns at uwaterloo.ca>:
>It's not generally possible to edit the message contained in an
>exception. However, you can add information to the program's stack using
>continuation marks. Here's an example:
>
>(define my-cont-mark-key (gensym))
>
>(define (parse-item item)
>??(with-continuation-mark my-cont-mark-key item
>????...))
>
>(with-handlers
>??([exn? (lambda(e)
>???????????((error-display-handler)
>????????????(format "Parser exception at item ~a:\n~a"
>????????????????????(first
>??????????????????????(continuation-mark-set->list
>????????????????????????(exn-continuation-marks e)
>????????????????????????my-cont-mark-key))
>????????????????????(exn-message e))
>????????????e))])
>??(parse-item start-item))
>
>The exception captures the continuation marks installed when it is
>thrown. You can extract the information you added to the stack with 
>these marks in the exception handler.
>
>On Wed, Sep 17, 2014 at 08:59:15PM +0400, Roman Klochkov wrote:
>>  I'm writeing a parser.
>> So I want to see in parser exception what happend and when (not only where).
>> 
>> (define (parse-item item)
>> ?? ??(with-handlers ([exn? (lambda (e) (raise (struct-copy exn e [message (format "At item ~a\n~a" item (exn-message e))])))])
>> ?? ?? ?? ?? ??....))
>> 
>> But this way I lose all exception types. Is there a way to change exception message, saving its type, except long copy/paste like
>> (with-handlers ([exn:fail:contract:arity? (raise (struct-copy exn:fail:contract:arity ...))]
>> ?? ?? ?? ?? ?? ?? ?? ?? ?? ?? ?? ?? ??[exn:fil:contract? (raise??(struct-copy exn:fail:contract ...))]
>> ?? ?? ?? ?? ?? ?? ?? ?? ?? ?? ?? ?? ??...) 
>> ?? ??...)
>> 
>> ?
>> 
>> -- 
>> Roman Klochkov
>
>> ____________________
>>   Racket Users list:
>>  http://lists.racket-lang.org/users
>


-- 
Roman Klochkov
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140918/bbaa0651/attachment-0001.html>

From mflatt at cs.utah.edu  Thu Sep 18 08:18:15 2014
From: mflatt at cs.utah.edu (Matthew Flatt)
Date: Thu, 18 Sep 2014 07:18:15 -0500
Subject: [racket] Text snip bleaches canvas space that does not belong
	to it
In-Reply-To: <540DF2A3.5020707@ipa.nw.ru>
References: <540DF2A3.5020707@ipa.nw.ru>
Message-ID: <20140918121816.EAF12650194@mail-svr1.cs.utah.edu>

The drawing model for editors is that an editor is allowed to draw
directly to the canvas, without going through `on-paint`, and typically
updating only a part of the canvas.

To draw a background behind the editor content, the intent is that you
use `on-paint` at the editor level (i.e., in `pasteboard%`). It's
called both before and after the editor content is drawn.

At Mon, 08 Sep 2014 22:17:07 +0400, Dmitry Pavlov wrote:
> Hello,
> 
> I need a text snip on a canvas that paints itself strictly in some defined WxH 
> bounds.
> I need no interactive moving or resizing, only interactive editing.
> Here is a code I have come up with, which also displays a window with the snip 
> and a grid.
> The grid facilitates the visual check of the text snip bounds.
> 
> (require racket/gui
>           racket/draw)
> 
> (define my-pasteboard%
>    (class pasteboard%
>      (super-new)
>      (define/augment (can-select? snip on?)
>        #f)
>      (define/augment (can-interactive-move? snip)
>        #f)
>      (define/augment (can-interactive-resize? snip)
>        #f)))
> 
> (define frame (new frame% (label "Test")))
> (define pasteboard (new my-pasteboard%))
> 
> (define my-canvas%
>    (class editor-canvas%
>      (super-new)
>      (define/override (on-paint)
>        (super on-paint)
>        (define dc (send this get-dc))
>        ;; Draw a 10x10 grid with cell size 20x20 pixels
>        (for ((x (in-range 10))) (send dc draw-line (* x 20) 0 (* x 20) 200))
>        (for ((y (in-range 10))) (send dc draw-line 0 (* y 20)  200 (* y 20)))
>        )))
> 
> (define editor-canvas
>    (new my-canvas%
>         (editor pasteboard)
>         (vertical-inset 0) (horizontal-inset 0)
>         (parent frame)
>         (style '(no-border no-hscroll no-vscroll))))
> 
> (send frame resize 200 200)
> (send frame show #t)
> 
> (define text-obj (new text%))
> (send text-obj insert "Hello" 0)
> (define text-snip (new editor-snip%
>                       (editor text-obj)
>                       (left-margin 0) (right-margin 0) (top-margin 0) 
> (bottom-margin 0)
>                       (left-inset 0) (right-inset 0) (top-inset 0) 
> (bottom-inset 0)
>                       (with-border? #f)
>                       (min-width 40)
>                       (max-width 40)
>                       (min-height 20)
>                       (max-height 20)))
> 
> (send pasteboard insert text-snip 40 60)
> (send pasteboard set-caret-owner text-snip)
> 
> 
> When I click on the snip in the window and try to add a "!" character after
> the "Hello" text, I get this:
> 
> http://imgur.com/cikwlqh
> 
> It is obvious that the text snip (or something associated with it)
> expanded its supposed area of rendering by 3-4 pixels in all directions.
> And if I do silly things, like pressing "Enter", things get even worse.
> Can that be avoided somehow?
> 
> 
> Best regards,
> 
> Dmitry
> 
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users

From dpavlov at ipa.nw.ru  Thu Sep 18 11:22:40 2014
From: dpavlov at ipa.nw.ru (Dmitry Pavlov)
Date: Thu, 18 Sep 2014 19:22:40 +0400
Subject: [racket] slider initial value is unexpectedly zero when min-value =
	max-value
Message-ID: <541AF8C0.2040901@ipa.nw.ru>

Hello,

The title says it all, but here is also a code snip:


(require racket/gui)

(define frame (new frame% (label "test")))

(define slider
   (new slider%
        (label "")
        (parent frame)
        (min-value 1)
        (max-value 1)))

(display (send slider get-value))


It prints "0", and this is a sign of violation
of the invariant min-value <= init-value <= max-value.
Passing (init-value 1) does not help.


Best regards,

Dmitry

From dpavlov at ipa.nw.ru  Thu Sep 18 11:25:13 2014
From: dpavlov at ipa.nw.ru (Dmitry Pavlov)
Date: Thu, 18 Sep 2014 19:25:13 +0400
Subject: [racket] Spreadsheet editor
In-Reply-To: <CABefVgz8iNFnsfWPR2kns05q8Sh3LAgNmCORZo++81wg1RU53Q@mail.gmail.com>
References: <6bc0c4456cc37557ac97f5e8ea2bed2f@ipa.nw.ru>	<CABefVgw-3v=0M_mq-Rjuh9CXwF-boiwBvtgA9Lo3id6FDzHdpw@mail.gmail.com>	<7e6737e7c1eb72dc9801e548b44ab112@ipa.nw.ru>
 <CABefVgz8iNFnsfWPR2kns05q8Sh3LAgNmCORZo++81wg1RU53Q@mail.gmail.com>
Message-ID: <541AF959.7050800@ipa.nw.ru>


> It should have said, that I don't know of way to fix it.

I just fixed it and committed to github.

Matthew, thank you for the answer which you gave in the original thread!



Regards,

Dmitry


From alexander at knauth.org  Thu Sep 18 16:37:10 2014
From: alexander at knauth.org (Alexander D. Knauth)
Date: Thu, 18 Sep 2014 16:37:10 -0400
Subject: [racket] filters and the type of partition
Message-ID: <F502A00D-E55B-4B57-A99E-8F915BD6DD81@knauth.org>

Is there any way for a predicate-ish thing to do something like this:
#lang typed/racket
(: vectorof-real? : [(U Number (Vectorof Real))
                     -> Boolean :
                     #:+ (Vectorof Real)
                     #:- Number])
(define (vectorof-real? x)
  (vector? x))

Because this works:
(let ([x : (U Number (Vectorof Real)) #(1 2 3)])
  (if (vector? x)
      (ann x (Vectorof Real))
      (ann x Number)))

And also I noticed the types for the functions filter and partition are:
 > filter
 - : (All (a b)
        (case->
        (-> (-> a Any : #:+ b) (Listof a) (Listof b))
        (-> (-> a Any) (Listof a) (Listof a))))
 #<procedure:filter>
 > partition
 - : (All (a b)
       (case->
        (-> (-> a Any) (Listof a) (values (Listof a) (Listof a)))
        (-> (-> Any Boolean : a) (Listof b) (values (Listof a) (Listof b)))))
 #<procedure:partition>

For the filter function, the predicate only has to accept the type a for both cases and can return Any for both cases.  Is there any reason why the second case of partition doesn?t do that?  And is there a reason why it doesn?t use #:+ for the filter?  

The current type for partition wouldn?t allow me to use my vectorof-real? as the predicate even if it did work, right?  But filter would?  




From ameliastahlman at gmail.com  Wed Sep 10 14:40:03 2014
From: ameliastahlman at gmail.com (Amy Stahlman)
Date: Wed, 10 Sep 2014 13:40:03 -0500
Subject: [racket] (no subject)
Message-ID: <CACfnStW_x-DoaasOkqOdsOX0085CZYVLx-B=eQ8uMeN4b5zJkw@mail.gmail.com>

I've been working on a program that uses "init-auto-scrollbars" and I
noticed that when the window is resized it actually changes the size of the
virtual canvas along with the literal canvas instead of just resizing the
scrollbar and changing the size of the literal canvas.

Here's an example that shows this:














*(require racket/gui)(define frame (new frame%                  [label
"Example"]                  [width 300]                  [height
300]))(define canv (new canvas% [parent frame]    [style '(vscroll
hscroll)]            [paint-callback             (lambda (canvas dc)(send
dc draw-ellipse 0 0 500 500))]))(send frame show #t)(send canv
init-auto-scrollbars 500 500 0.0 0.0)*

Resize the window, making it large enough to show the entire ellipse;
notice how the scrollbar's relative size doesn't change, and there is blank
space at the bottom.
Is this by design or some kind of bug?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140910/5373ee66/attachment.html>

From samth at cs.indiana.edu  Fri Sep 19 10:40:05 2014
From: samth at cs.indiana.edu (Sam Tobin-Hochstadt)
Date: Fri, 19 Sep 2014 10:40:05 -0400
Subject: [racket] filters and the type of partition
In-Reply-To: <F502A00D-E55B-4B57-A99E-8F915BD6DD81@knauth.org>
References: <F502A00D-E55B-4B57-A99E-8F915BD6DD81@knauth.org>
Message-ID: <CAK=HD+bjp6tMu729Z8vh=2m4Yr9=nAd_AR+_RpT7LcCmWCVOdA@mail.gmail.com>

I think your first function ought to type check, and that it doesn't is a bug.

Second, `partition` could be relaxed in the way you say, and could use
#:+ as well.

Sam

On Thu, Sep 18, 2014 at 4:37 PM, Alexander D. Knauth
<alexander at knauth.org> wrote:
> Is there any way for a predicate-ish thing to do something like this:
> #lang typed/racket
> (: vectorof-real? : [(U Number (Vectorof Real))
>                      -> Boolean :
>                      #:+ (Vectorof Real)
>                      #:- Number])
> (define (vectorof-real? x)
>   (vector? x))
>
> Because this works:
> (let ([x : (U Number (Vectorof Real)) #(1 2 3)])
>   (if (vector? x)
>       (ann x (Vectorof Real))
>       (ann x Number)))
>
> And also I noticed the types for the functions filter and partition are:
>  > filter
>  - : (All (a b)
>         (case->
>         (-> (-> a Any : #:+ b) (Listof a) (Listof b))
>         (-> (-> a Any) (Listof a) (Listof a))))
>  #<procedure:filter>
>  > partition
>  - : (All (a b)
>        (case->
>         (-> (-> a Any) (Listof a) (values (Listof a) (Listof a)))
>         (-> (-> Any Boolean : a) (Listof b) (values (Listof a) (Listof b)))))
>  #<procedure:partition>
>
> For the filter function, the predicate only has to accept the type a for both cases and can return Any for both cases.  Is there any reason why the second case of partition doesn?t do that?  And is there a reason why it doesn?t use #:+ for the filter?
>
> The current type for partition wouldn?t allow me to use my vectorof-real? as the predicate even if it did work, right?  But filter would?
>
>
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users


From dpavlov at ipa.nw.ru  Fri Sep 19 12:01:26 2014
From: dpavlov at ipa.nw.ru (Dmitry Pavlov)
Date: Fri, 19 Sep 2014 20:01:26 +0400
Subject: [racket] Extending DrRacket with non-text tabs
Message-ID: <541C5356.90502@ipa.nw.ru>

Hello,

I recently wrote a spreadsheet editor in Racket [1].
Now I am considering a crazy idea to use it within
DrRacket natively, not as a separate app.

The most fantastic scenario would be me running
the spreadsheet editor(s) in DrRacket tab(s), started
by opening files with a certain extension via DrRacket's
"Open" dialog. Is that possible? If yes, is it [2] where
I should be looking? Do any manuals or examples exist for that?
Because I hardly understand where to start when it
comes to extending DrRacket.

If the most fantastic scenario is not possible
(I do not really know), maybe I can open my spreadsheet
editor(s) in separate windows by pressing a custom
button on the DrRacket's tool panel, or choosing
a custom menu item? What would you, experienced Racketeers,
do if you were in my place?

Thanks.

[1] https://github.com/kugelblitz/spreadsheet-editor
[2] http://docs.racket-lang.org/tools/drracket_get_extend.html


Regards,

Dmitry

From jackhfirth at gmail.com  Fri Sep 19 17:41:31 2014
From: jackhfirth at gmail.com (Jack Firth)
Date: Fri, 19 Sep 2014 14:41:31 -0700
Subject: [racket] Provide transformers
Message-ID: <CAAXAoJXbO+D0mPzAonTh+uNPGW6ajG0zSyeh8=gsFxVr5OQQyw@mail.gmail.com>

I've been looking into how to design a macro that would specify tests in a
provide transformer, for example something like

    (provide (test-pred-out some-integer integer?))

I'd like that to expand to:

    (provide some-integer)
    (module+ test
      (check-pred integer? some-integer))

>From what I can gather from the reference, normally provide forms are done
with define-provide-syntax, but that macro doesn't let me splice anything
into the module body. The reference mentions provide pre-transformers, and
says:

"A provide pre-transformer is applied as part of the first phase of a
module?s expansion. Since it is used in the first phase, a provide
pre-transformer can use functions such as syntax-local-lift-expression to
introduce expressions and definitions in the enclosing module."

So it sounds like what I want is definitely possible, but I have no idea
how to actually *do* it. Does anyone have some simple well-explained
examples to guide me in the right direction?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140919/036e2577/attachment.html>

From asumu at ccs.neu.edu  Sat Sep 20 00:51:11 2014
From: asumu at ccs.neu.edu (Asumu Takikawa)
Date: Sat, 20 Sep 2014 00:51:11 -0400
Subject: [racket] RacketCon live video
Message-ID: <20140920045111.GI21882@localhost>

Hi all,

We are planning to air a RacketCon live stream again this year via
Google Hangouts. Barring any technical difficulties, we'll have the
stream up tomorrow at this link:

  https://plus.google.com/events/co9ghndpkko2q2p9jcd17a0fr1g

The program is here:

  http://con.racket-lang.org/

Last year there was a cap on stream length, so we may have to switch to
a new stream partway through. We'll send out a new link if we have to do
that.

Cheers,
Asumu

From leif at leifandersen.net  Sat Sep 20 10:31:29 2014
From: leif at leifandersen.net (Leif Andersen)
Date: Sat, 20 Sep 2014 09:31:29 -0500
Subject: [racket] RacketCon live video
In-Reply-To: <20140920045111.GI21882@localhost>
References: <20140920045111.GI21882@localhost>
Message-ID: <CAAVaeEBSeFT9LDpQOAXd1abQNsTF0OX8tqNWAs8T5a9bTyPB-A@mail.gmail.com>

Hello all,

The feed has started.

Again, the feed is here:

  https://plus.google.com/events/co9ghndpkko2q2p9jcd17a0fr1g

And the program is here:

  http://con.racket-lang.org/


~Leif Andersen

On Fri, Sep 19, 2014 at 11:51 PM, Asumu Takikawa <asumu at ccs.neu.edu> wrote:

> Hi all,
>
> We are planning to air a RacketCon live stream again this year via
> Google Hangouts. Barring any technical difficulties, we'll have the
> stream up tomorrow at this link:
>
>   https://plus.google.com/events/co9ghndpkko2q2p9jcd17a0fr1g
>
> The program is here:
>
>   http://con.racket-lang.org/
>
> Last year there was a cap on stream length, so we may have to switch to
> a new stream partway through. We'll send out a new link if we have to do
> that.
>
> Cheers,
> Asumu
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140920/450659f3/attachment.html>

From asumu at ccs.neu.edu  Sat Sep 20 13:25:55 2014
From: asumu at ccs.neu.edu (Asumu Takikawa)
Date: Sat, 20 Sep 2014 13:25:55 -0400
Subject: [racket] RacketCon live video
In-Reply-To: <20140920045111.GI21882@localhost>
References: <20140920045111.GI21882@localhost>
Message-ID: <20140920172555.GK21882@localhost>

On 2014-09-20 00:51:11 -0400, Asumu Takikawa wrote:
> Last year there was a cap on stream length, so we may have to switch to
> a new stream partway through. We'll send out a new link if we have to do
> that.

We'll be putting up Part 2 of the stream here:

  https://plus.google.com/events/crsu3iegmu73htksd9l6l8pfek8

We restart after the lunch break at 1:40PM Central Time.

Cheers,
Asumu

From asumu at ccs.neu.edu  Sat Sep 20 14:33:00 2014
From: asumu at ccs.neu.edu (Asumu Takikawa)
Date: Sat, 20 Sep 2014 14:33:00 -0400
Subject: [racket] RacketCon live video
In-Reply-To: <20140920172555.GK21882@localhost>
References: <20140920045111.GI21882@localhost>
 <20140920172555.GK21882@localhost>
Message-ID: <20140920183300.GL21882@localhost>

On 2014-09-20 13:25:55 -0400, Asumu Takikawa wrote:
> We'll be putting up Part 2 of the stream here:
>
>   https://plus.google.com/events/crsu3iegmu73htksd9l6l8pfek8

The stream is live again at the new link. The talks will resume in ~10
minutes at 1:40PM Central.

Cheers,
Asumu

From daniel.a.prager at gmail.com  Sun Sep 21 00:52:15 2014
From: daniel.a.prager at gmail.com (Daniel Prager)
Date: Sat, 20 Sep 2014 23:52:15 -0500
Subject: [racket] Rendering the PLT symbol in Racket?
Message-ID: <CAFKxZVUvUD=AHp1bEJO0soUHjfUKLZJhTFdKNcKWWM0t=maqmw@mail.gmail.com>

There was interest following my talk at RacketCon in a www.youpatch.com
quilt pattern featuring the PLT symbol
(file:///Users/dan/Documents/plt.svg).

I was having a play with youpatch.com by standard means, to put together a
commemorative pattern, but I think I can do a cleaner job given access to
Racket code that generates the image.

Can anyone point me to, or send me the code?


Dan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140920/4afee3af/attachment.html>

From lysseus at gmail.com  Sun Sep 21 01:47:21 2014
From: lysseus at gmail.com (Kevin Forchione)
Date: Sat, 20 Sep 2014 22:47:21 -0700
Subject: [racket]  ->i contract question
Message-ID: <7D717570-D4F1-4F93-A6D8-70A981189839@gmail.com>

Hi guys,
Is there a syntax for dependent-range that would allow 0 or more values to be specified? The documentation appears to describe only a specified number of values. Is there an analogue to list of and non-empty-listof?

-Kevin

From spencerflorence at gmail.com  Sun Sep 21 12:21:44 2014
From: spencerflorence at gmail.com (Spencer florence)
Date: Sun, 21 Sep 2014 09:21:44 -0700 (PDT)
Subject: [racket] Provide transformers
In-Reply-To: <CAAXAoJXbO+D0mPzAonTh+uNPGW6ajG0zSyeh8=gsFxVr5OQQyw@mail.gmail.com>
References: <CAAXAoJXbO+D0mPzAonTh+uNPGW6ajG0zSyeh8=gsFxVr5OQQyw@mail.gmail.com>
Message-ID: <1411316503678.d5b8e301@Nodemailer>

I think the way i have seen this done is to use a macro at the top level instead of a provide transformer.




for example define/provide-test-suite from rackunit probably works a little like




(define/provide-test-suite tests blah blah ?)




=>




(begin?

? ?(provide tests)

? ?(define tests (test-suite blah blah blah ?))







So maybe you could make a provide/test-pred form?

On Fri, Sep 19, 2014 at 4:42 PM, Jack Firth <jackhfirth at gmail.com> wrote:

> I've been looking into how to design a macro that would specify tests in a
> provide transformer, for example something like
>     (provide (test-pred-out some-integer integer?))
> I'd like that to expand to:
>     (provide some-integer)
>     (module+ test
>       (check-pred integer? some-integer))
> From what I can gather from the reference, normally provide forms are done
> with define-provide-syntax, but that macro doesn't let me splice anything
> into the module body. The reference mentions provide pre-transformers, and
> says:
> "A provide pre-transformer is applied as part of the first phase of a
> module?s expansion. Since it is used in the first phase, a provide
> pre-transformer can use functions such as syntax-local-lift-expression to
> introduce expressions and definitions in the enclosing module."
> So it sounds like what I want is definitely possible, but I have no idea
> how to actually *do* it. Does anyone have some simple well-explained
> examples to guide me in the right direction?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140921/620eb4f8/attachment.html>

From jarcane at gmail.com  Sun Sep 21 17:12:37 2014
From: jarcane at gmail.com (J Arcane)
Date: Mon, 22 Sep 2014 00:12:37 +0300
Subject: [racket] Hosting the try-racket REPL.
Message-ID: <CA+HVF0Djjj+mA-BYUeBNwYXOt+ELJt7CixdrO25S7B2Ktgeo+Q@mail.gmail.com>

Greetings,

For some time now, it's bothered me a bit that Racket doesn't have an
online REPL currently hosted anywhere. There's one written here:
https://github.com/voila/try-racket

But no one's hosted it anywhere. So I took it upon myself to fix that. I've
purchased try-racket.org and a basic DigitalOcean droplet to host it on,
but I've hit a snag: I can't get it to run.

It runs more or less without trouble on my personal FreeBSD box, but on my
Debian 7 droplet I waded through countless dependency issues until finally
I reached what seems to be this same error:

http://bugs.racket-lang.org/query/?cmd=view%20audit-trail&pr=12465

Presuming that perhaps this was in someway trying to run Racket in an
actual GUI window whether I wanted it to or not, I attempted to run it as
--script, but this just quits silently instead, leaving no servlet behind
either.

Any ideas what I might be doing wrong? or have I hit a proper bug? (IME
it's usually the former ... )

John Berry
http://jarcane.github.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140922/de71fc93/attachment.html>

From alexander at knauth.org  Sun Sep 21 19:31:01 2014
From: alexander at knauth.org (Alexander D. Knauth)
Date: Sun, 21 Sep 2014 19:31:01 -0400
Subject: [racket] define-type #:omit-define-syntaxes across different modules
Message-ID: <2F4CD703-EEB1-4BE4-9CE3-62D87615D462@knauth.org>

An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140921/f7f3f3b2/attachment.html>

From m4burns at csclub.uwaterloo.ca  Sun Sep 21 19:00:02 2014
From: m4burns at csclub.uwaterloo.ca (Marc Burns)
Date: Sun, 21 Sep 2014 19:00:02 -0400
Subject: [racket] Hosting the try-racket REPL.
In-Reply-To: <CA+HVF0Djjj+mA-BYUeBNwYXOt+ELJt7CixdrO25S7B2Ktgeo+Q@mail.gmail.com>
References: <CA+HVF0Djjj+mA-BYUeBNwYXOt+ELJt7CixdrO25S7B2Ktgeo+Q@mail.gmail.com>
Message-ID: <20140921230002.GD29774@happierface>

The culprit is the slideshow/code module loaded into the sandbox
evaluator. Requiring slideshow/code attempts to establish an X server
connection.

On Mon, Sep 22, 2014 at 12:12:37AM +0300, J Arcane wrote:
> Greetings,
> 
> For some time now, it's bothered me a bit that Racket doesn't have an
> online REPL currently hosted anywhere. There's one written here:
> https://github.com/voila/try-racket
> 
> But no one's hosted it anywhere. So I took it upon myself to fix that. I've
> purchased try-racket.org and a basic DigitalOcean droplet to host it on,
> but I've hit a snag: I can't get it to run.
> 
> It runs more or less without trouble on my personal FreeBSD box, but on my
> Debian 7 droplet I waded through countless dependency issues until finally
> I reached what seems to be this same error:
> 
> http://bugs.racket-lang.org/query/?cmd=view%20audit-trail&pr=12465
> 
> Presuming that perhaps this was in someway trying to run Racket in an
> actual GUI window whether I wanted it to or not, I attempted to run it as
> --script, but this just quits silently instead, leaving no servlet behind
> either.
> 
> Any ideas what I might be doing wrong? or have I hit a proper bug? (IME
> it's usually the former ... )
> 
> John Berry
> http://jarcane.github.com

> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users


From alexander at knauth.org  Sun Sep 21 19:59:01 2014
From: alexander at knauth.org (Alexander D. Knauth)
Date: Sun, 21 Sep 2014 19:59:01 -0400
Subject: [racket] Rendering the PLT symbol in Racket?
In-Reply-To: <CAFKxZVUvUD=AHp1bEJO0soUHjfUKLZJhTFdKNcKWWM0t=maqmw@mail.gmail.com>
References: <CAFKxZVUvUD=AHp1bEJO0soUHjfUKLZJhTFdKNcKWWM0t=maqmw@mail.gmail.com>
Message-ID: <989C0D02-3EED-40CB-8DB6-E4B36B6DFC4C@knauth.org>

I tried it with the images in the email, but it was too big, so here it is just links.
I think there was something on the list about this before, and I found these again from that:
http://docs.racket-lang.org/images/Logos.html?q=plt-logo#%28def._%28%28lib._images%2Flogos..rkt%29._plt-logo%29%29
(require images/logos)
(plt-logo)
and also
http://docs.racket-lang.org/draw/overview.html?q=racket%2Fdraw#%28part._.Drawing_.Paths%29 (down a little bit)


On Sep 21, 2014, at 12:52 AM, Daniel Prager <daniel.a.prager at gmail.com> wrote:

> There was interest following my talk at RacketCon in a www.youpatch.com quilt pattern featuring the PLT symbol (file:///Users/dan/Documents/plt.svg).
> 
> I was having a play with youpatch.com by standard means, to put together a commemorative pattern, but I think I can do a cleaner job given access to Racket code that generates the image.
> 
> Can anyone point me to, or send me the code?
> 
> 
> Dan
> ____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140921/23f972a2/attachment-0001.html>

From robby at eecs.northwestern.edu  Sun Sep 21 21:02:10 2014
From: robby at eecs.northwestern.edu (Robby Findler)
Date: Sun, 21 Sep 2014 20:02:10 -0500
Subject: [racket] Hosting the try-racket REPL.
In-Reply-To: <20140921230002.GD29774@happierface>
References: <CA+HVF0Djjj+mA-BYUeBNwYXOt+ELJt7CixdrO25S7B2Ktgeo+Q@mail.gmail.com>
 <20140921230002.GD29774@happierface>
Message-ID: <CAL3TdOPyujV7ork7Y+r+q---7w1WBpO-GN=H7wKDWK0yknjcMQ@mail.gmail.com>

Maybe you can use  slideshow/code-pict instead?

Robby

On Sun, Sep 21, 2014 at 6:00 PM, Marc Burns <m4burns at csclub.uwaterloo.ca> wrote:
> The culprit is the slideshow/code module loaded into the sandbox
> evaluator. Requiring slideshow/code attempts to establish an X server
> connection.
>
> On Mon, Sep 22, 2014 at 12:12:37AM +0300, J Arcane wrote:
>> Greetings,
>>
>> For some time now, it's bothered me a bit that Racket doesn't have an
>> online REPL currently hosted anywhere. There's one written here:
>> https://github.com/voila/try-racket
>>
>> But no one's hosted it anywhere. So I took it upon myself to fix that. I've
>> purchased try-racket.org and a basic DigitalOcean droplet to host it on,
>> but I've hit a snag: I can't get it to run.
>>
>> It runs more or less without trouble on my personal FreeBSD box, but on my
>> Debian 7 droplet I waded through countless dependency issues until finally
>> I reached what seems to be this same error:
>>
>> http://bugs.racket-lang.org/query/?cmd=view%20audit-trail&pr=12465
>>
>> Presuming that perhaps this was in someway trying to run Racket in an
>> actual GUI window whether I wanted it to or not, I attempted to run it as
>> --script, but this just quits silently instead, leaving no servlet behind
>> either.
>>
>> Any ideas what I might be doing wrong? or have I hit a proper bug? (IME
>> it's usually the former ... )
>>
>> John Berry
>> http://jarcane.github.com
>
>> ____________________
>>   Racket Users list:
>>   http://lists.racket-lang.org/users
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users

From daniel.a.prager at gmail.com  Sun Sep 21 21:09:23 2014
From: daniel.a.prager at gmail.com (Daniel Prager)
Date: Sun, 21 Sep 2014 18:09:23 -0700
Subject: [racket] Rendering the PLT symbol in Racket?
In-Reply-To: <E3D0D163-C3E4-4350-9144-609B617D3F1F@brinckerhoff.org>
References: <CAFKxZVUvUD=AHp1bEJO0soUHjfUKLZJhTFdKNcKWWM0t=maqmw@mail.gmail.com>
 <989C0D02-3EED-40CB-8DB6-E4B36B6DFC4C@knauth.org>
 <E3D0D163-C3E4-4350-9144-609B617D3F1F@brinckerhoff.org>
Message-ID: <CAFKxZVVKN9qVsdA_RLP92LXipvTuoPTitHZKaNbYY6yH9xcxUw@mail.gmail.com>

Thanks Alexander & John

On further examination I'm not sure that direct-drawing particularly helps
me, but I appreciate the references (and the history).

I'll make sure that there's a nice effect for the design: just bear with
me, I'm not quite satisfied with what can be currently be achieved.

Dan

On Sun, Sep 21, 2014 at 5:52 PM, John Clements <clements at brinckerhoff.org>
wrote:

> There are a whole bunch of ways of answering this question; you might want
> Neil?s fantastic ray-traced version, or just the simple paths. In case it?s
> helpful, here?s the original postscript file generated by Shriram, Robby,
> and Matthew. Oh, wait, here?s the comment from the file:
>
> %%Title: PLT Lambda
> %%Creator: Matthew, then Shriram, finally Robby
> %%CreationDate: June 24, 1997
> %%For: PLT
>
> (NB: Apple Mail tries to render this as an image, pretty badly. Save the
> attachment to read it.)
>
>
>
> On Sep 21, 2014, at 4:59 PM, Alexander D. Knauth <alexander at knauth.org>
> wrote:
>
> > I tried it with the images in the email, but it was too big, so here it
> is just links.
> > I think there was something on the list about this before, and I found
> these again from that:
> >
> http://docs.racket-lang.org/images/Logos.html?q=plt-logo#%28def._%28%28lib._images%2Flogos..rkt%29._plt-logo%29%29
> > (require images/logos)
> > (plt-logo)
> > and also
> >
> http://docs.racket-lang.org/draw/overview.html?q=racket%2Fdraw#%28part._.Drawing_.Paths%29
> (down a little bit)
> >
> >
> > On Sep 21, 2014, at 12:52 AM, Daniel Prager <daniel.a.prager at gmail.com>
> wrote:
> >
> >> There was interest following my talk at RacketCon in a www.youpatch.com
> quilt pattern featuring the PLT symbol
> (file:///Users/dan/Documents/plt.svg).
> >>
> >> I was having a play with youpatch.com by standard means, to put
> together a commemorative pattern, but I think I can do a cleaner job given
> access to Racket code that generates the image.
> >>
> >> Can anyone point me to, or send me the code?
> >>
> >>
> >> Dan
> >> ____________________
> >>  Racket Users list:
> >>  http://lists.racket-lang.org/users
> > ____________________
> >  Racket Users list:
> >  http://lists.racket-lang.org/users
>
>
>


-- 
*Daniel Prager*
Agile/Lean Coaching, Software Development and Leadership
Startup: www.youpatch.com
Twitter: @agilejitsu <https://twitter.com/agilejitsu>
Blog: agile-jitsu.blogspot.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140921/9856fc04/attachment.html>

From robby at eecs.northwestern.edu  Sun Sep 21 21:13:50 2014
From: robby at eecs.northwestern.edu (Robby Findler)
Date: Sun, 21 Sep 2014 20:13:50 -0500
Subject: [racket] Rendering the PLT symbol in Racket?
In-Reply-To: <CAFKxZVVKN9qVsdA_RLP92LXipvTuoPTitHZKaNbYY6yH9xcxUw@mail.gmail.com>
References: <CAFKxZVUvUD=AHp1bEJO0soUHjfUKLZJhTFdKNcKWWM0t=maqmw@mail.gmail.com>
 <989C0D02-3EED-40CB-8DB6-E4B36B6DFC4C@knauth.org>
 <E3D0D163-C3E4-4350-9144-609B617D3F1F@brinckerhoff.org>
 <CAFKxZVVKN9qVsdA_RLP92LXipvTuoPTitHZKaNbYY6yH9xcxUw@mail.gmail.com>
Message-ID: <CAL3TdOO46r9Ojef15JuwAU7w75203dcG-DC_gn_hYJxfUOYucg@mail.gmail.com>

There is some racket code, but it just duplicates the ps code, drawing
directly into a dc. (This can be lifted into a pict, via 'dc'.) If
that would be helpful, I can supply it.

Robby


On Sun, Sep 21, 2014 at 8:09 PM, Daniel Prager
<daniel.a.prager at gmail.com> wrote:
> Thanks Alexander & John
>
> On further examination I'm not sure that direct-drawing particularly helps
> me, but I appreciate the references (and the history).
>
> I'll make sure that there's a nice effect for the design: just bear with me,
> I'm not quite satisfied with what can be currently be achieved.
>
> Dan
>
> On Sun, Sep 21, 2014 at 5:52 PM, John Clements <clements at brinckerhoff.org>
> wrote:
>>
>> There are a whole bunch of ways of answering this question; you might want
>> Neil?s fantastic ray-traced version, or just the simple paths. In case it?s
>> helpful, here?s the original postscript file generated by Shriram, Robby,
>> and Matthew. Oh, wait, here?s the comment from the file:
>>
>> %%Title: PLT Lambda
>> %%Creator: Matthew, then Shriram, finally Robby
>> %%CreationDate: June 24, 1997
>> %%For: PLT
>>
>> (NB: Apple Mail tries to render this as an image, pretty badly. Save the
>> attachment to read it.)
>>
>>
>>
>>
>> On Sep 21, 2014, at 4:59 PM, Alexander D. Knauth <alexander at knauth.org>
>> wrote:
>>
>> > I tried it with the images in the email, but it was too big, so here it
>> > is just links.
>> > I think there was something on the list about this before, and I found
>> > these again from that:
>> >
>> > http://docs.racket-lang.org/images/Logos.html?q=plt-logo#%28def._%28%28lib._images%2Flogos..rkt%29._plt-logo%29%29
>> > (require images/logos)
>> > (plt-logo)
>> > and also
>> >
>> > http://docs.racket-lang.org/draw/overview.html?q=racket%2Fdraw#%28part._.Drawing_.Paths%29
>> > (down a little bit)
>> >
>> >
>> > On Sep 21, 2014, at 12:52 AM, Daniel Prager <daniel.a.prager at gmail.com>
>> > wrote:
>> >
>> >> There was interest following my talk at RacketCon in a www.youpatch.com
>> >> quilt pattern featuring the PLT symbol
>> >> (file:///Users/dan/Documents/plt.svg).
>> >>
>> >> I was having a play with youpatch.com by standard means, to put
>> >> together a commemorative pattern, but I think I can do a cleaner job given
>> >> access to Racket code that generates the image.
>> >>
>> >> Can anyone point me to, or send me the code?
>> >>
>> >>
>> >> Dan
>> >> ____________________
>> >>  Racket Users list:
>> >>  http://lists.racket-lang.org/users
>> > ____________________
>> >  Racket Users list:
>> >  http://lists.racket-lang.org/users
>>
>>
>
>
>
> --
> Daniel Prager
> Agile/Lean Coaching, Software Development and Leadership
> Startup: www.youpatch.com
> Twitter: @agilejitsu
> Blog: agile-jitsu.blogspot.com
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>


From daniel.a.prager at gmail.com  Sun Sep 21 21:30:00 2014
From: daniel.a.prager at gmail.com (Daniel Prager)
Date: Sun, 21 Sep 2014 18:30:00 -0700
Subject: [racket] Rendering the PLT symbol in Racket?
In-Reply-To: <CAL3TdOO46r9Ojef15JuwAU7w75203dcG-DC_gn_hYJxfUOYucg@mail.gmail.com>
References: <CAFKxZVUvUD=AHp1bEJO0soUHjfUKLZJhTFdKNcKWWM0t=maqmw@mail.gmail.com>
 <989C0D02-3EED-40CB-8DB6-E4B36B6DFC4C@knauth.org>
 <E3D0D163-C3E4-4350-9144-609B617D3F1F@brinckerhoff.org>
 <CAFKxZVVKN9qVsdA_RLP92LXipvTuoPTitHZKaNbYY6yH9xcxUw@mail.gmail.com>
 <CAL3TdOO46r9Ojef15JuwAU7w75203dcG-DC_gn_hYJxfUOYucg@mail.gmail.com>
Message-ID: <CAFKxZVUWqa0O+ipeKcoFq9_HcVpbiCHCrR+9Dr2_-h3uKpizBg@mail.gmail.com>

Thanks Robby

That may be useful.

FYI: I'm biting the bullet and trying to introduce a "good" triangular
rendering to reduce the jaggedness of curves (which are common in images
like the simple PLT logo) and while I've almost got something that almost
works from a sampled image, I'm not there yet.

Dan

On Sun, Sep 21, 2014 at 6:13 PM, Robby Findler <robby at eecs.northwestern.edu>
wrote:

> There is some racket code, but it just duplicates the ps code, drawing
> directly into a dc. (This can be lifted into a pict, via 'dc'.) If
> that would be helpful, I can supply it.
>
> Robby
>
>
> On Sun, Sep 21, 2014 at 8:09 PM, Daniel Prager
> <daniel.a.prager at gmail.com> wrote:
> > Thanks Alexander & John
> >
> > On further examination I'm not sure that direct-drawing particularly
> helps
> > me, but I appreciate the references (and the history).
> >
> > I'll make sure that there's a nice effect for the design: just bear with
> me,
> > I'm not quite satisfied with what can be currently be achieved.
> >
> > Dan
> >
> > On Sun, Sep 21, 2014 at 5:52 PM, John Clements <
> clements at brinckerhoff.org>
> > wrote:
> >>
> >> There are a whole bunch of ways of answering this question; you might
> want
> >> Neil?s fantastic ray-traced version, or just the simple paths. In case
> it?s
> >> helpful, here?s the original postscript file generated by Shriram,
> Robby,
> >> and Matthew. Oh, wait, here?s the comment from the file:
> >>
> >> %%Title: PLT Lambda
> >> %%Creator: Matthew, then Shriram, finally Robby
> >> %%CreationDate: June 24, 1997
> >> %%For: PLT
> >>
> >> (NB: Apple Mail tries to render this as an image, pretty badly. Save the
> >> attachment to read it.)
> >>
> >>
> >>
> >>
> >> On Sep 21, 2014, at 4:59 PM, Alexander D. Knauth <alexander at knauth.org>
> >> wrote:
> >>
> >> > I tried it with the images in the email, but it was too big, so here
> it
> >> > is just links.
> >> > I think there was something on the list about this before, and I found
> >> > these again from that:
> >> >
> >> >
> http://docs.racket-lang.org/images/Logos.html?q=plt-logo#%28def._%28%28lib._images%2Flogos..rkt%29._plt-logo%29%29
> >> > (require images/logos)
> >> > (plt-logo)
> >> > and also
> >> >
> >> >
> http://docs.racket-lang.org/draw/overview.html?q=racket%2Fdraw#%28part._.Drawing_.Paths%29
> >> > (down a little bit)
> >> >
> >> >
> >> > On Sep 21, 2014, at 12:52 AM, Daniel Prager <
> daniel.a.prager at gmail.com>
> >> > wrote:
> >> >
> >> >> There was interest following my talk at RacketCon in a
> www.youpatch.com
> >> >> quilt pattern featuring the PLT symbol
> >> >> (file:///Users/dan/Documents/plt.svg).
> >> >>
> >> >> I was having a play with youpatch.com by standard means, to put
> >> >> together a commemorative pattern, but I think I can do a cleaner job
> given
> >> >> access to Racket code that generates the image.
> >> >>
> >> >> Can anyone point me to, or send me the code?
> >> >>
> >> >>
> >> >> Dan
> >> >> ____________________
> >> >>  Racket Users list:
> >> >>  http://lists.racket-lang.org/users
> >> > ____________________
> >> >  Racket Users list:
> >> >  http://lists.racket-lang.org/users
> >>
> >>
> >
> >
> >
> > --
> > Daniel Prager
> > Agile/Lean Coaching, Software Development and Leadership
> > Startup: www.youpatch.com
> > Twitter: @agilejitsu
> > Blog: agile-jitsu.blogspot.com
> >
> > ____________________
> >   Racket Users list:
> >   http://lists.racket-lang.org/users
> >
>



-- 
*Daniel Prager*
Agile/Lean Coaching, Software Development and Leadership
Startup: www.youpatch.com
Twitter: @agilejitsu <https://twitter.com/agilejitsu>
Blog: agile-jitsu.blogspot.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140921/3a5d90e7/attachment-0001.html>

From jarcane at gmail.com  Mon Sep 22 01:17:35 2014
From: jarcane at gmail.com (J Arcane)
Date: Mon, 22 Sep 2014 08:17:35 +0300
Subject: [racket] Hosting the try-racket REPL.
In-Reply-To: <CAL3TdOPyujV7ork7Y+r+q---7w1WBpO-GN=H7wKDWK0yknjcMQ@mail.gmail.com>
References: <CA+HVF0Djjj+mA-BYUeBNwYXOt+ELJt7CixdrO25S7B2Ktgeo+Q@mail.gmail.com>
 <20140921230002.GD29774@happierface>
 <CAL3TdOPyujV7ork7Y+r+q---7w1WBpO-GN=H7wKDWK0yknjcMQ@mail.gmail.com>
Message-ID: <CA+HVF0Axwivp6GebTXfdoaHwyDwGAupe4aHP4yM0Z1qg3m8DLQ@mail.gmail.com>

Hmm.

Alright, replacing the calls to slideshow with calls directly to pict, as
well as commenting out the racket/gui/base calls successfully gets the code
to run without X.

However, something about how it converts the picts for web isn't working,
so instead of a neat little circle, for instance, I get "(pict '(prog
#<procedure:... etc." instead.

I suspect this is probably something to do with a difference in behavior in
how images are converted by the server, which does a convertible? test and
then runs:
 (run-code ev `(bytes-append #"data:image/png;base64,"
                         (base64-encode (convert ,res 'png-bytes) #"")))

At an idiot's guess, something is different about how pict is making the
images, so convertible? is #f and the code just prints the (pict '(prog...
stuff.

I've set up a fork here: https://github.com/jarcane/try-racket
I'm still working on getting the DNS set properly though, so no
try-racket.org yet; the servlet is public at http://104.131.18.192:8000/
but I haven't figured out how to get try-racket.org to go there yet.


<robby at eecs.northwestern.edu> wrote:

> Maybe you can use  slideshow/code-pict instead?
>
> Robby
>
> On Sun, Sep 21, 2014 at 6:00 PM, Marc Burns <m4burns at csclub.uwaterloo.ca>
> wrote:
> > The culprit is the slideshow/code module loaded into the sandbox
> > evaluator. Requiring slideshow/code attempts to establish an X server
> > connection.
> >
> > On Mon, Sep 22, 2014 at 12:12:37AM +0300, J Arcane wrote:
> >> Greetings,
> >>
> >> For some time now, it's bothered me a bit that Racket doesn't have an
> >> online REPL currently hosted anywhere. There's one written here:
> >> https://github.com/voila/try-racket
> >>
> >> But no one's hosted it anywhere. So I took it upon myself to fix that.
> I've
> >> purchased try-racket.org and a basic DigitalOcean droplet to host it
> on,
> >> but I've hit a snag: I can't get it to run.
> >>
> >> It runs more or less without trouble on my personal FreeBSD box, but on
> my
> >> Debian 7 droplet I waded through countless dependency issues until
> finally
> >> I reached what seems to be this same error:
> >>
> >> http://bugs.racket-lang.org/query/?cmd=view%20audit-trail&pr=12465
> >>
> >> Presuming that perhaps this was in someway trying to run Racket in an
> >> actual GUI window whether I wanted it to or not, I attempted to run it
> as
> >> --script, but this just quits silently instead, leaving no servlet
> behind
> >> either.
> >>
> >> Any ideas what I might be doing wrong? or have I hit a proper bug? (IME
> >> it's usually the former ... )
> >>
> >> John Berry
> >> http://jarcane.github.com
> >
> >> ____________________
> >>   Racket Users list:
> >>   http://lists.racket-lang.org/users
> >
> > ____________________
> >   Racket Users list:
> >   http://lists.racket-lang.org/users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140922/88ea1df5/attachment.html>

From fractallyte at csi.com  Mon Sep 22 02:28:28 2014
From: fractallyte at csi.com (Amir Ansari)
Date: Mon, 22 Sep 2014 07:28:28 +0100
Subject: [racket] Hosting the try-racket REPL.
Message-ID: <20140922072828.01b2011be8962eda327b3dfa@csi.com>

Have you tried Xvfb (http://en.wikipedia.org/wiki/Xvfb), a virtual X server? It runs headlessly, sidesteps the whole issue of having to fork the code...

Amir

From daniel.a.prager at gmail.com  Mon Sep 22 02:29:21 2014
From: daniel.a.prager at gmail.com (Daniel Prager)
Date: Sun, 21 Sep 2014 23:29:21 -0700
Subject: [racket] Rendering the PLT symbol in Racket?
In-Reply-To: <CAFKxZVUWqa0O+ipeKcoFq9_HcVpbiCHCrR+9Dr2_-h3uKpizBg@mail.gmail.com>
References: <CAFKxZVUvUD=AHp1bEJO0soUHjfUKLZJhTFdKNcKWWM0t=maqmw@mail.gmail.com>
 <989C0D02-3EED-40CB-8DB6-E4B36B6DFC4C@knauth.org>
 <E3D0D163-C3E4-4350-9144-609B617D3F1F@brinckerhoff.org>
 <CAFKxZVVKN9qVsdA_RLP92LXipvTuoPTitHZKaNbYY6yH9xcxUw@mail.gmail.com>
 <CAL3TdOO46r9Ojef15JuwAU7w75203dcG-DC_gn_hYJxfUOYucg@mail.gmail.com>
 <CAFKxZVUWqa0O+ipeKcoFq9_HcVpbiCHCrR+9Dr2_-h3uKpizBg@mail.gmail.com>
Message-ID: <CAFKxZVVPbRZuSpGE-NW2iriP6EQMXxdGMte9SMxR=m-gkVR9UA@mail.gmail.com>

This is what's achievable based on the simpler design with www.youpatch.com
right now (shown at medium resolution). As you can see, this kind of image
leads to heaps of "stair-casing".


I think it would be much nicer to smooth out the jaggies. I'm experimenting
with adding a facility to do this with some post-processing. This leads to
the following effect:



It will take me a little while to incorporate this into the PDF
instructional output (I'll estimate order of weeks, depending on my other
activities) .

So, what do *you* think of the difference in aesthetic quality? For those
who are keen to have their very own PLT quilt, is this improvement worth
the additional wait?

Dan
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140921/6116c171/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: PLT-Logo-Triangles.png
Type: image/png
Size: 10493 bytes
Desc: not available
URL: <http://lists.racket-lang.org/users/archive/attachments/20140921/6116c171/attachment-0002.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: plt-logo-squares.png
Type: image/png
Size: 2692 bytes
Desc: not available
URL: <http://lists.racket-lang.org/users/archive/attachments/20140921/6116c171/attachment-0003.png>

From neil at neilvandyke.org  Mon Sep 22 02:52:30 2014
From: neil at neilvandyke.org (Neil Van Dyke)
Date: Mon, 22 Sep 2014 02:52:30 -0400
Subject: [racket] Hosting the try-racket REPL.
In-Reply-To: <20140922072828.01b2011be8962eda327b3dfa@csi.com>
References: <20140922072828.01b2011be8962eda327b3dfa@csi.com>
Message-ID: <541FC72E.2090207@neilvandyke.org>

Amir Ansari wrote at 09/22/2014 02:28 AM:
> Have you tried Xvfb (http://en.wikipedia.org/wiki/Xvfb), a virtual X server? It runs headlessly, sidesteps the whole issue of having to fork the code...
>

I have no idea whether "xvfb" is the best way without looking closely at 
a particular case, but I've seen "xvfb" work in the past on production 
servers.  (Not for Racket, but for some R statistics tool that touched 
the X server, so that the tool could be used off-the-shelf, rather than 
maintaining a modified fork of the tool source.)

If you do this, just be write your startup and shutdown scripts to be 
resilient and stable (e.g., wrt available ports/sockets at the time, and 
to shut down cleanly, and so there are no race conditions or conflicts 
between multiple simultaneous users), and secure (e.g., ensure that it 
won't accidentally listen on a public network interface, and that an 
attacker can't synthesize operations to the xvfb to escalate privileges 
or DoS, and that the xvfb can't be used to leak data between users).

Neil V.


From jarcane at gmail.com  Mon Sep 22 02:54:52 2014
From: jarcane at gmail.com (J Arcane)
Date: Mon, 22 Sep 2014 09:54:52 +0300
Subject: [racket] Hosting the try-racket REPL.
In-Reply-To: <20140922072828.01b2011be8962eda327b3dfa@csi.com>
References: <20140922072828.01b2011be8962eda327b3dfa@csi.com>
Message-ID: <CA+HVF0CRTJALp_ExFHNnMSUC_RVDRp0JWuJhMD5o9kTqnvT_3g@mail.gmail.com>

I had not actually heard of Xvfb before now, but that could be a good
solution. I balk at using full X11 with only 512MB of RAM, but that could
be enough to satisfy the original code's expectations. I will have to
experiment with it.

On Mon, Sep 22, 2014 at 9:28 AM, Amir Ansari <fractallyte at csi.com> wrote:

> Have you tried Xvfb (http://en.wikipedia.org/wiki/Xvfb), a virtual X
> server? It runs headlessly, sidesteps the whole issue of having to fork the
> code...
>
> Amir
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140922/e458da45/attachment.html>

From jarcane at gmail.com  Mon Sep 22 03:11:25 2014
From: jarcane at gmail.com (J Arcane)
Date: Mon, 22 Sep 2014 10:11:25 +0300
Subject: [racket] Hosting the try-racket REPL.
In-Reply-To: <541FC72E.2090207@neilvandyke.org>
References: <20140922072828.01b2011be8962eda327b3dfa@csi.com>
 <541FC72E.2090207@neilvandyke.org>
Message-ID: <CA+HVF0DNhqPq4i7YL6VkPKajX=1wL75_gebVRf-Kg3bXRCPLBw@mail.gmail.com>

That does sound an awful headache for something that otherwise doesn't seem
to be using X for anything other than dependency expectations. It seems
like there should be a non-X-dependent way to do it.

On Mon, Sep 22, 2014 at 9:52 AM, Neil Van Dyke <neil at neilvandyke.org> wrote:

> Amir Ansari wrote at 09/22/2014 02:28 AM:
>
>> Have you tried Xvfb (http://en.wikipedia.org/wiki/Xvfb), a virtual X
>> server? It runs headlessly, sidesteps the whole issue of having to fork the
>> code...
>>
>>
> I have no idea whether "xvfb" is the best way without looking closely at a
> particular case, but I've seen "xvfb" work in the past on production
> servers.  (Not for Racket, but for some R statistics tool that touched the
> X server, so that the tool could be used off-the-shelf, rather than
> maintaining a modified fork of the tool source.)
>
> If you do this, just be write your startup and shutdown scripts to be
> resilient and stable (e.g., wrt available ports/sockets at the time, and to
> shut down cleanly, and so there are no race conditions or conflicts between
> multiple simultaneous users), and secure (e.g., ensure that it won't
> accidentally listen on a public network interface, and that an attacker
> can't synthesize operations to the xvfb to escalate privileges or DoS, and
> that the xvfb can't be used to leak data between users).
>
> Neil V.
>
>
> ____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140922/d040ea62/attachment.html>

From fractallyte at csi.com  Mon Sep 22 05:06:38 2014
From: fractallyte at csi.com (Amir Ansari)
Date: Mon, 22 Sep 2014 10:06:38 +0100
Subject: [racket] Hosting the try-racket REPL.
In-Reply-To: <CA+HVF0DNhqPq4i7YL6VkPKajX=1wL75_gebVRf-Kg3bXRCPLBw@mail.gmail.com>
References: <20140922072828.01b2011be8962eda327b3dfa@csi.com>
 <541FC72E.2090207@neilvandyke.org>
 <CA+HVF0DNhqPq4i7YL6VkPKajX=1wL75_gebVRf-Kg3bXRCPLBw@mail.gmail.com>
Message-ID: <20140922100638.8456bcb8d777504e45ae738f@csi.com>

If you run it inside a BSD jail, that should take care of a lot of the possible issues...

But you're right, of course. It should be able to run without X.


On Mon, 22 Sep 2014 10:11:25 +0300
J Arcane <jarcane at gmail.com> wrote:

> That does sound an awful headache for something that otherwise doesn't seem
> to be using X for anything other than dependency expectations. It seems
> like there should be a non-X-dependent way to do it.

From neil at neilvandyke.org  Mon Sep 22 05:09:27 2014
From: neil at neilvandyke.org (Neil Van Dyke)
Date: Mon, 22 Sep 2014 05:09:27 -0400
Subject: [racket] Hosting the try-racket REPL.
In-Reply-To: <CA+HVF0CRTJALp_ExFHNnMSUC_RVDRp0JWuJhMD5o9kTqnvT_3g@mail.gmail.com>
References: <20140922072828.01b2011be8962eda327b3dfa@csi.com>
 <CA+HVF0CRTJALp_ExFHNnMSUC_RVDRp0JWuJhMD5o9kTqnvT_3g@mail.gmail.com>
Message-ID: <541FE747.7070403@neilvandyke.org>


J Arcane wrote at 09/22/2014 02:54 AM:
> I had not actually heard of Xvfb before now, but that could be a good 
> solution. I balk at using full X11 with only 512MB of RAM, but that 
> could be enough to satisfy the original code's expectations. I will 
> have to experiment with it.

Side comment: I wouldn't worry too much about the memory requirements of 
an X11 server daemon (we used to run X11 and everything else on 16MB 
engineering workstations fine, and universities would sometimes run X on 
4MB workstations with megapixel displays, even without local swap.)

The X11 daemon on my computer right here is using only 3MB virtual 
memory total, and I have some server-intensive apps like Firefox running 
against it, on two heads.

(It's mostly only the modern ``desktop environments'' that tend to be 
bloatware (e.g. even a decade ago with those, the little clock panel 
applet was reporting as 50MB RSS), but a plain X11 server daemon itself 
can be pretty small.  A lot of the early development of the desktop 
environments you're using now was by smart-but-inexperienced 20yos who 
were tossed into the deep end, under 20yo managers who were trying to 
start a company.  With some noteworthy exceptions.  Since then, the 
desktop environments seem to have been evolved by often more-experienced 
people, and improved, but the desktop environments weren't built with 
the great architectures and practices from the start, which tends to be 
a technical and cultural burden for a very long time.  And developers 
have usually been targeting contemporary PC desktop hardware, when 
several gigabytes of RAM can be assumed, except for occasional forays 
into devices that are battery-powered and/or are intended to be 
ultra-cheap to manufacture.)

Neil V.


From alexander at knauth.org  Mon Sep 22 06:57:44 2014
From: alexander at knauth.org (Alexander D. Knauth)
Date: Mon, 22 Sep 2014 06:57:44 -0400
Subject: [racket] Hosting the try-racket REPL.
In-Reply-To: <CA+HVF0Axwivp6GebTXfdoaHwyDwGAupe4aHP4yM0Z1qg3m8DLQ@mail.gmail.com>
References: <CA+HVF0Djjj+mA-BYUeBNwYXOt+ELJt7CixdrO25S7B2Ktgeo+Q@mail.gmail.com>
 <20140921230002.GD29774@happierface>
 <CAL3TdOPyujV7ork7Y+r+q---7w1WBpO-GN=H7wKDWK0yknjcMQ@mail.gmail.com>
 <CA+HVF0Axwivp6GebTXfdoaHwyDwGAupe4aHP4yM0Z1qg3m8DLQ@mail.gmail.com>
Message-ID: <1F89AEEA-021A-44E8-A5BE-F7257E11AC93@knauth.org>



On Sep 22, 2014, at 1:17 AM, J Arcane <jarcane at gmail.com> wrote:

> 
> I've set up a fork here: https://github.com/jarcane/try-racket
> I'm still working on getting the DNS set properly though, so no try-racket.org yet; the servlet is public at http://104.131.18.192:8000/ but I haven't figured out how to get try-racket.org to go there yet.

Can you get it to give a newline if the parens don?t match yet, so for example:
> (for/list ([i (in-range 4)])
eval:1:0: read: expected a `)' to close `('
  context...:
   /usr/share/racket/pkgs/sandbox-lib/racket/sandbox.rkt:113:0: default-sandbox-reader
   /usr/share/racket/pkgs/sandbox-lib/racket/sandbox.rkt:536:0: input->code
   /usr/share/racket/pkgs/sandbox-lib/racket/sandbox.rkt:856:14
   /usr/share/racket/pkgs/sandbox-lib/racket/sandbox.rkt:379:0: call-with-limits
   /usr/share/racket/pkgs/sandbox-lib/racket/sandbox.rkt:475:0: call-with-custodian-shutdown
   /usr/share/racket/collects/racket/private/more-scheme.rkt:147:2: call-with-break-parameterization
   /usr/share/racket/pkgs/sandbox-lib/racket/sandbox.rkt:837:5: loop
>  

> 
> 
> <robby at eecs.northwestern.edu> wrote:
> Maybe you can use  slideshow/code-pict instead?
> 
> Robby
> 
> On Sun, Sep 21, 2014 at 6:00 PM, Marc Burns <m4burns at csclub.uwaterloo.ca> wrote:
> > The culprit is the slideshow/code module loaded into the sandbox
> > evaluator. Requiring slideshow/code attempts to establish an X server
> > connection.
> >
> > On Mon, Sep 22, 2014 at 12:12:37AM +0300, J Arcane wrote:
> >> Greetings,
> >>
> >> For some time now, it's bothered me a bit that Racket doesn't have an
> >> online REPL currently hosted anywhere. There's one written here:
> >> https://github.com/voila/try-racket
> >>
> >> But no one's hosted it anywhere. So I took it upon myself to fix that. I've
> >> purchased try-racket.org and a basic DigitalOcean droplet to host it on,
> >> but I've hit a snag: I can't get it to run.
> >>
> >> It runs more or less without trouble on my personal FreeBSD box, but on my
> >> Debian 7 droplet I waded through countless dependency issues until finally
> >> I reached what seems to be this same error:
> >>
> >> http://bugs.racket-lang.org/query/?cmd=view%20audit-trail&pr=12465
> >>
> >> Presuming that perhaps this was in someway trying to run Racket in an
> >> actual GUI window whether I wanted it to or not, I attempted to run it as
> >> --script, but this just quits silently instead, leaving no servlet behind
> >> either.
> >>
> >> Any ideas what I might be doing wrong? or have I hit a proper bug? (IME
> >> it's usually the former ... )
> >>
> >> John Berry
> >> http://jarcane.github.com
> >
> >> ____________________
> >>   Racket Users list:
> >>   http://lists.racket-lang.org/users
> >
> > ____________________
> >   Racket Users list:
> >   http://lists.racket-lang.org/users
> 
> ____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140922/805e313e/attachment-0001.html>

From robby at eecs.northwestern.edu  Mon Sep 22 08:30:56 2014
From: robby at eecs.northwestern.edu (Robby Findler)
Date: Mon, 22 Sep 2014 07:30:56 -0500
Subject: [racket] Hosting the try-racket REPL.
In-Reply-To: <CA+HVF0Axwivp6GebTXfdoaHwyDwGAupe4aHP4yM0Z1qg3m8DLQ@mail.gmail.com>
References: <CA+HVF0Djjj+mA-BYUeBNwYXOt+ELJt7CixdrO25S7B2Ktgeo+Q@mail.gmail.com>
 <20140921230002.GD29774@happierface>
 <CAL3TdOPyujV7ork7Y+r+q---7w1WBpO-GN=H7wKDWK0yknjcMQ@mail.gmail.com>
 <CA+HVF0Axwivp6GebTXfdoaHwyDwGAupe4aHP4yM0Z1qg3m8DLQ@mail.gmail.com>
Message-ID: <CAL3TdOPpj-TFeW2dD8QOQiE_nmOWa-Vn8WC8SGh4mEWp8jMuqg@mail.gmail.com>

Pict should be making them just the same and they should convert fine
even if the GUI libraries aren't yet required. For example:

Welcome to Racket v6.1.0.8.
> (require file/convertible pict)
> (convertible? (disk 1))
#t
> (convert (disk 1) 'png-bytes)
#"\211PNG\r\n\32\n\0\0\0\rIHDR\0\0\0\1\0\0\0\1\b\6\0\0\0\37\25\304\211\0\0\0\rIDAT\b\231c```8\2\0\0\311\0\305\336\357!\202\0\0\0\0IEND\256B`\202"


But, that said, it may be easiest to get X going, because people who
submit programs can easily run into this problem too.

Robby


On Mon, Sep 22, 2014 at 12:17 AM, J Arcane <jarcane at gmail.com> wrote:
> Hmm.
>
> Alright, replacing the calls to slideshow with calls directly to pict, as
> well as commenting out the racket/gui/base calls successfully gets the code
> to run without X.
>
> However, something about how it converts the picts for web isn't working, so
> instead of a neat little circle, for instance, I get "(pict '(prog
> #<procedure:... etc." instead.
>
> I suspect this is probably something to do with a difference in behavior in
> how images are converted by the server, which does a convertible? test and
> then runs:
>  (run-code ev `(bytes-append #"data:image/png;base64,"
>                          (base64-encode (convert ,res 'png-bytes) #"")))
>
> At an idiot's guess, something is different about how pict is making the
> images, so convertible? is #f and the code just prints the (pict '(prog...
> stuff.
>
> I've set up a fork here: https://github.com/jarcane/try-racket
> I'm still working on getting the DNS set properly though, so no
> try-racket.org yet; the servlet is public at http://104.131.18.192:8000/ but
> I haven't figured out how to get try-racket.org to go there yet.
>
>
> <robby at eecs.northwestern.edu> wrote:
>>
>> Maybe you can use  slideshow/code-pict instead?
>>
>> Robby
>>
>> On Sun, Sep 21, 2014 at 6:00 PM, Marc Burns <m4burns at csclub.uwaterloo.ca>
>> wrote:
>> > The culprit is the slideshow/code module loaded into the sandbox
>> > evaluator. Requiring slideshow/code attempts to establish an X server
>> > connection.
>> >
>> > On Mon, Sep 22, 2014 at 12:12:37AM +0300, J Arcane wrote:
>> >> Greetings,
>> >>
>> >> For some time now, it's bothered me a bit that Racket doesn't have an
>> >> online REPL currently hosted anywhere. There's one written here:
>> >> https://github.com/voila/try-racket
>> >>
>> >> But no one's hosted it anywhere. So I took it upon myself to fix that.
>> >> I've
>> >> purchased try-racket.org and a basic DigitalOcean droplet to host it
>> >> on,
>> >> but I've hit a snag: I can't get it to run.
>> >>
>> >> It runs more or less without trouble on my personal FreeBSD box, but on
>> >> my
>> >> Debian 7 droplet I waded through countless dependency issues until
>> >> finally
>> >> I reached what seems to be this same error:
>> >>
>> >> http://bugs.racket-lang.org/query/?cmd=view%20audit-trail&pr=12465
>> >>
>> >> Presuming that perhaps this was in someway trying to run Racket in an
>> >> actual GUI window whether I wanted it to or not, I attempted to run it
>> >> as
>> >> --script, but this just quits silently instead, leaving no servlet
>> >> behind
>> >> either.
>> >>
>> >> Any ideas what I might be doing wrong? or have I hit a proper bug? (IME
>> >> it's usually the former ... )
>> >>
>> >> John Berry
>> >> http://jarcane.github.com
>> >
>> >> ____________________
>> >>   Racket Users list:
>> >>   http://lists.racket-lang.org/users
>> >
>> > ____________________
>> >   Racket Users list:
>> >   http://lists.racket-lang.org/users
>
>

From jay.mccarthy at gmail.com  Mon Sep 22 09:14:21 2014
From: jay.mccarthy at gmail.com (Jay McCarthy)
Date: Mon, 22 Sep 2014 09:14:21 -0400
Subject: [racket] ->i contract question
In-Reply-To: <7D717570-D4F1-4F93-A6D8-70A981189839@gmail.com>
References: <7D717570-D4F1-4F93-A6D8-70A981189839@gmail.com>
Message-ID: <CAJYbDak0d2x+9B0B4P2ZBVAUioyoTxHQ_N2sTnFQY_eO8Th_pA@mail.gmail.com>

As far as I can tell, there is no contract form that allows treating
the values returned as a list. Here is how you could implement one if
you wanted:

http://pasterack.org/pastes/46111

Maybe Robby has an opinion about whether this should be done in racket/contract

Jay

On Sun, Sep 21, 2014 at 1:47 AM, Kevin Forchione <lysseus at gmail.com> wrote:
> Hi guys,
> Is there a syntax for dependent-range that would allow 0 or more values to be specified? The documentation appears to describe only a specified number of values. Is there an analogue to list of and non-empty-listof?
>
> -Kevin
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users



-- 
Jay McCarthy
http://jeapostrophe.github.io

           "Wherefore, be not weary in well-doing,
      for ye are laying the foundation of a great work.
And out of small things proceedeth that which is great."
                          - D&C 64:33

From jay.mccarthy at gmail.com  Mon Sep 22 09:24:53 2014
From: jay.mccarthy at gmail.com (Jay McCarthy)
Date: Mon, 22 Sep 2014 09:24:53 -0400
Subject: [racket] Provide transformers
In-Reply-To: <CAAXAoJXbO+D0mPzAonTh+uNPGW6ajG0zSyeh8=gsFxVr5OQQyw@mail.gmail.com>
References: <CAAXAoJXbO+D0mPzAonTh+uNPGW6ajG0zSyeh8=gsFxVr5OQQyw@mail.gmail.com>
Message-ID: <CAJYbDa=WUBPENTKG-2tzYXhV3fu-iaZYsvSJgfvdsrO4zZxmGg@mail.gmail.com>

Is this what you are looking for?

#lang racket/base
(require (for-syntax racket/base
                     syntax/parse
                     racket/provide-transform))

(define-syntax test-pred-out
  (make-provide-pre-transformer
   (? (stx _)
     (syntax-parse stx
       [(_ b:id p:expr)
        (syntax-local-lift-module-end-declaration
         #'(module+ test
             (if (p b)
               (printf "It passed.\n")
               (printf "It didn't pass.\n"))))
        (syntax/loc stx
          b)]))))

;; Example

(define some-integer "17")
(define some-other-integer 17)
(provide (test-pred-out some-integer integer?)
         (test-pred-out some-other-integer integer?))

On Fri, Sep 19, 2014 at 5:41 PM, Jack Firth <jackhfirth at gmail.com> wrote:
> I've been looking into how to design a macro that would specify tests in a
> provide transformer, for example something like
>
>     (provide (test-pred-out some-integer integer?))
>
> I'd like that to expand to:
>
>     (provide some-integer)
>     (module+ test
>       (check-pred integer? some-integer))
>
> From what I can gather from the reference, normally provide forms are done
> with define-provide-syntax, but that macro doesn't let me splice anything
> into the module body. The reference mentions provide pre-transformers, and
> says:
>
> "A provide pre-transformer is applied as part of the first phase of a
> module?s expansion. Since it is used in the first phase, a provide
> pre-transformer can use functions such as syntax-local-lift-expression to
> introduce expressions and definitions in the enclosing module."
>
> So it sounds like what I want is definitely possible, but I have no idea how
> to actually do it. Does anyone have some simple well-explained examples to
> guide me in the right direction?
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>



-- 
Jay McCarthy
http://jeapostrophe.github.io

           "Wherefore, be not weary in well-doing,
      for ye are laying the foundation of a great work.
And out of small things proceedeth that which is great."
                          - D&C 64:33


From robby at eecs.northwestern.edu  Mon Sep 22 10:01:53 2014
From: robby at eecs.northwestern.edu (Robby Findler)
Date: Mon, 22 Sep 2014 09:01:53 -0500
Subject: [racket] ->i contract question
In-Reply-To: <CAJYbDak0d2x+9B0B4P2ZBVAUioyoTxHQ_N2sTnFQY_eO8Th_pA@mail.gmail.com>
References: <7D717570-D4F1-4F93-A6D8-70A981189839@gmail.com>
 <CAJYbDak0d2x+9B0B4P2ZBVAUioyoTxHQ_N2sTnFQY_eO8Th_pA@mail.gmail.com>
Message-ID: <CAL3TdOMuC9fCevSaA2GuqK3oZjm2pvzQdP91M=BDQykR7mb+=A@mail.gmail.com>

Turning things back and forth between values and lists is not
something that our apis and our runtime infrastructure really
encourages. So I'd say that the first thing to try would be to either
fix the number of values returned from the function or to start
programming with lists instead.

If that doesn't work, then it is certainly possible to make function
contract combinators that implicitly do the call-with-values like that
and then list you specify the range contract kind of like how you
specify the #:rest contract. I'd be happy to help guide someone who
wanted to try to use the contract library to attempt that. It would be
a good exercise to learn about contracts and macros, no doubt!

Robby


On Mon, Sep 22, 2014 at 8:14 AM, Jay McCarthy <jay.mccarthy at gmail.com> wrote:
> As far as I can tell, there is no contract form that allows treating
> the values returned as a list. Here is how you could implement one if
> you wanted:
>
> http://pasterack.org/pastes/46111
>
> Maybe Robby has an opinion about whether this should be done in racket/contract
>
> Jay
>
> On Sun, Sep 21, 2014 at 1:47 AM, Kevin Forchione <lysseus at gmail.com> wrote:
>> Hi guys,
>> Is there a syntax for dependent-range that would allow 0 or more values to be specified? The documentation appears to describe only a specified number of values. Is there an analogue to list of and non-empty-listof?
>>
>> -Kevin
>> ____________________
>>   Racket Users list:
>>   http://lists.racket-lang.org/users
>
>
>
> --
> Jay McCarthy
> http://jeapostrophe.github.io
>
>            "Wherefore, be not weary in well-doing,
>       for ye are laying the foundation of a great work.
> And out of small things proceedeth that which is great."
>                           - D&C 64:33
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users

From lysseus at gmail.com  Mon Sep 22 12:05:00 2014
From: lysseus at gmail.com (Kevin Forchione)
Date: Mon, 22 Sep 2014 09:05:00 -0700
Subject: [racket]  keyword-apply questions
Message-ID: <E857EBAD-BB41-43C6-910F-8FC8B4A8B6A4@gmail.com>

Hi guys,
Apparently, with a desire to use keywords more in functions, I?ll have to wrap my head around keyword-apply better. Two questions I have are:

1. Why must the  kw-lst  be sorted in keyword<? rather than having those arguments appear in any order that makes sense to the caller, and then having keyword-apply sort both kw-lst and kw-val-lst appropriately to keep the key/value pairs synchronized?

2. The reference says:

(keyword-apply proc
				kw-lst
				kw-val-lst
				v ?
				lst
				#:<kw> kw-arg ?)

After playing around a bit all of the following are equivalent:

(define (bar #:x x #:y y #:z z u v . w) (list x y z u v w))

(keyword-apply bar '() '() '(6 7 8 9 10) #:x 3 #:y 4 #:z 5)
(keyword-apply bar '(#:x #:y #:z) '(3 4 5) '(6 7 8 9 10))
(keyword-apply bar #:x 3 #:y 4 #:z 5 '() '() '(6 7 8 9 10))
(keyword-apply bar #:z 5 #:y 4 #:x 3 '() '() '(6 7 8 9 10))
(keyword-apply bar #:z 5 #:x 3 '(#:y) '(4) '(6 7 8 9 10))
(keyword-apply bar #:z 5 #:x 3 '(#:y) '(4) 6 7 '(8 9 10))

So it appears that ordering of keywords is only mandatory in the kw-lst. What?s the advantage to using the kw-lst/kw-val-lst? If those lists are primarily built using automation, rather than by hand, then question #1 is probably explained by the idea that the automation is expected to sort the keywords prior to the call?

-Kevin

From lysseus at gmail.com  Mon Sep 22 14:24:18 2014
From: lysseus at gmail.com (Kevin Forchione)
Date: Mon, 22 Sep 2014 11:24:18 -0700
Subject: [racket]  keyword-apply questions
In-Reply-To: <E857EBAD-BB41-43C6-910F-8FC8B4A8B6A4@gmail.com>
References: <E857EBAD-BB41-43C6-910F-8FC8B4A8B6A4@gmail.com>
Message-ID: <AAF9175F-2894-4902-BEE7-467CC18A7559@gmail.com>


On Sep 22, 2014, at 9:05 AM, Kevin Forchione <lysseus at gmail.com> wrote:

> Hi guys,
> Apparently, with a desire to use keywords more in functions, I?ll have to wrap my head around keyword-apply better. Two questions I have are:
> 
> 1. Why must the  kw-lst  be sorted in keyword<? rather than having those arguments appear in any order that makes sense to the caller, and then having keyword-apply sort both kw-lst and kw-val-lst appropriately to keep the key/value pairs synchronized?
> 
> 2. The reference says:
> 
> (keyword-apply proc
> 				kw-lst
> 				kw-val-lst
> 				v ?
> 				lst
> 				#:<kw> kw-arg ?)
> 
> After playing around a bit all of the following are equivalent:
> 
> (define (bar #:x x #:y y #:z z u v . w) (list x y z u v w))
> 
> (keyword-apply bar '() '() '(6 7 8 9 10) #:x 3 #:y 4 #:z 5)
> (keyword-apply bar '(#:x #:y #:z) '(3 4 5) '(6 7 8 9 10))
> (keyword-apply bar #:x 3 #:y 4 #:z 5 '() '() '(6 7 8 9 10))
> (keyword-apply bar #:z 5 #:y 4 #:x 3 '() '() '(6 7 8 9 10))
> (keyword-apply bar #:z 5 #:x 3 '(#:y) '(4) '(6 7 8 9 10))
> (keyword-apply bar #:z 5 #:x 3 '(#:y) '(4) 6 7 '(8 9 10))
> 
> So it appears that ordering of keywords is only mandatory in the kw-lst. What?s the advantage to using the kw-lst/kw-val-lst? If those lists are primarily built using automation, rather than by hand, then question #1 is probably explained by the idea that the automation is expected to sort the keywords prior to the call?
> 
> -Kevin

Hmmm? ok, looks like the kws-lst and kvs-lst serve the similar function as common lisp?s &allow-other-keys, which allows passing keyword/keyword values as arguments to functions not directly defining them. 

-Kevin

From martindemello at gmail.com  Mon Sep 22 14:49:55 2014
From: martindemello at gmail.com (Martin DeMello)
Date: Mon, 22 Sep 2014 11:49:55 -0700
Subject: [racket] is anyone using racket to generate animated diagrams?
Message-ID: <CAFrFfuFHTK=nizbHCM+0nUX5LjVp7odPYwCgQFk6dG+Vey3K3w@mail.gmail.com>

I was looking through http://en.wikipedia.org/wiki/User:LucasVB/Gallery
(generated, according to the author, with a combination of php and libgd),
and wondered if anyone was doing similar stuff in racket. I'd love to see
code samples if so.

martin
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140922/f9ff2fd3/attachment.html>

From jay.mccarthy at gmail.com  Mon Sep 22 15:05:27 2014
From: jay.mccarthy at gmail.com (Jay McCarthy)
Date: Mon, 22 Sep 2014 15:05:27 -0400
Subject: [racket] Hosting the try-racket REPL.
In-Reply-To: <CA+HVF0Axwivp6GebTXfdoaHwyDwGAupe4aHP4yM0Z1qg3m8DLQ@mail.gmail.com>
References: <CA+HVF0Djjj+mA-BYUeBNwYXOt+ELJt7CixdrO25S7B2Ktgeo+Q@mail.gmail.com>
 <20140921230002.GD29774@happierface>
 <CAL3TdOPyujV7ork7Y+r+q---7w1WBpO-GN=H7wKDWK0yknjcMQ@mail.gmail.com>
 <CA+HVF0Axwivp6GebTXfdoaHwyDwGAupe4aHP4yM0Z1qg3m8DLQ@mail.gmail.com>
Message-ID: <CAJYbDanoXRAhnSA5vW8mF3wJSHA8AxJ7EsT8q9Rt01OUVxTGRA@mail.gmail.com>

J and I chatted a little and he got his DNS working and also I added

http://try.racket-lang.org/

as an alias for it.

In general, the Racket team is willing to give out subdomains of
r-l.org for projects that would like them, so they don't have to buy
their own DNS addresses. Just email the list or one of us directly to
ask.

Jay

On Mon, Sep 22, 2014 at 1:17 AM, J Arcane <jarcane at gmail.com> wrote:
> Hmm.
>
> Alright, replacing the calls to slideshow with calls directly to pict, as
> well as commenting out the racket/gui/base calls successfully gets the code
> to run without X.
>
> However, something about how it converts the picts for web isn't working, so
> instead of a neat little circle, for instance, I get "(pict '(prog
> #<procedure:... etc." instead.
>
> I suspect this is probably something to do with a difference in behavior in
> how images are converted by the server, which does a convertible? test and
> then runs:
>  (run-code ev `(bytes-append #"data:image/png;base64,"
>                          (base64-encode (convert ,res 'png-bytes) #"")))
>
> At an idiot's guess, something is different about how pict is making the
> images, so convertible? is #f and the code just prints the (pict '(prog...
> stuff.
>
> I've set up a fork here: https://github.com/jarcane/try-racket
> I'm still working on getting the DNS set properly though, so no
> try-racket.org yet; the servlet is public at http://104.131.18.192:8000/ but
> I haven't figured out how to get try-racket.org to go there yet.
>
>
> <robby at eecs.northwestern.edu> wrote:
>>
>> Maybe you can use  slideshow/code-pict instead?
>>
>> Robby
>>
>> On Sun, Sep 21, 2014 at 6:00 PM, Marc Burns <m4burns at csclub.uwaterloo.ca>
>> wrote:
>> > The culprit is the slideshow/code module loaded into the sandbox
>> > evaluator. Requiring slideshow/code attempts to establish an X server
>> > connection.
>> >
>> > On Mon, Sep 22, 2014 at 12:12:37AM +0300, J Arcane wrote:
>> >> Greetings,
>> >>
>> >> For some time now, it's bothered me a bit that Racket doesn't have an
>> >> online REPL currently hosted anywhere. There's one written here:
>> >> https://github.com/voila/try-racket
>> >>
>> >> But no one's hosted it anywhere. So I took it upon myself to fix that.
>> >> I've
>> >> purchased try-racket.org and a basic DigitalOcean droplet to host it
>> >> on,
>> >> but I've hit a snag: I can't get it to run.
>> >>
>> >> It runs more or less without trouble on my personal FreeBSD box, but on
>> >> my
>> >> Debian 7 droplet I waded through countless dependency issues until
>> >> finally
>> >> I reached what seems to be this same error:
>> >>
>> >> http://bugs.racket-lang.org/query/?cmd=view%20audit-trail&pr=12465
>> >>
>> >> Presuming that perhaps this was in someway trying to run Racket in an
>> >> actual GUI window whether I wanted it to or not, I attempted to run it
>> >> as
>> >> --script, but this just quits silently instead, leaving no servlet
>> >> behind
>> >> either.
>> >>
>> >> Any ideas what I might be doing wrong? or have I hit a proper bug? (IME
>> >> it's usually the former ... )
>> >>
>> >> John Berry
>> >> http://jarcane.github.com
>> >
>> >> ____________________
>> >>   Racket Users list:
>> >>   http://lists.racket-lang.org/users
>> >
>> > ____________________
>> >   Racket Users list:
>> >   http://lists.racket-lang.org/users
>
>
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>



-- 
Jay McCarthy
http://jeapostrophe.github.io

           "Wherefore, be not weary in well-doing,
      for ye are laying the foundation of a great work.
And out of small things proceedeth that which is great."
                          - D&C 64:33

From jensaxel at soegaard.net  Mon Sep 22 15:44:51 2014
From: jensaxel at soegaard.net (=?UTF-8?Q?Jens_Axel_S=C3=B8gaard?=)
Date: Mon, 22 Sep 2014 21:44:51 +0200
Subject: [racket] is anyone using racket to generate animated diagrams?
In-Reply-To: <CAFrFfuFHTK=nizbHCM+0nUX5LjVp7odPYwCgQFk6dG+Vey3K3w@mail.gmail.com>
References: <CAFrFfuFHTK=nizbHCM+0nUX5LjVp7odPYwCgQFk6dG+Vey3K3w@mail.gmail.com>
Message-ID: <CABefVgz58+HyFcrzFxBBUspCK3dHY3ioMYHiNSaW=Jd_TZPO0g@mail.gmail.com>

Hi Martin,

As an example I implemented the standing wave animation.

The MetaPict package is available through the package system.
Use the package manager in DrRacket to install install it.
The run the program below in DrRacket to see a standing wave.

The documentation of the MetaPict package is here:
    http://soegaard.github.io/docs/metapict/metapict.html

Enjoy,
Jens Axel

#lang racket
(require metapict
         metapict/graph
         (only-in 2htdp/universe animate)
         (only-in 2htdp/image freeze))

(set-curve-pict-size 400 100)  ; size 400 x 100 pixels

; red-circle : point -> pict
;  red disk with center in the point p drawn with black periphery
(define (red-circle p)
  (def c (circle p 0.1))
  (draw (color "red" (fill c)) ; red, filled circle
        c))                    ; black outline


(define N 100) ; number of frames in the animation
(define (f m)
  ; The function to graph at time t is cos(?t)*sin(kx), where t is in [0;2pi].
  ; First we convert the frame number to time:
  (define t (* 2 pi (/ (remainder m N) N)))
  (define ? 1) ; angular frequency
  (define k 1) ; wave number
  ; The actual function to graph
  (? (x) (* (cos (* ? t)) (sin (* k x)))))


(define (animation-frame m)
  (with-window (window/aspect -0.2 (+ (* 4 pi) 0.2) ; x in [-0.1, 4pi+0.1]
                              -1.2)             ; y in [-0.1, 1.1]
    (pict->bitmap ; animate needs bitmaps
     (draw
      ; graph the function f from 0 to 4pi
      (graph (f m) 0 (* 4 pi))
      ; five red dots at the nodes (0,0), (pi,0), ...
      (for/draw ([n 5])
          (red-circle (pt (* n pi) 0)))))))

; animate them!
(animate animation-frame)



2014-09-22 20:49 GMT+02:00 Martin DeMello <martindemello at gmail.com>:
> I was looking through http://en.wikipedia.org/wiki/User:LucasVB/Gallery
> (generated, according to the author, with a combination of php and libgd),
> and wondered if anyone was doing similar stuff in racket. I'd love to see
> code samples if so.
>
> martin
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>



-- 
--
Jens Axel S?gaard


From martindemello at gmail.com  Mon Sep 22 16:06:54 2014
From: martindemello at gmail.com (Martin DeMello)
Date: Mon, 22 Sep 2014 13:06:54 -0700
Subject: [racket] is anyone using racket to generate animated diagrams?
In-Reply-To: <CABefVgz58+HyFcrzFxBBUspCK3dHY3ioMYHiNSaW=Jd_TZPO0g@mail.gmail.com>
References: <CAFrFfuFHTK=nizbHCM+0nUX5LjVp7odPYwCgQFk6dG+Vey3K3w@mail.gmail.com>
 <CABefVgz58+HyFcrzFxBBUspCK3dHY3ioMYHiNSaW=Jd_TZPO0g@mail.gmail.com>
Message-ID: <CAFrFfuFEN0bSKXyE+L1bJ6yeXUmhm-FTFg8vn0mgz82ug3TyAQ@mail.gmail.com>

Thanks, metapict looks great!

martin

On Mon, Sep 22, 2014 at 12:44 PM, Jens Axel S?gaard <jensaxel at soegaard.net>
wrote:

> Hi Martin,
>
> As an example I implemented the standing wave animation.
>
> The MetaPict package is available through the package system.
> Use the package manager in DrRacket to install install it.
> The run the program below in DrRacket to see a standing wave.
>
> The documentation of the MetaPict package is here:
>     http://soegaard.github.io/docs/metapict/metapict.html
>
> Enjoy,
> Jens Axel
>
> #lang racket
> (require metapict
>          metapict/graph
>          (only-in 2htdp/universe animate)
>          (only-in 2htdp/image freeze))
>
> (set-curve-pict-size 400 100)  ; size 400 x 100 pixels
>
> ; red-circle : point -> pict
> ;  red disk with center in the point p drawn with black periphery
> (define (red-circle p)
>   (def c (circle p 0.1))
>   (draw (color "red" (fill c)) ; red, filled circle
>         c))                    ; black outline
>
>
> (define N 100) ; number of frames in the animation
> (define (f m)
>   ; The function to graph at time t is cos(?t)*sin(kx), where t is in
> [0;2pi].
>   ; First we convert the frame number to time:
>   (define t (* 2 pi (/ (remainder m N) N)))
>   (define ? 1) ; angular frequency
>   (define k 1) ; wave number
>   ; The actual function to graph
>   (? (x) (* (cos (* ? t)) (sin (* k x)))))
>
>
> (define (animation-frame m)
>   (with-window (window/aspect -0.2 (+ (* 4 pi) 0.2) ; x in [-0.1, 4pi+0.1]
>                               -1.2)             ; y in [-0.1, 1.1]
>     (pict->bitmap ; animate needs bitmaps
>      (draw
>       ; graph the function f from 0 to 4pi
>       (graph (f m) 0 (* 4 pi))
>       ; five red dots at the nodes (0,0), (pi,0), ...
>       (for/draw ([n 5])
>           (red-circle (pt (* n pi) 0)))))))
>
> ; animate them!
> (animate animation-frame)
>
>
>
> 2014-09-22 20:49 GMT+02:00 Martin DeMello <martindemello at gmail.com>:
> > I was looking through http://en.wikipedia.org/wiki/User:LucasVB/Gallery
> > (generated, according to the author, with a combination of php and
> libgd),
> > and wondered if anyone was doing similar stuff in racket. I'd love to see
> > code samples if so.
> >
> > martin
> >
> > ____________________
> >   Racket Users list:
> >   http://lists.racket-lang.org/users
> >
>
>
>
> --
> --
> Jens Axel S?gaard
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140922/18053100/attachment-0001.html>

From jsgrahamus at yahoo.com  Mon Sep 22 17:11:25 2014
From: jsgrahamus at yahoo.com (Steve Graham)
Date: Mon, 22 Sep 2014 14:11:25 -0700
Subject: [racket] Hosting the try-racket REPL.
In-Reply-To: <CAJYbDanoXRAhnSA5vW8mF3wJSHA8AxJ7EsT8q9Rt01OUVxTGRA@mail.gmail.com>
References: <CA+HVF0Djjj+mA-BYUeBNwYXOt+ELJt7CixdrO25S7B2Ktgeo+Q@mail.gmail.com>
 <20140921230002.GD29774@happierface>
 <CAL3TdOPyujV7ork7Y+r+q---7w1WBpO-GN=H7wKDWK0yknjcMQ@mail.gmail.com>
 <CA+HVF0Axwivp6GebTXfdoaHwyDwGAupe4aHP4yM0Z1qg3m8DLQ@mail.gmail.com>
 <CAJYbDanoXRAhnSA5vW8mF3wJSHA8AxJ7EsT8q9Rt01OUVxTGRA@mail.gmail.com>
Message-ID: <1411420285.72632.YahooMailNeo@web121405.mail.ne1.yahoo.com>

I get errors when trying to type in a multi-line function.  Am I doing something wrong?



________________________________
 From: Jay McCarthy <jay.mccarthy at gmail.com>
To: J Arcane <jarcane at gmail.com> 
Cc: Racket Users <users at racket-lang.org>; Robby Findler <robby at eecs.northwestern.edu>; Marc Burns <m4burns at csclub.uwaterloo.ca> 
Sent: Monday, September 22, 2014 12:05 PM
Subject: Re: [racket] Hosting the try-racket REPL.
 

J and I chatted a little and he got his DNS working and also I added

http://try.racket-lang.org/

as an alias for it.

In general, the Racket team is willing to give out subdomains of
r-l.org for projects that would like them, so they don't have to buy
their own DNS addresses. Just email the list or one of us directly to
ask.

Jay

On Mon, Sep 22, 2014 at 1:17 AM, J Arcane <jarcane at gmail.com> wrote:
> Hmm.
>
> Alright, replacing the calls to slideshow with calls directly to pict, as
> well as commenting out the racket/gui/base calls successfully gets the code
> to run without X.
>
> However, something about how it converts the picts for web isn't working, so
> instead of a neat little circle, for instance, I get "(pict '(prog
> #<procedure:... etc." instead.
>
> I suspect this is probably something to do with a difference in behavior in
> how images are converted by the server, which does a convertible? test and
> then runs:
>  (run-code ev `(bytes-append #"data:image/png;base64,"
>                          (base64-encode (convert ,res 'png-bytes) #"")))
>
> At an idiot's guess, something is different about how pict is making the
> images, so convertible? is #f and the code just prints the (pict '(prog...
> stuff.
>
> I've set up a fork here: https://github.com/jarcane/try-racket
> I'm still working on getting the DNS set properly though, so no
> try-racket.org yet; the servlet is public at http://104.131.18.192:8000/ but
> I haven't figured out how to get try-racket.org to go there yet.
>
>
> <robby at eecs.northwestern.edu> wrote:
>>
>> Maybe you can use  slideshow/code-pict instead?
>>
>> Robby
>>
>> On Sun, Sep 21, 2014 at 6:00 PM, Marc Burns <m4burns at csclub.uwaterloo.ca>
>> wrote:
>> > The culprit is the slideshow/code module loaded into the sandbox
>> > evaluator. Requiring slideshow/code attempts to establish an X server
>> > connection.
>> >
>> > On Mon, Sep 22, 2014 at 12:12:37AM +0300, J Arcane wrote:
>> >> Greetings,
>> >>
>> >> For some time now, it's bothered me a bit that Racket doesn't have an
>> >> online REPL currently hosted anywhere. There's one written here:
>> >> https://github.com/voila/try-racket
>> >>
>> >> But no one's hosted it anywhere. So I took it upon myself to fix that.
>> >> I've
>> >> purchased try-racket.org and a basic DigitalOcean droplet to host it
>> >> on,
>> >> but I've hit a snag: I can't get it to run.
>> >>
>> >> It runs more or less without trouble on my personal FreeBSD box, but on
>> >> my
>> >> Debian 7 droplet I waded through countless dependency issues until
>> >> finally
>> >> I reached what seems to be this same error:
>> >>
>> >> http://bugs.racket-lang.org/query/?cmd=view%20audit-trail?=12465
>> >>
>> >> Presuming that perhaps this was in someway trying to run Racket in an
>> >> actual GUI window whether I wanted it to or not, I attempted to run it
>> >> as
>> >> --script, but this just quits silently instead, leaving no servlet
>> >> behind
>> >> either.
>> >>
>> >> Any ideas what I might be doing wrong? or have I hit a proper bug? (IME
>> >> it's usually the former ... )
>> >>
>> >> John Berry
>> >> http://jarcane.github.com
>> >
>> >> ____________________
>> >>   Racket Users list:
>> >>  http://lists.racket-lang.org/users
>> >
>> > ____________________
>> >   Racket Users list:
>> >  http://lists.racket-lang.org/users
>
>
>
> ____________________
>   Racket Users list:
>  http://lists.racket-lang.org/users
>



-- 
Jay McCarthy
http://jeapostrophe.github.io

           "Wherefore, be not weary in well-doing,
      for ye are laying the foundation of a great work.
And out of small things proceedeth that which is great."
                          - D&C 64:33



____________________
  Racket Users list:
  http://lists.racket-lang.org/users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140922/3e4f6ee8/attachment.html>

From clements at brinckerhoff.org  Sun Sep 21 20:52:56 2014
From: clements at brinckerhoff.org (John Clements)
Date: Sun, 21 Sep 2014 17:52:56 -0700
Subject: [racket] Rendering the PLT symbol in Racket?
In-Reply-To: <989C0D02-3EED-40CB-8DB6-E4B36B6DFC4C@knauth.org>
References: <CAFKxZVUvUD=AHp1bEJO0soUHjfUKLZJhTFdKNcKWWM0t=maqmw@mail.gmail.com>
 <989C0D02-3EED-40CB-8DB6-E4B36B6DFC4C@knauth.org>
Message-ID: <E3D0D163-C3E4-4350-9144-609B617D3F1F@brinckerhoff.org>

There are a whole bunch of ways of answering this question; you might want Neil?s fantastic ray-traced version, or just the simple paths. In case it?s helpful, here?s the original postscript file generated by Shriram, Robby, and Matthew. Oh, wait, here?s the comment from the file:

%%Title: PLT Lambda
%%Creator: Matthew, then Shriram, finally Robby
%%CreationDate: June 24, 1997
%%For: PLT

(NB: Apple Mail tries to render this as an image, pretty badly. Save the attachment to read it.)

-------------- next part --------------
A non-text attachment was scrubbed...
Name: plt.ps
Type: image/ps
Size: 1899 bytes
Desc: not available
URL: <http://lists.racket-lang.org/users/archive/attachments/20140921/3febf5c2/attachment.bin>
-------------- next part --------------

On Sep 21, 2014, at 4:59 PM, Alexander D. Knauth <alexander at knauth.org> wrote:

> I tried it with the images in the email, but it was too big, so here it is just links.
> I think there was something on the list about this before, and I found these again from that:
> http://docs.racket-lang.org/images/Logos.html?q=plt-logo#%28def._%28%28lib._images%2Flogos..rkt%29._plt-logo%29%29
> (require images/logos)
> (plt-logo)
> and also
> http://docs.racket-lang.org/draw/overview.html?q=racket%2Fdraw#%28part._.Drawing_.Paths%29 (down a little bit)
> 
> 
> On Sep 21, 2014, at 12:52 AM, Daniel Prager <daniel.a.prager at gmail.com> wrote:
> 
>> There was interest following my talk at RacketCon in a www.youpatch.com quilt pattern featuring the PLT symbol (file:///Users/dan/Documents/plt.svg).
>> 
>> I was having a play with youpatch.com by standard means, to put together a commemorative pattern, but I think I can do a cleaner job given access to Racket code that generates the image.
>> 
>> Can anyone point me to, or send me the code?
>> 
>> 
>> Dan
>> ____________________
>>  Racket Users list:
>>  http://lists.racket-lang.org/users
> ____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users


From jarcane at gmail.com  Tue Sep 23 00:35:44 2014
From: jarcane at gmail.com (J Arcane)
Date: Tue, 23 Sep 2014 07:35:44 +0300
Subject: [racket] Hosting the try-racket REPL.
In-Reply-To: <1411420285.72632.YahooMailNeo@web121405.mail.ne1.yahoo.com>
References: <CA+HVF0Djjj+mA-BYUeBNwYXOt+ELJt7CixdrO25S7B2Ktgeo+Q@mail.gmail.com>
 <20140921230002.GD29774@happierface>
 <CAL3TdOPyujV7ork7Y+r+q---7w1WBpO-GN=H7wKDWK0yknjcMQ@mail.gmail.com>
 <CA+HVF0Axwivp6GebTXfdoaHwyDwGAupe4aHP4yM0Z1qg3m8DLQ@mail.gmail.com>
 <CAJYbDanoXRAhnSA5vW8mF3wJSHA8AxJ7EsT8q9Rt01OUVxTGRA@mail.gmail.com>
 <1411420285.72632.YahooMailNeo@web121405.mail.ne1.yahoo.com>
Message-ID: <CA+HVF0Ak38drhaGze_9=12KvK8HAjT8YVCwFynr3o2jMFXmTrg@mail.gmail.com>

It's a known issue. Alex Knauth has submitted a rough solution to my fork
of it; I'll get it merged and pushed to the server shortly.

On Tue, Sep 23, 2014 at 12:11 AM, Steve Graham <jsgrahamus at yahoo.com> wrote:

> I get errors when trying to type in a multi-line function.  Am I doing
> something wrong?
>
>   ------------------------------
>  *From:* Jay McCarthy <jay.mccarthy at gmail.com>
> *To:* J Arcane <jarcane at gmail.com>
> *Cc:* Racket Users <users at racket-lang.org>; Robby Findler <
> robby at eecs.northwestern.edu>; Marc Burns <m4burns at csclub.uwaterloo.ca>
> *Sent:* Monday, September 22, 2014 12:05 PM
> *Subject:* Re: [racket] Hosting the try-racket REPL.
>
> J and I chatted a little and he got his DNS working and also I added
>
> http://try.racket-lang.org/
>
> as an alias for it.
>
> In general, the Racket team is willing to give out subdomains of
> r-l.org for projects that would like them, so they don't have to buy
> their own DNS addresses. Just email the list or one of us directly to
> ask.
>
> Jay
>
> On Mon, Sep 22, 2014 at 1:17 AM, J Arcane <jarcane at gmail.com> wrote:
> > Hmm.
> >
> > Alright, replacing the calls to slideshow with calls directly to pict, as
> > well as commenting out the racket/gui/base calls successfully gets the
> code
> > to run without X.
> >
> > However, something about how it converts the picts for web isn't
> working, so
> > instead of a neat little circle, for instance, I get "(pict '(prog
> > #<procedure:... etc." instead.
> >
> > I suspect this is probably something to do with a difference in behavior
> in
> > how images are converted by the server, which does a convertible? test
> and
> > then runs:
> >  (run-code ev `(bytes-append #"data:image/png;base64,"
> >                          (base64-encode (convert ,res 'png-bytes) #"")))
> >
> > At an idiot's guess, something is different about how pict is making the
> > images, so convertible? is #f and the code just prints the (pict
> '(prog...
> > stuff.
> >
> > I've set up a fork here: https://github.com/jarcane/try-racket
> > I'm still working on getting the DNS set properly though, so no
> > try-racket.org yet; the servlet is public at http://104.131.18.192:8000/
> but
> > I haven't figured out how to get try-racket.org to go there yet.
> >
> >
> > <robby at eecs.northwestern.edu> wrote:
> >>
> >> Maybe you can use  slideshow/code-pict instead?
> >>
> >> Robby
> >>
> >> On Sun, Sep 21, 2014 at 6:00 PM, Marc Burns <
> m4burns at csclub.uwaterloo.ca>
> >> wrote:
> >> > The culprit is the slideshow/code module loaded into the sandbox
> >> > evaluator. Requiring slideshow/code attempts to establish an X server
> >> > connection.
> >> >
> >> > On Mon, Sep 22, 2014 at 12:12:37AM +0300, J Arcane wrote:
> >> >> Greetings,
> >> >>
> >> >> For some time now, it's bothered me a bit that Racket doesn't have an
> >> >> online REPL currently hosted anywhere. There's one written here:
> >> >> https://github.com/voila/try-racket
> >> >>
> >> >> But no one's hosted it anywhere. So I took it upon myself to fix
> that.
> >> >> I've
> >> >> purchased try-racket.org and a basic DigitalOcean droplet to host it
> >> >> on,
> >> >> but I've hit a snag: I can't get it to run.
> >> >>
> >> >> It runs more or less without trouble on my personal FreeBSD box, but
> on
> >> >> my
> >> >> Debian 7 droplet I waded through countless dependency issues until
> >> >> finally
> >> >> I reached what seems to be this same error:
> >> >>
> >> >> http://bugs.racket-lang.org/query/?cmd=view%20audit-trail?=12465
> <http://bugs.racket-lang.org/query/?cmd=view%20audit-trail&pr=12465>
>
> >> >>
> >> >> Presuming that perhaps this was in someway trying to run Racket in an
> >> >> actual GUI window whether I wanted it to or not, I attempted to run
> it
> >> >> as
> >> >> --script, but this just quits silently instead, leaving no servlet
> >> >> behind
> >> >> either.
> >> >>
> >> >> Any ideas what I might be doing wrong? or have I hit a proper bug?
> (IME
> >> >> it's usually the former ... )
> >> >>
> >> >> John Berry
> >> >> http://jarcane.github.com
> >> >
> >> >> ____________________
> >> >>  Racket Users list:
> >> >>  http://lists.racket-lang.org/users
> >> >
> >> > ____________________
> >> >  Racket Users list:
> >> >  http://lists.racket-lang.org/users
> >
> >
> >
> > ____________________
> >  Racket Users list:
> >  http://lists.racket-lang.org/users
> >
>
>
>
> --
> Jay McCarthy
> http://jeapostrophe.github.io
>
>           "Wherefore, be not weary in well-doing,
>       for ye are laying the foundation of a great work.
> And out of small things proceedeth that which is great."
>                           - D&C 64:33
>
>
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>
>
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140923/3176579e/attachment-0001.html>

From gneuner2 at comcast.net  Tue Sep 23 16:42:24 2014
From: gneuner2 at comcast.net (George Neuner)
Date: Tue, 23 Sep 2014 16:42:24 -0400
Subject: [racket] web server: module servlets
Message-ID: <5421DB30.3030803@comcast.net>

Hi,

I'm using 6.0.1 and I have a couple of questions regarding module
servlets.  I'm new to the server side of Racket so please be gentle.


1) Is there any way to speed up initial servlet loading?  Even for a
very simple servlet (does nothing, canned response for testing)
several seconds elapse before it responds to the browser.  I tried
compiling it to bytecode with raco, but having the ./compiled/.zo file
available doesn't seem to make any noticeable difference.  I know the
bytecode still is JIT'd before execution, but I thought eliminating
the source compile step would have helped more.

Does a web server require the .zo file to be in a different location
than ./compiled?  Or does it not deal with bytecode files at all?


2) Debugging module servlets is a pain due to having to stop/restart
the server.  Is there some clever way to force the server to unload a
particular servlet but keep running?

I thought about starting the listener in a thread, but I can't figure
how to get at either the thread or the custodian from within a
servlet.  Would listener termination have to be done from a hardcoded
function at the top level?


Apologies if these questions are stupid.

Thanks,
George


From jay.mccarthy at gmail.com  Tue Sep 23 18:04:58 2014
From: jay.mccarthy at gmail.com (Jay McCarthy)
Date: Tue, 23 Sep 2014 18:04:58 -0400
Subject: [racket] web server: module servlets
In-Reply-To: <5421DB30.3030803@comcast.net>
References: <5421DB30.3030803@comcast.net>
Message-ID: <CAJYbDa=uR26UAdfPgg5RzLo3HJv_rWosXXiCx+Rp9vbxPEbe9Q@mail.gmail.com>

On Tue, Sep 23, 2014 at 4:42 PM, George Neuner <gneuner2 at comcast.net> wrote:
> Hi,
>
> I'm using 6.0.1 and I have a couple of questions regarding module
> servlets.  I'm new to the server side of Racket so please be gentle.
>
>
> 1) Is there any way to speed up initial servlet loading?  Even for a
> very simple servlet (does nothing, canned response for testing)
> several seconds elapse before it responds to the browser.  I tried
> compiling it to bytecode with raco, but having the ./compiled/.zo file
> available doesn't seem to make any noticeable difference.  I know the
> bytecode still is JIT'd before execution, but I thought eliminating
> the source compile step would have helped more.
>
> Does a web server require the .zo file to be in a different location
> than ./compiled?  Or does it not deal with bytecode files at all?
>

This is why I recommend that new users use serve/servlet, which is
much faster because it does not use namespaces or live loading.

> 2) Debugging module servlets is a pain due to having to stop/restart
> the server.  Is there some clever way to force the server to unload a
> particular servlet but keep running?
>
> I thought about starting the listener in a thread, but I can't figure
> how to get at either the thread or the custodian from within a
> servlet.  Would listener termination have to be done from a hardcoded
> function at the top level?

Doing this pretty much requires not following my advice in question 1,
but the default server (used by the command line tool) allows you to
go to "/conf/refresh-servlets" and unload/reload all the servlet code.

The way that you would know that is by following the link from the
command-line docs

http://docs.racket-lang.org/web-server/run.html#%28part._command-line-tools%29

to the web-server@ unit which is the implementation of the dispatch
sequence the command-line tool uses:

http://docs.racket-lang.org/web-server-internal/Web_Servers.html#%28def._%28%28lib._web-server%2Fweb-server-unit..rkt%29._web-server~40%29%29

>
> Apologies if these questions are stupid.
>
> Thanks,
> George
>
> ____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users



-- 
Jay McCarthy
http://jeapostrophe.github.io

           "Wherefore, be not weary in well-doing,
      for ye are laying the foundation of a great work.
And out of small things proceedeth that which is great."
                          - D&C 64:33

From jsgrahamus at yahoo.com  Tue Sep 23 20:32:23 2014
From: jsgrahamus at yahoo.com (Steve Graham)
Date: Tue, 23 Sep 2014 17:32:23 -0700
Subject: [racket] Hosting the try-racket REPL.
In-Reply-To: <CA+HVF0Ak38drhaGze_9=12KvK8HAjT8YVCwFynr3o2jMFXmTrg@mail.gmail.com>
References: <CA+HVF0Djjj+mA-BYUeBNwYXOt+ELJt7CixdrO25S7B2Ktgeo+Q@mail.gmail.com>	<20140921230002.GD29774@happierface>	<CAL3TdOPyujV7ork7Y+r+q---7w1WBpO-GN=H7wKDWK0yknjcMQ@mail.gmail.com>	<CA+HVF0Axwivp6GebTXfdoaHwyDwGAupe4aHP4yM0Z1qg3m8DLQ@mail.gmail.com>	<CAJYbDanoXRAhnSA5vW8mF3wJSHA8AxJ7EsT8q9Rt01OUVxTGRA@mail.gmail.com>	<1411420285.72632.YahooMailNeo@web121405.mail.ne1.yahoo.com>
 <CA+HVF0Ak38drhaGze_9=12KvK8HAjT8YVCwFynr3o2jMFXmTrg@mail.gmail.com>
Message-ID: <1411518743.40110.YahooMailNeo@web121406.mail.ne1.yahoo.com>

Thanks.



________________________________
 From: J Arcane <jarcane at gmail.com>
To: Steve Graham <jsgrahamus at yahoo.com> 
Cc: Racket Users <users at racket-lang.org> 
Sent: Monday, September 22, 2014 9:35 PM
Subject: Re: [racket] Hosting the try-racket REPL.
 


It's a known issue. Alex Knauth has submitted a rough solution to my fork of it; I'll get it merged and pushed to the server shortly.


On Tue, Sep 23, 2014 at 12:11 AM, Steve Graham <jsgrahamus at yahoo.com> wrote:


>
>
>I get errors when trying to type in a multi-line function.  Am I doing something wrong?
>
>
>
>
>________________________________
> From: Jay McCarthy <jay.mccarthy at gmail.com>
>To: J Arcane <jarcane at gmail.com> 
>Cc: Racket Users <users at racket-lang.org>; Robby Findler <robby at eecs.northwestern.edu>; Marc Burns <m4burns at csclub.uwaterloo.ca> 
>Sent: Monday, September 22, 2014 12:05 PM
>Subject: Re: [racket] Hosting the try-racket REPL.
> 
>
>J and I chatted a little and he got his DNS working and also I added
>
>http://try.racket-lang.org/
>
>as an alias for it.
>
>In general, the Racket team is willing to give out subdomains of
>r-l.org for projects that would like them, so they don't have to buy
>their own DNS addresses. Just email the list or one of us directly to
>ask.
>
>Jay
>
>On Mon, Sep 22, 2014 at 1:17 AM, J Arcane <jarcane at gmail.com> wrote:
>> Hmm.
>>
>> Alright, replacing the calls to slideshow with calls directly to pict, as
>> well as commenting out the racket/gui/base calls successfully gets the code
>> to run without X.
>>
>> However, something about how it converts the picts for web isn't working, so
>> instead of a neat little circle, for instance, I get "(pict '(prog
>> #<procedure:... etc." instead.
>>
>> I suspect this is probably something to do with a difference in behavior in
>> how images are converted by the server, which does a convertible? test and
>> then runs:
>>  (run-code ev `(bytes-append
 #"data:image/png;base64,"
>>                          (base64-encode (convert ,res 'png-bytes) #"")))
>>
>> At an idiot's guess, something is different about how pict is making the
>> images, so convertible? is #f and the code just prints the (pict '(prog...
>> stuff.
>>
>> I've set up a fork here: https://github.com/jarcane/try-racket
>> I'm still working on getting the DNS set properly though, so no
>> try-racket.org yet; the servlet is public at http://104.131.18.192:8000/ but
>> I haven't figured out how to get try-racket.org to go there yet.
>>
>>
>> <robby at eecs.northwestern.edu> wrote:
>>>
>>> Maybe you can use  slideshow/code-pict instead?
>>>
>>> Robby
>>>
>>> On Sun, Sep 21, 2014 at 6:00 PM, Marc Burns <m4burns at csclub.uwaterloo.ca>
>>> wrote:
>>> > The culprit is the slideshow/code module loaded into the sandbox
>>> > evaluator. Requiring slideshow/code attempts to establish an X server
>>> > connection.
>>> >
>>> > On Mon, Sep 22, 2014 at
 12:12:37AM +0300, J Arcane wrote:
>>> >> Greetings,
>>> >>
>>> >> For some time now, it's bothered me a bit that Racket doesn't have an
>>> >> online REPL currently hosted anywhere. There's one written here:
>>> >> https://github.com/voila/try-racket
>>> >>
>>> >> But no one's hosted it anywhere. So I took it upon myself to fix that.
>>> >> I've
>>> >> purchased try-racket.org and a basic DigitalOcean droplet to host it
>>> >> on,
>>> >> but I've hit a snag: I can't get it to run.
>>> >>
>>> >> It runs more or less without trouble
 on my personal FreeBSD box, but on
>>> >> my
>>> >> Debian 7 droplet I waded through countless dependency issues until
>>> >> finally
>>> >> I reached what seems to be this same error:
>>> >>
>>> >> http://bugs.racket-lang.org/query/?cmd=view%20audit-trail?=12465
>
>>> >>
>>> >> Presuming that perhaps this was in someway trying to run Racket in an
>>> >> actual GUI window whether I wanted it to or not, I attempted to run it
>>> >> as
>>> >> --script, but this just quits silently instead, leaving no servlet
>>> >> behind
>>> >> either.
>>> >>
>>> >> Any ideas what I might be doing wrong? or have I hit a proper bug? (IME
>>> >> it's usually the former ... )
>>> >>
>>> >> John Berry
>>> >> http://jarcane.github.com
>>> >
>>> >> ____________________
>>> >>   Racket Users list:
>>> >>  http://lists.racket-lang.org/users
>>> >
>>> > ____________________
>>> >   Racket Users list:
>>> >  http://lists.racket-lang.org/users
>>
>>
>>
>> ____________________
>>   Racket Users list:
>>  http://lists.racket-lang.org/users
>>
>
>
>
>-- 
>Jay McCarthy
>http://jeapostrophe.github.io
>
>           "Wherefore, be not weary in well-doing,
>      for ye are laying the foundation of a great work.
>And out of small things proceedeth that which is great."
>             
             - D&C 64:33
>
>
>
>
>____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users
>
>
>
>____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140923/af8abab6/attachment-0001.html>

From gneuner2 at comcast.net  Tue Sep 23 21:19:41 2014
From: gneuner2 at comcast.net (George Neuner)
Date: Tue, 23 Sep 2014 21:19:41 -0400
Subject: [racket] web server: module servlets
In-Reply-To: <CAJYbDa=uR26UAdfPgg5RzLo3HJv_rWosXXiCx+Rp9vbxPEbe9Q@mail.gmail.com>
References: <5421DB30.3030803@comcast.net>
 <CAJYbDa=uR26UAdfPgg5RzLo3HJv_rWosXXiCx+Rp9vbxPEbe9Q@mail.gmail.com>
Message-ID: <54221C2D.9020109@comcast.net>

Hi Jay,

On 9/23/2014 6:04 PM, Jay McCarthy wrote:
> On Tue, Sep 23, 2014 at 4:42 PM, George Neuner <gneuner2 at comcast.net> wrote:
> >
> > 1) Is there any way to speed up initial [module] servlet loading?
>
> This is why I recommend that new users use serve/servlet, which is
> much faster because it does not use namespaces or live loading.

That's an option for a finished application ... if a web application 
ever really can
be said to be finished ... but it's tough when you're developing.

I guess the question is "where is all the time spent"?  Loading required 
modules
from the library I suppose?

Not really knowing much about Racket's internals he naively asks: Could 
a server
pre-load the commonly used webserver modules and make them available
to new module servlets, or does the custodian implementation make doing 
that
difficult/impossible?

> > 2) Debugging module servlets is a pain due to having to stop/restart
> > the server.  Is there some clever way to force the server to unload a
> > particular servlet but keep running?
>
> Doing this pretty much requires not following my advice in question 1,
> but the default server (used by the command line tool) allows you to
> go to "/conf/refresh-servlets" and unload/reload all the servlet code.

That might help.  Does unloading & reloading affect a stateless servlet 
that isn't
been changed?  More to the point, can you stop/restart the server and 
continue
a stateless servlet?

My application so far is based on stateful servlets and AJAX ... 
stateful mainly because
it's easier for me to understand.  Currently there is little use of 
continuations, but
some planned functionality will use them extensively and it certainly 
would help if
debugging didn't always mean starting over setting up conditions in the 
application.

George
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140923/187c2d08/attachment.html>

From jay.mccarthy at gmail.com  Tue Sep 23 22:03:38 2014
From: jay.mccarthy at gmail.com (Jay McCarthy)
Date: Tue, 23 Sep 2014 22:03:38 -0400
Subject: [racket] web server: module servlets
In-Reply-To: <54221C2D.9020109@comcast.net>
References: <5421DB30.3030803@comcast.net>
 <CAJYbDa=uR26UAdfPgg5RzLo3HJv_rWosXXiCx+Rp9vbxPEbe9Q@mail.gmail.com>
 <54221C2D.9020109@comcast.net>
Message-ID: <CAJYbDak29o47-yU5ONQy3DH+V+2DVCzT8UWqtaZNMsHAdb=JdA@mail.gmail.com>

On Tue, Sep 23, 2014 at 9:19 PM, George Neuner <gneuner2 at comcast.net> wrote:
> Hi Jay,
>
> On 9/23/2014 6:04 PM, Jay McCarthy wrote:
>
> On Tue, Sep 23, 2014 at 4:42 PM, George Neuner <gneuner2 at comcast.net> wrote:
>>
>> 1) Is there any way to speed up initial [module] servlet loading?
>
> This is why I recommend that new users use serve/servlet, which is
> much faster because it does not use namespaces or live loading.
>
>
> That's an option for a finished application ... if a web application ever
> really can
> be said to be finished ... but it's tough when you're developing.

Hi George,

The command-line tool is basically deprecated and only provided for
backwards compatibility. There is a huge amount that it can't do at
all and it hasn't been the primary way that we recommend using the Web
server for a very long time.

> I guess the question is "where is all the time spent"?  Loading required
> modules
> from the library I suppose?

Yes.

> Not really knowing much about Racket's internals he naively asks: Could a
> server
> pre-load the commonly used webserver modules and make them available
> to new module servlets, or does the custodian implementation make doing that
> difficult/impossible?

This is the purpose of the make-servlet-namespace argument of
configuration-table->web-config@ but there is no option in the
configuration file for that argument.

http://docs.racket-lang.org/web-server-internal/Web_Servers.html?q=servlet-namespace#%28def._web-config._%28%28lib._web-server%2Fweb-config-unit..rkt%29._configuration-table-~3eweb-config~40%29%29

>> 2) Debugging module servlets is a pain due to having to stop/restart
>> the server.  Is there some clever way to force the server to unload a
>> particular servlet but keep running?
>
> Doing this pretty much requires not following my advice in question 1,
> but the default server (used by the command line tool) allows you to
> go to "/conf/refresh-servlets" and unload/reload all the servlet code.
>
>
> That might help.  Does unloading & reloading affect a stateless servlet that
> isn't
> been changed?  More to the point, can you stop/restart the server and
> continue
> a stateless servlet?

Yes, that's the goal of stateless servlets :)

> My application so far is based on stateful servlets and AJAX ... stateful
> mainly because
> it's easier for me to understand.  Currently there is little use of
> continuations, but
> some planned functionality will use them extensively and it certainly would
> help if
> debugging didn't always mean starting over setting up conditions in the
> application.

This comment/question is related to questions 4 and 5 from the FAQ:

http://docs.racket-lang.org/web-server/faq.html?q=servlet-namespace#%28part._update-servlets%29


-- 
Jay McCarthy
http://jeapostrophe.github.io

           "Wherefore, be not weary in well-doing,
      for ye are laying the foundation of a great work.
And out of small things proceedeth that which is great."
                          - D&C 64:33

From gustavo at oma.org.ar  Tue Sep 23 23:22:28 2014
From: gustavo at oma.org.ar (Gustavo Massaccesi)
Date: Wed, 24 Sep 2014 00:22:28 -0300
Subject: [racket] Exception decorator. is it possible?
In-Reply-To: <1410973155.471904896@f186.i.mail.ru>
References: <1410973155.471904896@f186.i.mail.ru>
Message-ID: <CAPaha9P9jQcMMSUuKMF9UWqkTY7QCUZjfv-GRXuF58gEYQBDiw@mail.gmail.com>

This code works, but it needs a lot more of error checking.

It uses that all the built-in exceptions are transparent and that all
of them inherit from exn, so the first field is the message.

;---
#lang racket

(define (copy-struct/new-first str first)
  (define-values (str-struct-info _) (struct-info str))
  (define str-maker (struct-type-make-constructor str-struct-info))
  (apply str-maker first (cdr (cdr (vector->list (struct->vector str))))))

(define (parse-item item)
  (with-handlers ([exn? (lambda (e) (raise (copy-struct/new-first e
(format "At item ~a\n~a" item (exn-message e)))))])
    (cons 1 2 3) ;==> cons: arity mismatch;
    #;(/ 1 0) ;==> /: division by zero
    ))

(parse-item "Hello")
;---

Gustavo




On Wed, Sep 17, 2014 at 1:59 PM, Roman Klochkov <kalimehtar at mail.ru> wrote:
> I'm writeing a parser.
> So I want to see in parser exception what happend and when (not only where).
>
> (define (parse-item item)
>    (with-handlers ([exn? (lambda (e) (raise (struct-copy exn e [message
> (format "At item ~a\n~a" item (exn-message e))])))])
>          ....))
>
> But this way I lose all exception types. Is there a way to change exception
> message, saving its type, except long copy/paste like
> (with-handlers ([exn:fail:contract:arity? (raise (struct-copy
> exn:fail:contract:arity ...))]
>                          [exn:fil:contract? (raise (struct-copy
> exn:fail:contract ...))]
>                          ...)
>    ...)
>
> ?
>
> --
> Roman Klochkov
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>

From dtp at mindstory.com  Wed Sep 24 00:29:09 2014
From: dtp at mindstory.com (David T. Pierson)
Date: Wed, 24 Sep 2014 00:29:09 -0400
Subject: [racket] ranking items by their appearance in sets
Message-ID: <1411532547.17822%dtp@mindstory.com>

Hi all,

Given a list of sets of symbols, I want to rank the symbols based on how
many sets they appear in.  I prefer to be completely functional.

Below are 2 different implementations I came up with, with tests.  I'm
looking for criticism in any aspect, but most importantly I'm wondering
whether there is a simpler implementation.

Thanks.

David

;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
#lang racket

(require rackunit)

(define (rank-set-elements/1 list-of-sets)
  (define list-of-lists (map set->list list-of-sets))
  (define flat-list (flatten list-of-lists))
  (define sorted-flat-list (sort flat-list symbol<?))
  (define list-of-item-frequency-pairs
          (foldl (lambda (v accum)
                   (match accum
                     [(list-rest (cons sym cnt) rst)
                      (if (eq? sym v)
                          (cons (cons sym (add1 cnt)) rst)
                          (cons (cons v 1) accum))]
                     [else
                      (list (cons v 1))]))
                 '()
                 sorted-flat-list))
  (sort list-of-item-frequency-pairs > #:key cdr))

(define (rank-set-elements/2 list-of-sets)
  (define hash-of-item-frequencies
          (for*/fold ([frequencies (hash)])
                     ([a-set list-of-sets]
                      [sym a-set])
            (hash-update frequencies sym add1 0)))
  (define list-of-item-frequency-pairs (hash->list hash-of-item-frequencies))
  (sort list-of-item-frequency-pairs > #:key cdr))

(define input
        (shuffle
         (map list->set
              '((a)
                (a b)
                (a b c)
                (a b c d)
                (a b c d e)))))

(define expected
        '((a . 5)
          (b . 4)
          (c . 3)
          (d . 2)
          (e . 1)))

(check-equal? (rank-set-elements/1 '()) '())
(check-equal? (rank-set-elements/1 input)
              expected)

(check-equal? (rank-set-elements/2 '()) '())
(check-equal? (rank-set-elements/2 input)
              expected)

From kalimehtar at mail.ru  Wed Sep 24 07:50:45 2014
From: kalimehtar at mail.ru (=?UTF-8?B?Um9tYW4gS2xvY2hrb3Y=?=)
Date: Wed, 24 Sep 2014 15:50:45 +0400
Subject: [racket] =?utf-8?q?Exception_decorator=2E_is_it_possible=3F?=
In-Reply-To: <CAPaha9P9jQcMMSUuKMF9UWqkTY7QCUZjfv-GRXuF58gEYQBDiw@mail.gmail.com>
References: <1410973155.471904896@f186.i.mail.ru>
 <CAPaha9P9jQcMMSUuKMF9UWqkTY7QCUZjfv-GRXuF58gEYQBDiw@mail.gmail.com>
Message-ID: <1411559445.443202175@f436.i.mail.ru>

 Thank you very much!
I didn't find struct-info function. This is enough for my needs.


Wed, 24 Sep 2014 00:22:28 -0300 ?? Gustavo Massaccesi <gustavo at oma.org.ar>:
> This code works, but it needs a lot more of error checking.
> 
> It uses that all the built-in exceptions are transparent and that all
> of them inherit from exn, so the first field is the message.
> 
> ;---
> #lang racket
> 
> (define (copy-struct/new-first str first)
>   (define-values (str-struct-info _) (struct-info str))
>   (define str-maker (struct-type-make-constructor str-struct-info))
>   (apply str-maker first (cdr (cdr (vector->list (struct->vector str))))))
> 
> (define (parse-item item)
>   (with-handlers ([exn? (lambda (e) (raise (copy-struct/new-first e
> (format "At item ~a\n~a" item (exn-message e)))))])
>     (cons 1 2 3) ;==> cons: arity mismatch;
>     #;(/ 1 0) ;==> /: division by zero
>     ))
> 
> (parse-item "Hello")
> ;---
> 
> Gustavo
> 
> 
> 
> 
> On Wed, Sep 17, 2014 at 1:59 PM, Roman Klochkov <kalimehtar at mail.ru> wrote:
> > I'm writeing a parser.
> > So I want to see in parser exception what happend and when (not only where).
> >
> > (define (parse-item item)
> >    (with-handlers ([exn? (lambda (e) (raise (struct-copy exn e [message
> > (format "At item ~a\n~a" item (exn-message e))])))])
> >          ....))
> >
> > But this way I lose all exception types. Is there a way to change exception
> > message, saving its type, except long copy/paste like
> > (with-handlers ([exn:fail:contract:arity? (raise (struct-copy
> > exn:fail:contract:arity ...))]
> >                          [exn:fil:contract? (raise (struct-copy
> > exn:fail:contract ...))]
> >                          ...)
> >    ...)
> >
> > ?
> >
> > --
> > Roman Klochkov
> >
> > ____________________
> >   Racket Users list:
> >   http://lists.racket-lang.org/users
> >
> 

-- 
Roman Klochkov


From matthias at ccs.neu.edu  Wed Sep 24 10:44:45 2014
From: matthias at ccs.neu.edu (Matthias Felleisen)
Date: Wed, 24 Sep 2014 10:44:45 -0400
Subject: [racket] ranking items by their appearance in sets
In-Reply-To: <1411532547.17822%dtp@mindstory.com>
References: <1411532547.17822%dtp@mindstory.com>
Message-ID: <7081BD31-40BE-4648-8890-E4006F8B157D@ccs.neu.edu>


On Sep 24, 2014, at 12:29 AM, David T. Pierson <dtp at mindstory.com> wrote:

> I'm looking for criticism in any aspect,

Style: Here is my rewrite with style suggestions. I liked to have a single point that tests all versions of a function when I conduct such design experiments. I also want to have all my test code in the test submodule so that it doesn't get deployed. That's why I am using a syntactic abstraction that unfolds into a submodule rather than a function that calls things. 

Performance: I have not checked their performance but I cannot imagine the first version to be performant. See my annotations. 

My choice: Your second version is what I would have written. 

#lang racket

(module+ test
  (require rackunit))

;; ---------------------------------------------------------------------------------------------------
;; [Listof [Setof Symbol]] -> [Listof [Cons Symbol Natural]]
;; rank the elements of the given sets based on how many sets they appear in

;; version 1
(define (rank-set-elements/1 list-of-sets)               ;; n is number of symbols in sets 
  (define list-of-lists    (map set->list list-of-sets)) ;; O(n)
  (define flat-list        (flatten list-of-lists))      ;; allocates a lot  
  ;; I would have fused the next two lines and possibly all four 
  (define sorted-flat-list (sort flat-list symbol<?))    ;; O(n^2)? 
  (define list-of-item-frequency-pairs                   ;; critically relies on (sorted v)
    (for/fold ((accum '())) ((v sorted-flat-list))
      (match accum
        ;; this matches until the next, distinct symbol shows up 
        [(list-rest (cons sym cnt) rst)
         (if (eq? sym v)
             (cons (cons sym (add1 cnt)) rst)
             (cons (cons v 1) accum))]
        [else (list (cons v 1))])))
  (sort list-of-item-frequency-pairs > #:key cdr))

;; version 2 [[ that's the version I would have written ]]
(define (rank-set-elements/2 list-of-sets)
  (define hash-of-item-frequencies
    (for*/fold ([frequencies (hash)]) ([a-set list-of-sets] [sym a-set])
      (hash-update frequencies sym add1 0)))
  ;; I would fuse the next two lines 
  (define list-of-item-frequency-pairs (hash->list hash-of-item-frequencies))
  (sort list-of-item-frequency-pairs > #:key cdr))

;; unit tests <
(define-syntax-rule 
  (test rank)
  ;; => 
  (module+ test 
    (define input
      (shuffle
       (map list->set
            '((a)
              (a b)
              (a b c)
              (a b c d)
              (a b c d e)))))
    
    (define expected
      '((a . 5)
        (b . 4)
        (c . 3)
        (d . 2)
        (e . 1)))
    
    (check-equal? (rank '()) '())
    (check-equal? (rank input) expected)))

(test rank-set-elements/1)
(test rank-set-elements/2)



From terefv at ltu.sld.cu  Wed Sep 24 22:35:49 2014
From: terefv at ltu.sld.cu (Alejandro Zamora)
Date: Wed, 24 Sep 2014 22:35:49 -0400
Subject: [racket] macro question
Message-ID: <20140924223549.7f15c912@alejandro-H61H2-CM>

Hi Racket people:
	I have a (maybe too simple) question:
	How I create one macro like this (but using define-macro &
	R5RS language only)?:
	
	(vars sym var-cant val-of-vars)

	Example:
	(vars id 5 null) -> (prog (define id1 null) 
				  (define id2 null)
				  (define id3 null)
				  (define id4 null)
				  (define id5 null)) 

--
Nunca digas nunca, di mejor: gracias, permiso, disculpe.

Este mensaje le ha llegado mediante el servicio de correo electronico que ofrece Infomed para respaldar el cumplimiento de las misiones del Sistema Nacional de Salud. La persona que envia este correo asume el compromiso de usar el servicio a tales fines y cumplir con las regulaciones establecidas

Infomed: http://www.sld.cu/


From dtp at mindstory.com  Wed Sep 24 23:07:13 2014
From: dtp at mindstory.com (David T. Pierson)
Date: Wed, 24 Sep 2014 23:07:13 -0400
Subject: [racket] ranking items by their appearance in sets
In-Reply-To: <7081BD31-40BE-4648-8890-E4006F8B157D@ccs.neu.edu>
References: <1411532547.17822%dtp@mindstory.com>
 <7081BD31-40BE-4648-8890-E4006F8B157D@ccs.neu.edu>
Message-ID: <1411613699.19584%dtp@mindstory.com>

On Wed, Sep 24, 2014 at 10:44:45AM -0400, Matthias Felleisen wrote:
> Style: Here is my rewrite with style suggestions. I liked to have a
> single point that tests all versions of a function when I conduct such
> design experiments. I also want to have all my test code in the test
> submodule so that it doesn't get deployed. That's why I am using a
> syntactic abstraction that unfolds into a submodule rather than a
> function that calls things. 

Thanks for the style suggestions.  I am reminded I need to reread the
style guide.

> Performance: I have not checked their performance but I cannot imagine
> the first version to be performant. See my annotations. 

Good point.  I had put performance out of my mind because my real data
is relatively small and I just wanted to play a bit with the methods,
but I definitely appreciate the notes.

> My choice: Your second version is what I would have written. 

Glad to know I was on the right track.

Thank you.

David

From ckkashyap at gmail.com  Thu Sep 25 02:17:57 2014
From: ckkashyap at gmail.com (C K Kashyap)
Date: Thu, 25 Sep 2014 11:47:57 +0530
Subject: [racket] Switching from editor to repl
Message-ID: <CAGdT1gqUp-0hdF15FYo2T_Qpm+qf9AkKMak_iLV9HTC4WkDrtQ@mail.gmail.com>

Hi,
I am new to racket (but am really excited by the idea - everything is a
program) - need a quick help with Dr Racket -

What's the keyboard shortcut to switch between editor and repl?

I got to know from a video on youtube that I could use C-e to toggle the
repl visibility but not able to find out how to get to the rpl using the
keyboard.
Regards,
Kashyap
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140925/39d4382b/attachment.html>

From jay.mccarthy at gmail.com  Thu Sep 25 02:33:55 2014
From: jay.mccarthy at gmail.com (Jay McCarthy)
Date: Thu, 25 Sep 2014 02:33:55 -0400
Subject: [racket] Switching from editor to repl
In-Reply-To: <CAGdT1gqUp-0hdF15FYo2T_Qpm+qf9AkKMak_iLV9HTC4WkDrtQ@mail.gmail.com>
References: <CAGdT1gqUp-0hdF15FYo2T_Qpm+qf9AkKMak_iLV9HTC4WkDrtQ@mail.gmail.com>
Message-ID: <CAJYbDamt+_RvG1utwzjJ1KW-m5ynRDxPr2xVivTOGQj8Y2R3DA@mail.gmail.com>

I use C-e and C-d to do this and I just hit it twice, once hides the
side and the second shows it and focuses it.

On Thu, Sep 25, 2014 at 2:17 AM, C K Kashyap <ckkashyap at gmail.com> wrote:
> Hi,
> I am new to racket (but am really excited by the idea - everything is a
> program) - need a quick help with Dr Racket -
>
> What's the keyboard shortcut to switch between editor and repl?
>
> I got to know from a video on youtube that I could use C-e to toggle the
> repl visibility but not able to find out how to get to the rpl using the
> keyboard.
> Regards,
> Kashyap
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>



-- 
Jay McCarthy
http://jeapostrophe.github.io

           "Wherefore, be not weary in well-doing,
      for ye are laying the foundation of a great work.
And out of small things proceedeth that which is great."
                          - D&C 64:33

From ckkashyap at gmail.com  Thu Sep 25 03:43:58 2014
From: ckkashyap at gmail.com (C K Kashyap)
Date: Thu, 25 Sep 2014 13:13:58 +0530
Subject: [racket] Switching from editor to repl
In-Reply-To: <CAJYbDamt+_RvG1utwzjJ1KW-m5ynRDxPr2xVivTOGQj8Y2R3DA@mail.gmail.com>
References: <CAGdT1gqUp-0hdF15FYo2T_Qpm+qf9AkKMak_iLV9HTC4WkDrtQ@mail.gmail.com>
 <CAJYbDamt+_RvG1utwzjJ1KW-m5ynRDxPr2xVivTOGQj8Y2R3DA@mail.gmail.com>
Message-ID: <CAGdT1gqYYRz1Kk11U1==oUgP93ELFR0PFPsdG5v8C_tYjwqH=g@mail.gmail.com>

C-d works for me ... thanks Jay.
I was wondering though if I could've found this out in the documentation
somewhere. I could not find it in the guide/google.

Regards,
Kashyap

On Thu, Sep 25, 2014 at 12:03 PM, Jay McCarthy <jay.mccarthy at gmail.com>
wrote:

> I use C-e and C-d to do this and I just hit it twice, once hides the
> side and the second shows it and focuses it.
>
> On Thu, Sep 25, 2014 at 2:17 AM, C K Kashyap <ckkashyap at gmail.com> wrote:
> > Hi,
> > I am new to racket (but am really excited by the idea - everything is a
> > program) - need a quick help with Dr Racket -
> >
> > What's the keyboard shortcut to switch between editor and repl?
> >
> > I got to know from a video on youtube that I could use C-e to toggle the
> > repl visibility but not able to find out how to get to the rpl using the
> > keyboard.
> > Regards,
> > Kashyap
> >
> > ____________________
> >   Racket Users list:
> >   http://lists.racket-lang.org/users
> >
>
>
>
> --
> Jay McCarthy
> http://jeapostrophe.github.io
>
>            "Wherefore, be not weary in well-doing,
>       for ye are laying the foundation of a great work.
> And out of small things proceedeth that which is great."
>                           - D&C 64:33
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140925/93aa6ef8/attachment.html>

From robby at eecs.northwestern.edu  Thu Sep 25 08:56:53 2014
From: robby at eecs.northwestern.edu (Robby Findler)
Date: Thu, 25 Sep 2014 07:56:53 -0500
Subject: [racket] Switching from editor to repl
In-Reply-To: <CAGdT1gqYYRz1Kk11U1==oUgP93ELFR0PFPsdG5v8C_tYjwqH=g@mail.gmail.com>
References: <CAGdT1gqUp-0hdF15FYo2T_Qpm+qf9AkKMak_iLV9HTC4WkDrtQ@mail.gmail.com>
 <CAJYbDamt+_RvG1utwzjJ1KW-m5ynRDxPr2xVivTOGQj8Y2R3DA@mail.gmail.com>
 <CAGdT1gqYYRz1Kk11U1==oUgP93ELFR0PFPsdG5v8C_tYjwqH=g@mail.gmail.com>
Message-ID: <CAL3TdONC6=U8EibFfxLYDNVVW0+SEo9+Do5ZuMSu2bLQRm8y8A@mail.gmail.com>

The guide focuses on the language, not the DrRacket IDE. This is a
link to the IDE docs that has a list of keyboard shortcuts.

You might also try looking at the menu items; things are a little more
discoverable that way too.

http://docs.racket-lang.org/drracket/Keyboard_Shortcuts.html

hth,
Robby

On Thu, Sep 25, 2014 at 2:43 AM, C K Kashyap <ckkashyap at gmail.com> wrote:
> C-d works for me ... thanks Jay.
> I was wondering though if I could've found this out in the documentation
> somewhere. I could not find it in the guide/google.
>
> Regards,
> Kashyap
>
> On Thu, Sep 25, 2014 at 12:03 PM, Jay McCarthy <jay.mccarthy at gmail.com>
> wrote:
>>
>> I use C-e and C-d to do this and I just hit it twice, once hides the
>> side and the second shows it and focuses it.
>>
>> On Thu, Sep 25, 2014 at 2:17 AM, C K Kashyap <ckkashyap at gmail.com> wrote:
>> > Hi,
>> > I am new to racket (but am really excited by the idea - everything is a
>> > program) - need a quick help with Dr Racket -
>> >
>> > What's the keyboard shortcut to switch between editor and repl?
>> >
>> > I got to know from a video on youtube that I could use C-e to toggle the
>> > repl visibility but not able to find out how to get to the rpl using the
>> > keyboard.
>> > Regards,
>> > Kashyap
>> >
>> > ____________________
>> >   Racket Users list:
>> >   http://lists.racket-lang.org/users
>> >
>>
>>
>>
>> --
>> Jay McCarthy
>> http://jeapostrophe.github.io
>>
>>            "Wherefore, be not weary in well-doing,
>>       for ye are laying the foundation of a great work.
>> And out of small things proceedeth that which is great."
>>                           - D&C 64:33
>
>
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>

From jarcane at gmail.com  Thu Sep 25 09:52:47 2014
From: jarcane at gmail.com (J Arcane)
Date: Thu, 25 Sep 2014 16:52:47 +0300
Subject: [racket]  Unquoting a variable without unquoting its contents?
Message-ID: <CA+HVF0Ddf_WybAgJAAnJBvYeVoYMXhcN=4Kp4YF1_v-a9TSwGg@mail.gmail.com>

http://try-racket.org continues to splutter to life, but there's still
something of a glaring error in that it doesn't really like list or quote
at all.

AlexKnauth has narrowed the problem down here:
https://github.com/jarcane/try-racket/issues/4#issuecomment-56751603

It seems that unquoting a variable does funny things to the variable's
contents when they contain a bare quoted list or symbol, so we get weird
results and errors. Quotes get lost, and then the eval finds itself looking
at mangled forms.

The best I can think of to do is maybe make a simple macro rather than
using quasiquote? Any ideas?
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140925/e97f0d3f/attachment.html>

From jay.mccarthy at gmail.com  Thu Sep 25 10:17:49 2014
From: jay.mccarthy at gmail.com (Jay McCarthy)
Date: Thu, 25 Sep 2014 10:17:49 -0400
Subject: [racket] macro question
In-Reply-To: <20140924223549.7f15c912@alejandro-H61H2-CM>
References: <20140924223549.7f15c912@alejandro-H61H2-CM>
Message-ID: <CAJYbDakR7Sxa55rRiZoFinREGEdZyT7V4d+QRD0f_G1CA4j-GA@mail.gmail.com>

I won't answer the question about R5RS, but in Racket you can do this
fairly succinctly:

#lang racket/base
(require (for-syntax racket/base
                     racket/syntax))

(define-syntax (vars stx)
  (syntax-case stx ()
    [(_ id count val)
     (with-syntax
         ([(id_i ...)
           (for/list ([i (in-range 1 (add1 (syntax->datum #'count)))])
             (format-id #'id "~a~a" #'id i))])
     (syntax/loc stx
       (begin (define id_i val)
              ...)))]))

(vars id 5 null)

(vector id1 id2 id3 id4 id5)

2014-09-24 22:35 GMT-04:00 Alejandro Zamora <terefv at ltu.sld.cu>:
> Hi Racket people:
>         I have a (maybe too simple) question:
>         How I create one macro like this (but using define-macro &
>         R5RS language only)?:
>
>         (vars sym var-cant val-of-vars)
>
>         Example:
>         (vars id 5 null) -> (prog (define id1 null)
>                                   (define id2 null)
>                                   (define id3 null)
>                                   (define id4 null)
>                                   (define id5 null))
>
> --
> Nunca digas nunca, di mejor: gracias, permiso, disculpe.
>
> Este mensaje le ha llegado mediante el servicio de correo electronico que ofrece Infomed para respaldar el cumplimiento de las misiones del Sistema Nacional de Salud. La persona que envia este correo asume el compromiso de usar el servicio a tales fines y cumplir con las regulaciones establecidas
>
> Infomed: http://www.sld.cu/
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users



-- 
Jay McCarthy
http://jeapostrophe.github.io

           "Wherefore, be not weary in well-doing,
      for ye are laying the foundation of a great work.
And out of small things proceedeth that which is great."
                          - D&C 64:33


From terefv at ltu.sld.cu  Thu Sep 25 10:49:15 2014
From: terefv at ltu.sld.cu (Alejandro Zamora)
Date: Thu, 25 Sep 2014 10:49:15 -0400
Subject: [racket] macro question
In-Reply-To: <CAJYbDakR7Sxa55rRiZoFinREGEdZyT7V4d+QRD0f_G1CA4j-GA@mail.gmail.com>
References: <20140924223549.7f15c912@alejandro-H61H2-CM>
 <CAJYbDakR7Sxa55rRiZoFinREGEdZyT7V4d+QRD0f_G1CA4j-GA@mail.gmail.com>
Message-ID: <20140925104915.59291dc0@alejandro-H61H2-CM>

Thanks Jay!!
nice solution!!
I'm asking for define-macro & R5RS because I'm trying to use one
R5RS Embedded Scheme Interpreter on Python that have only define-macro
construction for make macros and I want to use it.

greetings!
 
 El Thu, 25 Sep 2014 10:17:49 -0400
Jay McCarthy <jay.mccarthy at gmail.com> escribi?:
> I won't answer the question about R5RS, but in Racket you can do this
> fairly succinctly:
> 
> #lang racket/base
> (require (for-syntax racket/base
>                      racket/syntax))
> 
> (define-syntax (vars stx)
>   (syntax-case stx ()
>     [(_ id count val)
>      (with-syntax
>          ([(id_i ...)
>            (for/list ([i (in-range 1 (add1 (syntax->datum #'count)))])
>              (format-id #'id "~a~a" #'id i))])
>      (syntax/loc stx
>        (begin (define id_i val)
>               ...)))]))
> 
> (vars id 5 null)
> 
> (vector id1 id2 id3 id4 id5)
> 
> 2014-09-24 22:35 GMT-04:00 Alejandro Zamora <terefv at ltu.sld.cu>:
> > Hi Racket people:
> >         I have a (maybe too simple) question:
> >         How I create one macro like this (but using define-macro &
> >         R5RS language only)?:
> >
> >         (vars sym var-cant val-of-vars)
> >
> >         Example:
> >         (vars id 5 null) -> (prog (define id1 null)
> >                                   (define id2 null)
> >                                   (define id3 null)
> >                                   (define id4 null)
> >                                   (define id5 null))
> >
> > --
> > Nunca digas nunca, di mejor: gracias, permiso, disculpe.
> >
> > Este mensaje le ha llegado mediante el servicio de correo
> > electronico que ofrece Infomed para respaldar el cumplimiento de
> > las misiones del Sistema Nacional de Salud. La persona que envia
> > este correo asume el compromiso de usar el servicio a tales fines y
> > cumplir con las regulaciones establecidas
> >
> > Infomed: http://www.sld.cu/
> >
> > ____________________
> >   Racket Users list:
> >   http://lists.racket-lang.org/users
> 
> 
> 


--
Nunca digas nunca, di mejor: gracias, permiso, disculpe.

Este mensaje le ha llegado mediante el servicio de correo electronico que ofrece Infomed para respaldar el cumplimiento de las misiones del Sistema Nacional de Salud. La persona que envia este correo asume el compromiso de usar el servicio a tales fines y cumplir con las regulaciones establecidas

Infomed: http://www.sld.cu/



From dbastos at toledo.com  Thu Sep 25 12:48:05 2014
From: dbastos at toledo.com (Daniel Bastos)
Date: Thu, 25 Sep 2014 13:48:05 -0300
Subject: [racket] htdp: exercise 21.1.3
Message-ID: <CAB08-qr1MMDnJ6uSztfx+qnJMGh3QYztK5iC3_xWiMhKz-onFA@mail.gmail.com>

The solution is missing, but I thought of exposing this one here so that it
gets scrutinized before sending it to Robby Findler.

;; Exercise 21.1.3. Define natural-f, which is the abstraction of the
;; following two functions:
;;
;; ;; copy : N X  ->  (listof X)
;; ;; to create a list that contains
;; ;; obj n times
;; (define (copy n obj)
;;   (cond
;;     [(zero? n) empty]
;;     [else (cons obj
;;                 (copy (sub1 n) obj))]))
;;
;; ;; n-adder : N number  ->  number
;; ;; to add n to x using
;; ;; (+ 1 ...) only
;; (define (n-adder n x)
;;   (cond
;;     [(zero? n) x]
;;     [else (+ 1
;;              (n-adder (sub1 n) x))]))
;;
;; Don't forget to test natural-f. Also use natural-f to define
;; n-multiplier, which consumes n and x and produces n times x with
;; additions only. Use the examples to formulate a contract.
;;
;; Hint: The two function differ more than, say, the functions sum and
;; product in exercise 21.1.2. In particular, the base case in one
;; instance is a argument of the function, where in the other it is just
;; a constant value.
;;
;; (*) Solution
;;
;; After following the design recipe for abstraction, we get natural-f as
;; shown below. To write the contract, let's first only consider copy,
;; then we consider only n-adder and then we consider both.
;;
;; ;; natural-f-copy: Nat X (X (listof X) -> (listof X)) (listof X) ->
(listof X)
;; ;; natural-f-adder: Nat number (number number -> number) number -> number
;;
;; The contract for natural-f-adder is more specific than
;; natural-f-copy. The difference between these is that in the first, two
;; different types are involved --- X and (listof X). Therefore, (listof
;; X) works as another variable, so let's call it Y in the final version.

;; natural-f: Nat X (X Y -> Y) Y -> Y
(define (natural-f n x f init)
  (cond
    [(zero? n) init]
    [else (f x
              (natural-f (sub1 n) x f init))]))

;; (*) Tests

;; Let's use two different types in copy, say symbol and then number.

(check-expect (copy 3 'a) '(a a a))
(check-expect (natural-f 3 'a cons empty) (copy 3 'a))

(check-expect (copy 0 'a) empty)
(check-expect (natural-f 0 'a cons empty) (copy 0 'a))

(check-expect (copy 3 1) (list 1 1 1))
(check-expect (natural-f 3 1 cons empty) (copy 3 1))

;; For n-adder, let's make sure to test different init-values.

(check-expect (n-adder 3 3.14) 6.14)
(check-expect (natural-f 3 1 + 3.14) (n-adder 3 3.14))

(check-expect (n-adder 3 1) 4)
(check-expect (natural-f 3 1 + 1) (n-adder 3 1))

(check-expect (n-adder 3 2) 5)
(check-expect (natural-f 3 1 + 2) (n-adder 3 2))
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140925/ea097a92/attachment.html>

From gneuner2 at comcast.net  Thu Sep 25 12:51:56 2014
From: gneuner2 at comcast.net (George Neuner)
Date: Thu, 25 Sep 2014 12:51:56 -0400
Subject: [racket] web server: module servlets
In-Reply-To: <CAJYbDak29o47-yU5ONQy3DH+V+2DVCzT8UWqtaZNMsHAdb=JdA@mail.gmail.com>
References: <5421DB30.3030803@comcast.net>	<CAJYbDa=uR26UAdfPgg5RzLo3HJv_rWosXXiCx+Rp9vbxPEbe9Q@mail.gmail.com>	<54221C2D.9020109@comcast.net>
 <CAJYbDak29o47-yU5ONQy3DH+V+2DVCzT8UWqtaZNMsHAdb=JdA@mail.gmail.com>
Message-ID: <5424482C.1030901@comcast.net>

Hi Jay,

On 9/23/2014 10:03 PM, Jay McCarthy wrote:
> The command-line tool is basically deprecated and only provided for
> backwards compatibility. There is a huge amount that it can't do at
> all and it hasn't been the primary way that we recommend using the Web
> server for a very long time.

I'm not using the plt-web-server app - I created a minimal application 
that set up the environment: directories, ports, etc. and a start 
function that just returns a 404 if called.  My server sits behind 
Apache and only handles servlets, which I would like to be demand load 
modules so they can be added to and updated easily. That's why I am 
interested in being able to unload servlets on command, though not 
necessarily all of them at once (although that also is helpful).


> > Not really knowing much about Racket's internals he naively asks:
> >Coulda server  pre-load the commonly used webserver modules and make
> >themavailableto new module servlets, or does the custodian
> >implementation  make doing thatdifficult/impossible?
>
> This is the purpose of the make-servlet-namespace argument of
> configuration-table->web-config@ but there is no option in the
> configuration file for that argument.
>
> http://docs.racket-lang.org/web-server-internal/Web_Servers.html?q=servlet-namespace#%28def._web-config._%28%28lib._web-server%2Fweb-config-unit..rkt%29._configuration-table-~3eweb-config~40%29%29

The namespace facility seems designed more for user written modules than 
for library modules.

My analysis may be off-base as I have very little experience working 
directly with custodians, but it seems that when a dynamic servlet is 
first loaded, its custodian spends a lot of additional time (re)loading 
library modules that already exist in other custodians. [ Though I can't 
tell exactly what's happening, I can see a lot of disk activity when I 
think I'm loading a 5KB servlet. ]  Reasonably I would have expected 
that after loading/linking the first servlet, the libraries common to 
the servlets would be already in memory. But loading additional servlets 
is no quicker [ and causes a similar disk hit ] so clearly I don't 
understand what is happening internally with the custodians.

If a Racket library is deliberately put into the servlet-namespace, does 
that streamline linking?


> > My application so far is based on stateful servlets and AJAX ... stateful
> > mainly because it's easier for me to understand.  Currently there is
> >little use ofcontinuations, but some planned functionality will use them
> >  extensively and it certainly wouldhelp if debugging didn't always mean
> >starting over setting up conditions in theapplication.
>
> This comment/question is related to questions 4 and 5 from the FAQ:
>
> http://docs.racket-lang.org/web-server/faq.html?q=servlet-namespace#%28part._update-servlets%29

That link leads to a "troubleshooting" page 8-).  I didn't consider the 
issue to be a "problem" per se - I already knew that restarting a 
stateful servlet would lose saved state.  Most of my existing servlets 
are one-shots that don't save any state, but they are written using the 
stateful language.  Only a few use continuations and not extensively (so 
far).  I was just thinking ahead to stuff that will need to use 
continuations more extensively.

Thanks for putting up with my questions.
George

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140925/7c368b80/attachment.html>

From jay.mccarthy at gmail.com  Thu Sep 25 13:02:00 2014
From: jay.mccarthy at gmail.com (Jay McCarthy)
Date: Thu, 25 Sep 2014 13:02:00 -0400
Subject: [racket] ANN: DOS,
	Delimited-continuation-based Operating-system Simulator
Message-ID: <CAJYbDam0maAG7ZmCBjbha1uqhJyCpX9J+L3=kxqAL4GjdKUvkg@mail.gmail.com>

I've just released the game architecture library I talked about in
part 2 of my RacketCon talk.

The big picture of this library is to make World-style programs more
 compositional by (a) using continuations to hide the internal state
(including control state) of components and (b) using environments as
a standard monoid-based inter-component communication channel. A
monoid is used to ensure that the components can be evaluated in any
order. Despite assumptions some have about continuations and pure
functional programming, it is incredibly efficient and can run at
60FPS, as demonstrated in get-bonus.

You can get it with

raco pkg install dos

And you can try out the demo with

racket -l dos/examples/win

The demo source is a bare 39 lines:

https://github.com/jeapostrophe/dos/blob/master/dos/examples/win.rkt

and I provide the non-DOS version for comparison:

https://github.com/jeapostrophe/dos/blob/master/dos/examples/win-long.rkt

The core library has a mere 36 lines:

https://github.com/jeapostrophe/dos/blob/master/dos/main.rkt

-- 
Jay McCarthy
http://jeapostrophe.github.io

           "Wherefore, be not weary in well-doing,
      for ye are laying the foundation of a great work.
And out of small things proceedeth that which is great."
                          - D&C 64:33

From jay.mccarthy at gmail.com  Thu Sep 25 13:04:44 2014
From: jay.mccarthy at gmail.com (Jay McCarthy)
Date: Thu, 25 Sep 2014 13:04:44 -0400
Subject: [racket] web server: module servlets
In-Reply-To: <5424482C.1030901@comcast.net>
References: <5421DB30.3030803@comcast.net>
 <CAJYbDa=uR26UAdfPgg5RzLo3HJv_rWosXXiCx+Rp9vbxPEbe9Q@mail.gmail.com>
 <54221C2D.9020109@comcast.net>
 <CAJYbDak29o47-yU5ONQy3DH+V+2DVCzT8UWqtaZNMsHAdb=JdA@mail.gmail.com>
 <5424482C.1030901@comcast.net>
Message-ID: <CAJYbDa=-DuRMVy7JnCt32Y8waG2+qkHZQw5cVDL4L_d6zth45w@mail.gmail.com>

On Thu, Sep 25, 2014 at 12:51 PM, George Neuner <gneuner2 at comcast.net> wrote:
> Hi Jay,
>
> On 9/23/2014 10:03 PM, Jay McCarthy wrote:
>
> The command-line tool is basically deprecated and only provided for
> backwards compatibility. There is a huge amount that it can't do at
> all and it hasn't been the primary way that we recommend using the Web
> server for a very long time.
>
>
> I'm not using the plt-web-server app - I created a minimal application that
> set up the environment: directories, ports, etc. and a start function that
> just returns a 404 if called.  My server sits behind Apache and only handles
> servlets, which I would like to be demand load modules so they can be added
> to and updated easily.  That's why I am interested in being able to unload
> servlets on command, though not necessarily all of them at once (although
> that also is helpful).
>
>
>> Not really knowing much about Racket's internals he naively asks:
>> Could a server  pre-load the commonly used webserver modules and make
>> them available to new module servlets, or does the custodian
>> implementation make doing that difficult/impossible?
>
> This is the purpose of the make-servlet-namespace argument of
> configuration-table->web-config@ but there is no option in the
> configuration file for that argument.
>
> http://docs.racket-lang.org/web-server-internal/Web_Servers.html?q=servlet-namespace#%28def._web-config._%28%28lib._web-server%2Fweb-config-unit..rkt%29._configuration-table-~3eweb-config~40%29%29
>
>
> The namespace facility seems designed more for user written modules than for
> library modules.
>
> My analysis may be off-base as I have very little experience working
> directly with custodians, but it seems that when a dynamic servlet is first
> loaded, its custodian spends a lot of additional time (re)loading library
> modules that already exist in other custodians.  [ Though I can't tell
> exactly what's happening, I can see a lot of disk activity when I think I'm
> loading a 5KB servlet. ]  Reasonably I would have expected that after
> loading/linking the first servlet, the libraries common to the servlets
> would be already in memory.  But loading additional servlets is no quicker [
> and causes a similar disk hit ] so clearly I don't understand what is
> happening internally with the custodians.
>
> If a Racket library is deliberately put into the servlet-namespace, does
> that streamline linking?
>

Your assumption about the purpose of this is not correct. Anything in
the servlet-namespace will be shared between all servlets, and thus
not loaded per-servlet.

>> My application so far is based on stateful servlets and AJAX ... stateful
>> mainly because it's easier for me to understand.  Currently there is
>> little use of continuations, but some planned functionality will use them
>> extensively and it certainly would help if debugging didn't always mean
>> starting over setting up conditions in the application.
>
> This comment/question is related to questions 4 and 5 from the FAQ:
>
> http://docs.racket-lang.org/web-server/faq.html?q=servlet-namespace#%28part._update-servlets%29
>
>
> That link leads to a "troubleshooting" page 8-).  I didn't consider the
> issue to be a "problem" per se - I already knew that restarting a stateful
> servlet would lose saved state.  Most of my existing servlets are one-shots
> that don't save any state, but they are written using the stateful language.
> Only a few use continuations and not extensively (so far).  I was just
> thinking ahead to stuff that will need to use continuations more
> extensively.
>
> Thanks for putting up with my questions.
> George
>



-- 
Jay McCarthy
http://jeapostrophe.github.io

           "Wherefore, be not weary in well-doing,
      for ye are laying the foundation of a great work.
And out of small things proceedeth that which is great."
                          - D&C 64:33

From fweytjens.android at gmail.com  Thu Sep 25 16:19:51 2014
From: fweytjens.android at gmail.com (Frank Weytjens)
Date: Thu, 25 Sep 2014 22:19:51 +0200
Subject: [racket] My DrRacket on Linux doesn't evaluate a thing
Message-ID: <CAKP-afq3bwSdJ8dKmasOSE7Wst4DD8LFaQLDQQDPmUxZqjHa_g@mail.gmail.com>

Hi,

My DrRacket doesn't evaluate any expression.
Here is a link <http://bugs.racket-lang.org/query/?cmd=view&pr=14746> to my
bug report
Any suggestions?

Thanks,

Frank
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140925/90e65835/attachment.html>

From gneuner2 at comcast.net  Thu Sep 25 16:34:09 2014
From: gneuner2 at comcast.net (George Neuner)
Date: Thu, 25 Sep 2014 16:34:09 -0400
Subject: [racket] web server: module servlets
In-Reply-To: <CAJYbDa=-DuRMVy7JnCt32Y8waG2+qkHZQw5cVDL4L_d6zth45w@mail.gmail.com>
References: <5421DB30.3030803@comcast.net>	<CAJYbDa=uR26UAdfPgg5RzLo3HJv_rWosXXiCx+Rp9vbxPEbe9Q@mail.gmail.com>	<54221C2D.9020109@comcast.net>	<CAJYbDak29o47-yU5ONQy3DH+V+2DVCzT8UWqtaZNMsHAdb=JdA@mail.gmail.com>	<5424482C.1030901@comcast.net>
 <CAJYbDa=-DuRMVy7JnCt32Y8waG2+qkHZQw5cVDL4L_d6zth45w@mail.gmail.com>
Message-ID: <54247C41.1030408@comcast.net>

Hi Jay,

On 9/25/2014 1:04 PM, Jay McCarthy wrote:
> > If a Racket library is deliberately put into the servlet-namespace, does
> > that streamline linking?
>
> Your assumption about the purpose of this is not correct. Anything in
> the servlet-namespace will be shared between all servlets, and thus
> not loaded per-servlet.

I think you misunderstood my question, but you may have answered it anyway.

What I really was asking was whether each custodian was having to 
individually load common Racket libraries [ web-server/*, net/*, etc. ] 
for a new servlet rather than all servlet custodians sharing libraries 
that are already loaded.  Or, if not actually "loading" the libraries, 
then having to go to disk to check dependencies.

So if I create a shared servlet namespace and put, e.g., 
"web-server/servlet" into it, would that in any way speed up starting a 
new dynamically loaded servlet?

George

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140925/95bfa8ff/attachment.html>

From gneuner2 at comcast.net  Thu Sep 25 17:07:49 2014
From: gneuner2 at comcast.net (George Neuner)
Date: Thu, 25 Sep 2014 17:07:49 -0400
Subject: [racket] Off Topic:  Anybody reading this via news.gmane.org?
In-Reply-To: <CAJYbDam0maAG7ZmCBjbha1uqhJyCpX9J+L3=kxqAL4GjdKUvkg@mail.gmail.com>
References: <CAJYbDam0maAG7ZmCBjbha1uqhJyCpX9J+L3=kxqAL4GjdKUvkg@mail.gmail.com>
Message-ID: <54248425.90509@comcast.net>


Just wondering if anyone else reads this list via gmane and also is 
having problems?

I follow several lists via gmane and a few days ago all my posts started 
being rejected - no authorization email, just  an immediate "you are not 
allowed to post" error.   Then 2 days ago, I also lost read access to 
all the lists - attempts get "480 Read access denied".  I can ping the 
NNTP server but I can't connect to it.

I have sent a couple of emails to gmane's adminstrator but so far I 
haven't gotten any response.  Gmane is an open service that doesn't 
require accounts, so I am uncertain how I suddenly could be denied 
reading privilege ... even to post it only requires that you be 
subscribed to the particular list.  I can't think of any reason my 
address would be black-listed ... unless, of course, all of comcast.net 
has been black-listed.

Thanks,
George

From neil at neilvandyke.org  Thu Sep 25 17:26:12 2014
From: neil at neilvandyke.org (Neil Van Dyke)
Date: Thu, 25 Sep 2014 17:26:12 -0400
Subject: [racket] Off Topic:  Anybody reading this via news.gmane.org?
In-Reply-To: <54248425.90509@comcast.net>
References: <CAJYbDam0maAG7ZmCBjbha1uqhJyCpX9J+L3=kxqAL4GjdKUvkg@mail.gmail.com>
 <54248425.90509@comcast.net>
Message-ID: <54248874.80201@neilvandyke.org>

You can check your IP addr against the list at 
"http://gmane.org/denied.php".

Neil V.


From mflatt at cs.utah.edu  Thu Sep 25 17:57:20 2014
From: mflatt at cs.utah.edu (Matthew Flatt)
Date: Thu, 25 Sep 2014 15:57:20 -0600
Subject: [racket] My DrRacket on Linux doesn't evaluate a thing
In-Reply-To: <CAKP-afq3bwSdJ8dKmasOSE7Wst4DD8LFaQLDQQDPmUxZqjHa_g@mail.gmail.com>
References: <CAKP-afq3bwSdJ8dKmasOSE7Wst4DD8LFaQLDQQDPmUxZqjHa_g@mail.gmail.com>
Message-ID: <20140925215722.0D540650190@mail-svr1.cs.utah.edu>

After you type "(+ 1 1)", are you hitting the Enter key on the main
part of the keyboard, or the one that's part of the numpad?

If you're pressing the latter (the numpad Enter), then DrRacket doesn't
evaluate the expression, and it instead lets you continue typing on the
next line. If that's it, use the Enter key that's not in the numpad;
the two keys are intentionally treated differently.

At Thu, 25 Sep 2014 22:19:51 +0200, Frank Weytjens wrote:
> Hi,
> 
> My DrRacket doesn't evaluate any expression.
> Here is a link <http://bugs.racket-lang.org/query/?cmd=view&pr=14746> to my
> bug report
> Any suggestions?
> 
> Thanks,
> 
> Frank
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users

From alexander at knauth.org  Thu Sep 25 18:15:50 2014
From: alexander at knauth.org (Alexander D. Knauth)
Date: Thu, 25 Sep 2014 18:15:50 -0400
Subject: [racket] typed racket, filters, and polymorphism
Message-ID: <D276D7BD-D6AF-4F26-A79B-C1016641EF06@knauth.org>

Do any of you have any advice for getting a function like this to type-check?
#lang typed/racket

(: check-int : (All (a) (case-> [a -> a]
                                [Any -> Integer])))
(define (check-int int)
  (unless (exact-integer? int)
    (error 'check-int "expected Integer, given ~v" int))
  int)

;. Type Checker: type mismatch
;  expected: a
;  given: Integer in: int


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140925/6490e5ad/attachment.html>

From mflatt at cs.utah.edu  Thu Sep 25 18:16:15 2014
From: mflatt at cs.utah.edu (Matthew Flatt)
Date: Thu, 25 Sep 2014 16:16:15 -0600
Subject: [racket] My DrRacket on Linux doesn't evaluate a thing
In-Reply-To: <CAKP-afoydYZC2P58SjimCv0Tub7mN2dCZo_BzimfBwr9ptG08g@mail.gmail.com>
References: <CAKP-afq3bwSdJ8dKmasOSE7Wst4DD8LFaQLDQQDPmUxZqjHa_g@mail.gmail.com>
 <20140925215722.0D540650190@mail-svr1.cs.utah.edu>
 <CAKP-afoydYZC2P58SjimCv0Tub7mN2dCZo_BzimfBwr9ptG08g@mail.gmail.com>
Message-ID: <20140925221616.B5B3665018E@mail-svr1.cs.utah.edu>

Ok, so much for my guess... unless your keyboard is mapped in an
unusual way. To rule that out completely, you could download

https://raw.githubusercontent.com/plt/racket/stable/pkgs/gui-pkgs/gui-test/tests/gracket/showkey.rkt

run it (assuming that the "Run" button works!), and make sure that the
Enter key is reported as "\r".


I expect that something else is happening, though, and I don't have
another guess so far.

At Fri, 26 Sep 2014 00:06:27 +0200, Frank Weytjens wrote:
> Hi Matthew,
> 
> I use the Enter from the main part of the keyboard.
> 
> 
> 
> 
> 
> 
> 2014-09-25 23:57 GMT+02:00 Matthew Flatt <mflatt at cs.utah.edu>:
> 
> > After you type "(+ 1 1)", are you hitting the Enter key on the main
> > part of the keyboard, or the one that's part of the numpad?
> >
> > If you're pressing the latter (the numpad Enter), then DrRacket doesn't
> > evaluate the expression, and it instead lets you continue typing on the
> > next line. If that's it, use the Enter key that's not in the numpad;
> > the two keys are intentionally treated differently.
> >
> > At Thu, 25 Sep 2014 22:19:51 +0200, Frank Weytjens wrote:
> > > Hi,
> > >
> > > My DrRacket doesn't evaluate any expression.
> > > Here is a link <http://bugs.racket-lang.org/query/?cmd=view&pr=14746>
> > to my
> > > bug report
> > > Any suggestions?
> > >
> > > Thanks,
> > >
> > > Frank
> > > ____________________
> > >   Racket Users list:
> > >   http://lists.racket-lang.org/users
> >
> 
> 
> 
> -- 
> Hazelarenlaan 34 bus 6
> 3500 Hasselt
> +32484538321

From samth at cs.indiana.edu  Thu Sep 25 18:19:37 2014
From: samth at cs.indiana.edu (Sam Tobin-Hochstadt)
Date: Thu, 25 Sep 2014 18:19:37 -0400
Subject: [racket] typed racket, filters, and polymorphism
In-Reply-To: <D276D7BD-D6AF-4F26-A79B-C1016641EF06@knauth.org>
References: <D276D7BD-D6AF-4F26-A79B-C1016641EF06@knauth.org>
Message-ID: <CAK=HD+ZpPuvYTMZJ1_LXrnPGjcPamZ73Z8aiPizCussR+UxYsw@mail.gmail.com>

No, I don't think you can do this. Can you say more about what you're
trying to accomplish?

Sam

On Thu, Sep 25, 2014 at 6:15 PM, Alexander D. Knauth
<alexander at knauth.org> wrote:
> Do any of you have any advice for getting a function like this to
> type-check?
> #lang typed/racket
>
> (: check-int : (All (a) (case-> [a -> a]
>                                 [Any -> Integer])))
> (define (check-int int)
>   (unless (exact-integer? int)
>     (error 'check-int "expected Integer, given ~v" int))
>   int)
>
> ;. Type Checker: type mismatch
> ;  expected: a
> ;  given: Integer in: int
>
>
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>

From jay.mccarthy at gmail.com  Thu Sep 25 18:26:58 2014
From: jay.mccarthy at gmail.com (Jay McCarthy)
Date: Thu, 25 Sep 2014 18:26:58 -0400
Subject: [racket] web server: module servlets
In-Reply-To: <54247C41.1030408@comcast.net>
References: <5421DB30.3030803@comcast.net>
 <CAJYbDa=uR26UAdfPgg5RzLo3HJv_rWosXXiCx+Rp9vbxPEbe9Q@mail.gmail.com>
 <54221C2D.9020109@comcast.net>
 <CAJYbDak29o47-yU5ONQy3DH+V+2DVCzT8UWqtaZNMsHAdb=JdA@mail.gmail.com>
 <5424482C.1030901@comcast.net>
 <CAJYbDa=-DuRMVy7JnCt32Y8waG2+qkHZQw5cVDL4L_d6zth45w@mail.gmail.com>
 <54247C41.1030408@comcast.net>
Message-ID: <CAJYbDamsu2F=97vnZJQdx2dRO-Ss46NU0icrf8XwNA_2mMg9gw@mail.gmail.com>

On Thu, Sep 25, 2014 at 4:34 PM, George Neuner <gneuner2 at comcast.net> wrote:
> Hi Jay,
>
> On 9/25/2014 1:04 PM, Jay McCarthy wrote:
>
>> If a Racket library is deliberately put into the servlet-namespace, does
>> that streamline linking?
>
> Your assumption about the purpose of this is not correct. Anything in
> the servlet-namespace will be shared between all servlets, and thus
> not loaded per-servlet.
>
>
> I think you misunderstood my question, but you may have answered it anyway.
>
> What I really was asking was whether each custodian was having to
> individually load common Racket libraries [ web-server/*, net/*, etc. ] for
> a new servlet rather than all servlet custodians sharing libraries that are
> already loaded.  Or, if not actually "loading" the libraries, then having to
> go to disk to check dependencies.
>
> So if I create a shared servlet namespace and put, e.g.,
> "web-server/servlet" into it, would that in any way speed up starting a new
> dynamically loaded servlet?

There is a very small set that is automatically shared:
racket/base
web-server/private/servlet
web-server/http
web-server/servlet/web
...

web-server/servlet is NOT shared, nor is any net library or common
racket library like racket/list. They are all totally unique per
servlet.

This, by the way, is part of why I don't recommend using dynamic
servlets at all and suggest using serve/servlet.

Jay



-- 
Jay McCarthy
http://jeapostrophe.github.io

           "Wherefore, be not weary in well-doing,
      for ye are laying the foundation of a great work.
And out of small things proceedeth that which is great."
                          - D&C 64:33

From gneuner2 at comcast.net  Thu Sep 25 20:13:54 2014
From: gneuner2 at comcast.net (George Neuner)
Date: Thu, 25 Sep 2014 20:13:54 -0400
Subject: [racket] [GENERAL] Off Topic: Anybody reading this via
	news.gmane.org?
In-Reply-To: <5424A05A.9020809@aklaver.com>
References: <CAJYbDam0maAG7ZmCBjbha1uqhJyCpX9J+L3=kxqAL4GjdKUvkg@mail.gmail.com>
 <54248425.90509@comcast.net> <5424A05A.9020809@aklaver.com>
Message-ID: <5424AFC2.6020302@comcast.net>


On 9/25/2014 5:26 PM, Neil Van Dyke wrote:
> You can check your IP addr against the list at 
> "http://gmane.org/denied.php".
>
> Neil V.

On 9/25/2014 7:08 PM, Adrian Klaver wrote:
>
> Take a look here:
>
> http://gmane.org/denied.php
>
> My guess is you are the fourth one from the bottom.
>
>
> Might want to take a look at this thread to see what your options are 
> and what the turn around time is on your request:
>
> http://thread.gmane.org/gmane.discuss/16309

Thanks to both of you - I missed seeing the entry about the denied list 
in the FAQ.

Adrian you were right - it seems that I have been blocked for some 
reason.  Based on the description of infractions I really don't think I 
did anything to warrant it ... but there it is.  I'll have to follow my 
lists by email until I can unblocked.

Thanks again,
George
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140925/0d2a1f73/attachment.html>

From alexander at knauth.org  Thu Sep 25 21:42:00 2014
From: alexander at knauth.org (Alexander D. Knauth)
Date: Thu, 25 Sep 2014 21:42:00 -0400
Subject: [racket] typed racket, filters, and polymorphism
In-Reply-To: <CAK=HD+ZpPuvYTMZJ1_LXrnPGjcPamZ73Z8aiPizCussR+UxYsw@mail.gmail.com>
References: <D276D7BD-D6AF-4F26-A79B-C1016641EF06@knauth.org>
 <CAK=HD+ZpPuvYTMZJ1_LXrnPGjcPamZ73Z8aiPizCussR+UxYsw@mail.gmail.com>
Message-ID: <F16C1892-F09C-4A9F-94E2-16D5DE72F57D@knauth.org>

What I?m trying to accomplish is something more like this:
#lang typed/racket

(require "dimensions.rkt")

(struct (d) unit ([name : Any] [scalar : Positive-Real] [dimension : d]) #:transparent
  #:guard (lambda (name scalar dimension _)
            (unless (dimension? dimension)
              (error 'unit "expected Dimension, given ~v" dimension))
            (values name scalar dimension)))

(define-type (Unitof d) (unit d))

(define-type Unit (Unitof Dimension))

(define Unit? (make-predicate Unit))

(define-type Unitish
  (U (Unitof Any)
     Dimension
     Positive-Real))

(: ->unit : (All (d) (case-> [(Unitof d) -> (Unitof d)]
                             [Unitish -> Unit])))
(define (->unit u)
  (cond [(unit? u)
         (unless (Unit? u) ; this should never happen anyway because of the guard
           (error '->unit "expected (Unitof Dimension), given ~v" u))
         u]
        [(dimension? u) (unit u 1 u)]
        [(positive-real? u) (unit u u dimensionless-dimension)]))


On Sep 25, 2014, at 6:19 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> wrote:

> No, I don't think you can do this. Can you say more about what you're
> trying to accomplish?
> 
> Sam
> 
> On Thu, Sep 25, 2014 at 6:15 PM, Alexander D. Knauth
> <alexander at knauth.org> wrote:
>> Do any of you have any advice for getting a function like this to
>> type-check?
>> #lang typed/racket
>> 
>> (: check-int : (All (a) (case-> [a -> a]
>>                                [Any -> Integer])))
>> (define (check-int int)
>>  (unless (exact-integer? int)
>>    (error 'check-int "expected Integer, given ~v" int))
>>  int)
>> 
>> ;. Type Checker: type mismatch
>> ;  expected: a
>> ;  given: Integer in: int
>> 
>> 
>> 
>> ____________________
>>  Racket Users list:
>>  http://lists.racket-lang.org/users
>> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140925/dc09b97f/attachment.html>

From gneuner2 at comcast.net  Fri Sep 26 00:16:56 2014
From: gneuner2 at comcast.net (George Neuner)
Date: Fri, 26 Sep 2014 00:16:56 -0400
Subject: [racket] web server: module servlets
In-Reply-To: <CAJYbDamsu2F=97vnZJQdx2dRO-Ss46NU0icrf8XwNA_2mMg9gw@mail.gmail.com>
References: <5421DB30.3030803@comcast.net>	<CAJYbDa=uR26UAdfPgg5RzLo3HJv_rWosXXiCx+Rp9vbxPEbe9Q@mail.gmail.com>	<54221C2D.9020109@comcast.net>	<CAJYbDak29o47-yU5ONQy3DH+V+2DVCzT8UWqtaZNMsHAdb=JdA@mail.gmail.com>	<5424482C.1030901@comcast.net>	<CAJYbDa=-DuRMVy7JnCt32Y8waG2+qkHZQw5cVDL4L_d6zth45w@mail.gmail.com>	<54247C41.1030408@comcast.net>
 <CAJYbDamsu2F=97vnZJQdx2dRO-Ss46NU0icrf8XwNA_2mMg9gw@mail.gmail.com>
Message-ID: <5424E8B8.8040501@comcast.net>

On 9/25/2014 6:26 PM, Jay McCarthy wrote:
> web-server/servlet is NOT shared, nor is any net library or common
> racket library like racket/list. They are all totally unique per
> servlet.
>
> This, by the way, is part of why I don't recommend using dynamic
> servlets at all and suggest using serve/servlet.

But ... IIUC ... a listener started by  serve/servlet  will still 
dynamically load from <servlet-root>, <server-root>/htdocs/*, etc. if 
the request doesn't match a hard coded dispatch URL - the regex of the 
initial servlet or an entry in a dispatcher table.  At least that's the 
behavior I see with my own application: my initial servlet returns 404 
when called - the call to  serve/servlet  just sets up the environment 
and and all the "real" servlets are demand loaded from disk when first 
touched.

Are you advocating *static* linking and essentially just dispatching to 
internal functions by URL?  ISTM that that defeats the purpose.

George
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140926/ffaf6bde/attachment-0001.html>

From konrad.hinsen at fastmail.net  Fri Sep 26 12:07:19 2014
From: konrad.hinsen at fastmail.net (Konrad Hinsen)
Date: Fri, 26 Sep 2014 18:07:19 +0200
Subject: [racket] Tricky case of occurrence typing in Typed Racket
Message-ID: <21541.36663.168649.25871@Konrad-Hinsens-MacBook-Pro-2.local>

Hi everyone,

I am trying to convince Typed Racket that after a run-time check on
some list, its elements are all of a given type. This is just like
occurrence typing based on a predicate, except that my predicate
is more complicated.

My first attempt was this:


   #lang typed/racket

   (struct: foo ())
   (struct: bar ())
   (define-type FooOrBar (U foo bar))

   (: f-bar (-> (Listof bar) Void))
   (define (f-bar xs)
     (void))

   (: f-mixed (-> (Listof FooOrBar) Void))
   (define (f-mixed xs)
     (void))

   (: f (-> (Listof FooOrBar) Void))
   (define (f xs)
     (if (for/and : Boolean ([x xs]) (bar? x))
         (f-bar xs) ; contains only bars
         (f-mixed xs) ; may contain foos as well
         ))

This yields a type error:

   ; /Users/hinsen/projects/racket/foobar.rkt:18:13: Type Checker: type mismatch
   ;   expected: (Listof bar)
   ;   given: (Listof FooOrBar)
   ;   in: xs

After reading section 5 of the Typed Racket Guide, in particular section
5.2 entitled "Filters and Predicates", I thought I could simply define my
own predicate with a type annotation:

   #lang typed/racket

   (struct: foo ())
   (struct: bar ())
   (define-type FooOrBar (U foo bar))

   (: f-bar (-> (Listof bar) Void))
   (define (f-bar xs)
     (void))

   (: f-mixed (-> (Listof FooOrBar) Void))
   (define (f-mixed xs)
     (void))

   (: pure-bars? (-> (Listof FooOrBar) Boolean : (Listof bar)))
   (define (pure-bars? xs)
     (for/and : Boolean ([x xs]) (bar? x)))

   (: f (-> (Listof FooOrBar) Void))
   (define (f xs)
     (if (pure-bars? xs)
         (f-bar xs) ; contains only bars
         (f-mixed xs) ; may contain foos as well
         ))

But this got me only into more trouble - now I don't even understand the
error message any more:

   ; /Users/hinsen/projects/racket/foobar.rkt:17:2: Type Checker: type mismatch;
   ;  mismatch in filter
   ;   expected: (((Listof bar) @ xs) | (! (Listof bar) @ xs))
   ;   given: (Top | Top)
   ;   in: (for/and : Boolean ((x xs)) (bar? x))

Is there a way to do this kind of test?

Konrad.

From amk.kent at gmail.com  Fri Sep 26 12:12:29 2014
From: amk.kent at gmail.com (Andrew Kent)
Date: Fri, 26 Sep 2014 12:12:29 -0400
Subject: [racket] Tricky case of occurrence typing in Typed Racket
In-Reply-To: <21541.36663.168649.25871@Konrad-Hinsens-MacBook-Pro-2.local>
References: <21541.36663.168649.25871@Konrad-Hinsens-MacBook-Pro-2.local>
Message-ID: <CAMf1y4JeW-jdYeuJP-0x=U=TJB3tahxqvAmuM3TSbeFvGSAYJg@mail.gmail.com>

Will andmap work for you?

   (struct: foo ())
   (struct: bar ())
   (define-type FooOrBar (U foo bar))

   (: f-bar (-> (Listof bar) Void))
   (define (f-bar xs)
     (void))

   (: f-mixed (-> (Listof FooOrBar) Void))
   (define (f-mixed xs)
     (void))

   (: f (-> (Listof FooOrBar) Void))
   (define (f xs)
     (if (andmap bar? xs)
         (f-bar xs) ; contains only bars
         (f-mixed xs) ; may contain foos as well
         ))

On Fri, Sep 26, 2014 at 12:07 PM, Konrad Hinsen <konrad.hinsen at fastmail.net>
wrote:

> Hi everyone,
>
> I am trying to convince Typed Racket that after a run-time check on
> some list, its elements are all of a given type. This is just like
> occurrence typing based on a predicate, except that my predicate
> is more complicated.
>
> My first attempt was this:
>
>
>    #lang typed/racket
>
>    (struct: foo ())
>    (struct: bar ())
>    (define-type FooOrBar (U foo bar))
>
>    (: f-bar (-> (Listof bar) Void))
>    (define (f-bar xs)
>      (void))
>
>    (: f-mixed (-> (Listof FooOrBar) Void))
>    (define (f-mixed xs)
>      (void))
>
>    (: f (-> (Listof FooOrBar) Void))
>    (define (f xs)
>      (if (for/and : Boolean ([x xs]) (bar? x))
>          (f-bar xs) ; contains only bars
>          (f-mixed xs) ; may contain foos as well
>          ))
>
> This yields a type error:
>
>    ; /Users/hinsen/projects/racket/foobar.rkt:18:13: Type Checker: type
> mismatch
>    ;   expected: (Listof bar)
>    ;   given: (Listof FooOrBar)
>    ;   in: xs
>
> After reading section 5 of the Typed Racket Guide, in particular section
> 5.2 entitled "Filters and Predicates", I thought I could simply define my
> own predicate with a type annotation:
>
>    #lang typed/racket
>
>    (struct: foo ())
>    (struct: bar ())
>    (define-type FooOrBar (U foo bar))
>
>    (: f-bar (-> (Listof bar) Void))
>    (define (f-bar xs)
>      (void))
>
>    (: f-mixed (-> (Listof FooOrBar) Void))
>    (define (f-mixed xs)
>      (void))
>
>    (: pure-bars? (-> (Listof FooOrBar) Boolean : (Listof bar)))
>    (define (pure-bars? xs)
>      (for/and : Boolean ([x xs]) (bar? x)))
>
>    (: f (-> (Listof FooOrBar) Void))
>    (define (f xs)
>      (if (pure-bars? xs)
>          (f-bar xs) ; contains only bars
>          (f-mixed xs) ; may contain foos as well
>          ))
>
> But this got me only into more trouble - now I don't even understand the
> error message any more:
>
>    ; /Users/hinsen/projects/racket/foobar.rkt:17:2: Type Checker: type
> mismatch;
>    ;  mismatch in filter
>    ;   expected: (((Listof bar) @ xs) | (! (Listof bar) @ xs))
>    ;   given: (Top | Top)
>    ;   in: (for/and : Boolean ((x xs)) (bar? x))
>
> Is there a way to do this kind of test?
>
> Konrad.
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140926/72332d84/attachment.html>

From konrad.hinsen at fastmail.net  Fri Sep 26 12:28:03 2014
From: konrad.hinsen at fastmail.net (Konrad Hinsen)
Date: Fri, 26 Sep 2014 18:28:03 +0200
Subject: [racket] Tricky case of occurrence typing in Typed Racket
In-Reply-To: <CAMf1y4JeW-jdYeuJP-0x=U=TJB3tahxqvAmuM3TSbeFvGSAYJg@mail.gmail.com>
References: <21541.36663.168649.25871@Konrad-Hinsens-MacBook-Pro-2.local>
 <CAMf1y4JeW-jdYeuJP-0x=U=TJB3tahxqvAmuM3TSbeFvGSAYJg@mail.gmail.com>
Message-ID: <21541.37907.627925.114871@Konrad-Hinsens-MacBook-Pro-2.local>

Andrew Kent writes:

 > Will andmap work for you?

Interesting... I didn't know about that one.

For my demonstration code, that's indeed a good solution. In my real
application, the test is more complicated. I need to check all
elements of a list for conformance to a union type, so I have no
prefabricated predicate, not even for the elements of my list.

Konrad.

From amk.kent at gmail.com  Fri Sep 26 12:48:05 2014
From: amk.kent at gmail.com (Andrew Kent)
Date: Fri, 26 Sep 2014 12:48:05 -0400
Subject: [racket] Tricky case of occurrence typing in Typed Racket
In-Reply-To: <21541.37907.627925.114871@Konrad-Hinsens-MacBook-Pro-2.local>
References: <21541.36663.168649.25871@Konrad-Hinsens-MacBook-Pro-2.local>
 <CAMf1y4JeW-jdYeuJP-0x=U=TJB3tahxqvAmuM3TSbeFvGSAYJg@mail.gmail.com>
 <21541.37907.627925.114871@Konrad-Hinsens-MacBook-Pro-2.local>
Message-ID: <CAMf1y4LenSG_Ti2S_sFeaDXY6Zo_P5nO9C8-RC6q_TNScG4G9w@mail.gmail.com>

Does using andmap with the custom predicates you alluded to for that union
you mentioned work?

Something like this perhaps?:

#lang typed/racket

(struct: T1 ())
(struct: T2 ())
(define-type T1or2 (U T1 T2))

(: T1or2? (-> Any Boolean : T1or2))
(define (T1or2? a)
  (or (T1? a) (T2? a)))


(: listof-T1or2 (-> Any Boolean : (Listof T1or2)))
(define (listof-T1or2 l)
  (and (list? l) (andmap T1or2? l)))



On Fri, Sep 26, 2014 at 12:28 PM, Konrad Hinsen <konrad.hinsen at fastmail.net>
wrote:

> Andrew Kent writes:
>
>  > Will andmap work for you?
>
> Interesting... I didn't know about that one.
>
> For my demonstration code, that's indeed a good solution. In my real
> application, the test is more complicated. I need to check all
> elements of a list for conformance to a union type, so I have no
> prefabricated predicate, not even for the elements of my list.
>
> Konrad.
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140926/30da28f4/attachment.html>

From michaelsbradleyjr at gmail.com  Fri Sep 26 14:15:31 2014
From: michaelsbradleyjr at gmail.com (Michael =?utf-8?b?QnJhZGxleSw=?= Jr.)
Date: Fri, 26 Sep 2014 18:15:31 +0000 (UTC)
Subject: [racket]
	=?utf-8?q?ANN=3A_DOS=2C=09Delimited-continuation-based_O?=
	=?utf-8?q?perating-system_Simulator?=
References: <CAJYbDam0maAG7ZmCBjbha1uqhJyCpX9J+L3=kxqAL4GjdKUvkg@mail.gmail.com>
Message-ID: <loom.20140926T201005-291@post.gmane.org>

Jay McCarthy <jay.mccarthy at ...> writes:

> 
> I've just released the game architecture library I talked about in
> part 2 of my RacketCon talk.
> 
> The big picture of this library is to make World-style programs more
>  compositional by (a) using continuations to hide the internal state
> (including control state) of components and (b) using environments as
> a standard monoid-based inter-component communication channel. A
> monoid is used to ensure that the components can be evaluated in any
> order. Despite assumptions some have about continuations and pure
> functional programming, it is incredibly efficient and can run at
> 60FPS, as demonstrated in get-bonus.
> 
> You can get it with
> 
> raco pkg install dos
> 
> And you can try out the demo with
> 
> racket -l dos/examples/win
> 
> The demo source is a bare 39 lines:
> 
> https://github.com/jeapostrophe/dos/blob/master/dos/examples/win.rkt
> 
> and I provide the non-DOS version for comparison:
> 
> https://github.com/jeapostrophe/dos/blob/master/dos/examples/win-long.rkt
> 
> The core library has a mere 36 lines:
> 
> https://github.com/jeapostrophe/dos/blob/master/dos/main.rkt
>



Hi Jay,

I really enjoyed your talk on get-bonus at RacketCon last weekend, and
appreciate your sharing DOS with all of us. Also, your presentation at
Strange Loop 2013 was my first introduction to delimited continuations in
Racket!

I have benefited from a number of your blog posts in the past, e.g. the one
on an improved threading macro, and was wondering if you would consider
doing a write-up on DOS? 

Best regards,

--
Michael Bradley, Jr.
@michaelsbradley



From alexander at knauth.org  Fri Sep 26 19:35:24 2014
From: alexander at knauth.org (Alexander D. Knauth)
Date: Fri, 26 Sep 2014 19:35:24 -0400
Subject: [racket] typed racket, filters, and polymorphism
In-Reply-To: <F16C1892-F09C-4A9F-94E2-16D5DE72F57D@knauth.org>
References: <D276D7BD-D6AF-4F26-A79B-C1016641EF06@knauth.org>
 <CAK=HD+ZpPuvYTMZJ1_LXrnPGjcPamZ73Z8aiPizCussR+UxYsw@mail.gmail.com>
 <F16C1892-F09C-4A9F-94E2-16D5DE72F57D@knauth.org>
Message-ID: <8CF041CE-65AD-488D-B2AE-FAE6C8C1C352@knauth.org>

Is it possible to have a struct that does certain things according to the guard?
#lang typed/racket

(struct (a) foo ([a : a]) #:transparent
  #:guard (lambda (a _)
            (unless (exact-integer? a)
              (error 'foo "expected Integer, given ~v" a))
            a))

(ann (foo (ann 1 Any)) (foo Integer))

(: x : (foo Any))
(define x (foo 1))

(ann (foo-a x) Integer)

;. Type Checker: Polymorphic function `foo1' could not be applied to arguments:
;Argument 1:
;  Expected: a
;  Given:    Any
;
;Result type:     (foo a)
;Expected result: (foo Integer)
; in: (foo (ann 1 Any))
;. Type Checker: Polymorphic function `foo-a' could not be applied to arguments:
;Argument 1:
;  Expected: (foo a)
;  Given:    (foo Any)
;
;Result type:     (a : ....)
;Expected result: Integer
; in: (foo-a x)

On Sep 25, 2014, at 9:42 PM, Alexander D. Knauth <alexander at knauth.org> wrote:

> What I?m trying to accomplish is something more like this:
> #lang typed/racket
> 
> (require "dimensions.rkt")
> 
> (struct (d) unit ([name : Any] [scalar : Positive-Real] [dimension : d]) #:transparent
>   #:guard (lambda (name scalar dimension _)
>             (unless (dimension? dimension)
>               (error 'unit "expected Dimension, given ~v" dimension))
>             (values name scalar dimension)))
> 
> (define-type (Unitof d) (unit d))
> 
> (define-type Unit (Unitof Dimension))
> 
> (define Unit? (make-predicate Unit))
> 
> (define-type Unitish
>   (U (Unitof Any)
>      Dimension
>      Positive-Real))
> 
> (: ->unit : (All (d) (case-> [(Unitof d) -> (Unitof d)]
>                              [Unitish -> Unit])))
> (define (->unit u)
>   (cond [(unit? u)
>          (unless (Unit? u) ; this should never happen anyway because of the guard
>            (error '->unit "expected (Unitof Dimension), given ~v" u))
>          u]
>         [(dimension? u) (unit u 1 u)]
>         [(positive-real? u) (unit u u dimensionless-dimension)]))
> 
> 
> On Sep 25, 2014, at 6:19 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> wrote:
> 
>> No, I don't think you can do this. Can you say more about what you're
>> trying to accomplish?
>> 
>> Sam
>> 
>> On Thu, Sep 25, 2014 at 6:15 PM, Alexander D. Knauth
>> <alexander at knauth.org> wrote:
>>> Do any of you have any advice for getting a function like this to
>>> type-check?
>>> #lang typed/racket
>>> 
>>> (: check-int : (All (a) (case-> [a -> a]
>>>                                [Any -> Integer])))
>>> (define (check-int int)
>>>  (unless (exact-integer? int)
>>>    (error 'check-int "expected Integer, given ~v" int))
>>>  int)
>>> 
>>> ;. Type Checker: type mismatch
>>> ;  expected: a
>>> ;  given: Integer in: int
>>> 
>>> 
>>> 
>>> ____________________
>>>  Racket Users list:
>>>  http://lists.racket-lang.org/users
>>> 
> 
> ____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140926/bbbbc5ac/attachment.html>

From jay.mccarthy at gmail.com  Fri Sep 26 20:03:57 2014
From: jay.mccarthy at gmail.com (Jay McCarthy)
Date: Fri, 26 Sep 2014 20:03:57 -0400
Subject: [racket] ANN: DOS,
	Delimited-continuation-based Operating-system Simulator
In-Reply-To: <loom.20140926T201005-291@post.gmane.org>
References: <CAJYbDam0maAG7ZmCBjbha1uqhJyCpX9J+L3=kxqAL4GjdKUvkg@mail.gmail.com>
 <loom.20140926T201005-291@post.gmane.org>
Message-ID: <CAJYbDanLdZACFv41w+CAtNjvLernOomC0L7goMVHH-kwe07fXw@mail.gmail.com>

Sure, I can make a post about it. However, the core idea is already in
this post:

http://jeapostrophe.github.io/2012-07-12-cont-sys-post.html

I can make another one, however, to rehash it and talk about the
benefits of monoid states.

Jay

On Fri, Sep 26, 2014 at 2:15 PM, Michael Bradley, Jr.
<michaelsbradleyjr at gmail.com> wrote:
> Jay McCarthy <jay.mccarthy at ...> writes:
>
>>
>> I've just released the game architecture library I talked about in
>> part 2 of my RacketCon talk.
>>
>> The big picture of this library is to make World-style programs more
>>  compositional by (a) using continuations to hide the internal state
>> (including control state) of components and (b) using environments as
>> a standard monoid-based inter-component communication channel. A
>> monoid is used to ensure that the components can be evaluated in any
>> order. Despite assumptions some have about continuations and pure
>> functional programming, it is incredibly efficient and can run at
>> 60FPS, as demonstrated in get-bonus.
>>
>> You can get it with
>>
>> raco pkg install dos
>>
>> And you can try out the demo with
>>
>> racket -l dos/examples/win
>>
>> The demo source is a bare 39 lines:
>>
>> https://github.com/jeapostrophe/dos/blob/master/dos/examples/win.rkt
>>
>> and I provide the non-DOS version for comparison:
>>
>> https://github.com/jeapostrophe/dos/blob/master/dos/examples/win-long.rkt
>>
>> The core library has a mere 36 lines:
>>
>> https://github.com/jeapostrophe/dos/blob/master/dos/main.rkt
>>
>
>
>
> Hi Jay,
>
> I really enjoyed your talk on get-bonus at RacketCon last weekend, and
> appreciate your sharing DOS with all of us. Also, your presentation at
> Strange Loop 2013 was my first introduction to delimited continuations in
> Racket!
>
> I have benefited from a number of your blog posts in the past, e.g. the one
> on an improved threading macro, and was wondering if you would consider
> doing a write-up on DOS?
>
> Best regards,
>
> --
> Michael Bradley, Jr.
> @michaelsbradley
>
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users



-- 
Jay McCarthy
http://jeapostrophe.github.io

           "Wherefore, be not weary in well-doing,
      for ye are laying the foundation of a great work.
And out of small things proceedeth that which is great."
                          - D&C 64:33

From jay.mccarthy at gmail.com  Fri Sep 26 20:08:30 2014
From: jay.mccarthy at gmail.com (Jay McCarthy)
Date: Fri, 26 Sep 2014 20:08:30 -0400
Subject: [racket] web server: module servlets
In-Reply-To: <5424E8B8.8040501@comcast.net>
References: <5421DB30.3030803@comcast.net>
 <CAJYbDa=uR26UAdfPgg5RzLo3HJv_rWosXXiCx+Rp9vbxPEbe9Q@mail.gmail.com>
 <54221C2D.9020109@comcast.net>
 <CAJYbDak29o47-yU5ONQy3DH+V+2DVCzT8UWqtaZNMsHAdb=JdA@mail.gmail.com>
 <5424482C.1030901@comcast.net>
 <CAJYbDa=-DuRMVy7JnCt32Y8waG2+qkHZQw5cVDL4L_d6zth45w@mail.gmail.com>
 <54247C41.1030408@comcast.net>
 <CAJYbDamsu2F=97vnZJQdx2dRO-Ss46NU0icrf8XwNA_2mMg9gw@mail.gmail.com>
 <5424E8B8.8040501@comcast.net>
Message-ID: <CAJYbDa=aYrXcKh6bZK-dBai4CYR8tE+yuXOMHp6P9a-sYs_f4Q@mail.gmail.com>

Hi George,

I suggest that all Racket web-server apps not use the dynamic features
of serve/servlet either, but instead write the servlet as the single
"request -> response" function that serve/servlet provides. This would
ensure that all the libraries are loaded the one time and has the best
performance out of all the options of the Web server. Furthermore,
there are many features (such as total URL control) that are only
available in this model.

In case it is not completely clear, here's a little example (<500
lines) of a Web application that works this way:

https://github.com/plt/racket/blob/master/pkgs/plt-services/meta/pkg-index/official/dynamic.rkt#L423

Here's another that uses 'serve' directly for non-servlet needs:

https://github.com/jeapostrophe/exp/blob/master/dir-serve.rkt

And here's a considerably longer one:

https://github.com/jeapostrophe/grade-samurai/blob/master/app.rkt#L2020

I highly recommend using serve/servlet like this over any other
options. In particular, many of the problems you're talking about were
the reason that I made serve/servlet in the first place.

Jay

On Fri, Sep 26, 2014 at 12:16 AM, George Neuner <gneuner2 at comcast.net> wrote:
> On 9/25/2014 6:26 PM, Jay McCarthy wrote:
>
> web-server/servlet is NOT shared, nor is any net library or common
> racket library like racket/list. They are all totally unique per
> servlet.
>
> This, by the way, is part of why I don't recommend using dynamic
> servlets at all and suggest using serve/servlet.
>
>
> But ... IIUC ... a listener started by  serve/servlet  will still
> dynamically load from <servlet-root>, <server-root>/htdocs/*, etc. if the
> request doesn't match a hard coded dispatch URL - the regex of the initial
> servlet or an entry in a dispatcher table.  At least that's the behavior I
> see with my own application: my initial servlet returns 404 when called -
> the call to  serve/servlet  just sets up the environment and and all the
> "real" servlets are demand loaded from disk when first touched.
>
> Are you advocating *static* linking and essentially just dispatching to
> internal functions by URL?  ISTM that that defeats the purpose.
>
> George



-- 
Jay McCarthy
http://jeapostrophe.github.io

           "Wherefore, be not weary in well-doing,
      for ye are laying the foundation of a great work.
And out of small things proceedeth that which is great."
                          - D&C 64:33

From acarrico at memebeam.org  Fri Sep 26 23:07:55 2014
From: acarrico at memebeam.org (Anthony Carrico)
Date: Fri, 26 Sep 2014 23:07:55 -0400
Subject: [racket] typed racket and generic interfaces,
 is there a workaround using properties?
In-Reply-To: <B952C6DE-40D5-4A84-8BCC-608F06E31261@knauth.org>
References: <2021E4C9-B660-444D-8E4B-7C41169D61F7@knauth.org>
 <B952C6DE-40D5-4A84-8BCC-608F06E31261@knauth.org>
Message-ID: <54262A0B.2010901@memebeam.org>

Dredging up an old thread: So how exactly do you define custom-write for
a typed racket struct?

On 05/25/2014 03:30 PM, Alexander D. Knauth wrote:
> Never mind I just found prop:dict.
> 
> By the way it would probably be helpful if instead of the value for
> prop:dict being a vector, it was a hash table with the method names
> as the keys, so then it would be a lot easier to tell which procedure
> goes with which method, and also a lot easier to extend it while
> maintaining backwards compatibility.
> 
> On May 24, 2014, at 10:38 PM, Alexander D. Knauth
> <alexander at knauth.org> wrote:
> 
>> Do generic interfaces work using structure type properties, and if
>> they do, is there a way to use generic interfaces through
>> properties so that I can do it in typed racket?
>> 
>> Specifically I?m trying to use the gen:custom-write and gen:dict
>> generic interfaces.  I can use prop:custom-write for the first one,
>> but I don?t know what to do for gen:dict.
>> 
>> Otherwise I?ll just put it in an untyped submodule.

-- 
Anthony Carrico

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 181 bytes
Desc: OpenPGP digital signature
URL: <http://lists.racket-lang.org/users/archive/attachments/20140926/d14018a4/attachment-0001.sig>

From martindemello at gmail.com  Sat Sep 27 00:56:52 2014
From: martindemello at gmail.com (Martin DeMello)
Date: Fri, 26 Sep 2014 21:56:52 -0700
Subject: [racket] "do what i mean" for https://github links
Message-ID: <CAFrFfuH93VtSwahAz7Zo9EBGoTgP2_z=xZU04WdjrmEerd0Smw@mail.gmail.com>

It took me a while of reading through the docs to figure out why trying to
install a package via "https://github.com/user/pkg" failed for want of a
MANIFEST file. I think the "do what I mean" installer should have
autoconverted it to git:// , seeing as how github is already special-cased.

(Also my first try of installing via the git clone url
https://github.com/user/pkg.git should ideally have been supported too,
since that's what github encourages you to copy/paste as a reference)

martin
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140926/a07de57f/attachment.html>

From spencerflorence at gmail.com  Sat Sep 27 01:07:43 2014
From: spencerflorence at gmail.com (Spencer florence)
Date: Fri, 26 Sep 2014 22:07:43 -0700 (PDT)
Subject: [racket] typed racket and generic interfaces,
 is there a workaround using properties?
In-Reply-To: <2021E4C9-B660-444D-8E4B-7C41169D61F7@knauth.org>
References: <2021E4C9-B660-444D-8E4B-7C41169D61F7@knauth.org>
Message-ID: <1411794462639.6720ba81@Nodemailer>

I don?t think you can. You would need define the struct in an untyped module then require it via require/typed. This is why dict?s don?t work in typed/racket either.

On Sat, May 24, 2014 at 9:39 PM, Alexander D. Knauth <alexander at knauth.org>
wrote:

> Do generic interfaces work using structure type properties, and if they do, is there a way to use generic interfaces through properties so that I can do it in typed racket?  
> Specifically I?m trying to use the gen:custom-write and gen:dict generic interfaces.  I can use prop:custom-write for the first one, but I don?t know what to do for gen:dict.  
> Otherwise I?ll just put it in an untyped submodule.  
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140926/e61d292b/attachment.html>

From martindemello at gmail.com  Sat Sep 27 03:27:29 2014
From: martindemello at gmail.com (Martin DeMello)
Date: Sat, 27 Sep 2014 00:27:29 -0700
Subject: [racket] problem with pict->bitmap and mrlib/write-animated-gif
Message-ID: <CAFrFfuHsyYBhWB01PnOf8E5HzwDM_69n0nAKfLyNn2UY1+nkhQ@mail.gmail.com>

Can't figure it out, but something in the interaction between pict->bitmap
and write-animated-gif is causing the frames to display one on top of the
other when viewing the gif.

#lang racket

(require pict
         mrlib/gif)

(define (draw-frame i)
  (pict->bitmap (circle (* 50 i))))

(write-animated-gif
 (map draw-frame (sequence->list (in-range 1 10)))
 10
 "test1.gif"
 #:loop? true
 #:one-at-a-time? true)

martin
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140927/9a4a2d74/attachment.html>

From richter at math.northwestern.edu  Sat Sep 27 03:38:48 2014
From: richter at math.northwestern.edu (Bill Richter)
Date: Sat, 27 Sep 2014 02:38:48 -0500
Subject: [racket] proof assistants, DrRacket and Bootstrap
Message-ID: <201409270738.s8R7cmcd022184@poisson.math.northwestern.edu>

I have a few questions that might be off-topic.  Are you interested in formal proofs?  Have you considered adapting DrRacket to give an integrated editor for a proof assistant?  The proof assistants Coq and Isabelle use jedit and ProofGeneral, which I think aren't nearly as nice as DrRacket.  I actually use HOL Light, which nobody uses an integrated editor for.  Here are the slides for a talk I gave at the Institut Henri Poincar? in the workshop ``Formalization of mathematics in proof assistants''
http://www.math.northwestern.edu/~richter/RichterIHPslide.pdf

HOL Light and Coq are written in OCaml, a dialect of ML, which is therefore similar to Scheme, but it has one difference that I wonder if anyone here's knows how to deal with.  Scheme is well-suited for writing a Scheme interpreter, because of the quote feature.  OCaml doesn't have a quote feature, so the question arises how to write an OCaml interpreter inside OCaml.  That's not quite what I want to do, but if anyone could explain how to do it, I'd be grateful.

I have a 7th grade student I'm trying to teach Racket, and I'm a bit confused about Bootstrap and Program By Design.  Is there any reason for my student not to just read HtDP?

-- 
Best,
Bill 

From robby at eecs.northwestern.edu  Sat Sep 27 08:38:54 2014
From: robby at eecs.northwestern.edu (Robby Findler)
Date: Sat, 27 Sep 2014 07:38:54 -0500
Subject: [racket] Fwd: JFP CFP
In-Reply-To: <ECBE1B1A-9676-4D4A-9B9A-E41CFF8C24C1@cse.unsw.edu.au>
References: <ECBE1B1A-9676-4D4A-9B9A-E41CFF8C24C1@cse.unsw.edu.au>
Message-ID: <CAL3TdOMU-PHh++5qUz2RE9fRxHGd-zjB8=nMNdQA8_FOMSbiag@mail.gmail.com>

If you're working on parallel or concurrent programming, consider submitting!

Robby


------------------------------------------------------------------------------
                        CALL FOR PAPERS

                       JFP Special Issue
                                on
                  Parallel and Concurrent Programming

                Submission Deadline: 22 December 2014
       Expected Publication Date: November/December issue 2015


* Scope

Functional languages are uniquely suited to provide programmers with a
programming model for parallel and concurrent computing. This is reflected in
the wide range of work that is currently underway, both on parallel and
concurrent functional languages, but also on bringing functional language
features to other programming languages. This has resulted in a rapidly
growing number of practical applications.

The Journal of Functional Programming will devote a special issue to
functional programming for concurrent and parallel computing. The purpose of
this special issue is to showcase the state of the art on how functional
languages and functional concepts currently assist programmers with the task
of managing the challenges of creating parallel and concurrent systems.

Language designers as well as systems builders and application programmers are
invited to submit papers describing the current state of the art of language
support and experiences with functional programming in building real-world
systems.

We encourage the submission of consolidated, condensed and extended work based
on prior conference and workshop publications.

* Submission Details

Manuscripts should be submitted to CUP's Manuscript Central formatted
according to JFP's standards.  Authors should specify in their cover message
that the paper is intended for the special issue and choose the appropriate
desired editor from the menu. For submission and formatting details, please
consult the Journal of Functional Programming [ http://bit.ly/jfpadvice ]
website.


Guest Editors
---------------------------------------------------------------------------------
Gabriele Keller                             Fritz Henglein
keller @cse.unsw.edu.au                     henglein at diku.dk
School of Computer Science & Engineering    Department of Computer
Science (DIKU)
UNSW Sydney NSW 2052                        University of Copenhagen
and                                         Universitetsparken 5
The Software Systems Research Group@ NICTA  DK-2100 Copenhagen
Australia                                   Denmark
---------------------------------------------------------------------------------

From alexander at knauth.org  Sat Sep 27 09:49:48 2014
From: alexander at knauth.org (Alexander D. Knauth)
Date: Sat, 27 Sep 2014 09:49:48 -0400
Subject: [racket] typed racket and generic interfaces,
	is there a workaround using properties?
In-Reply-To: <54262A0B.2010901@memebeam.org>
References: <2021E4C9-B660-444D-8E4B-7C41169D61F7@knauth.org>
 <B952C6DE-40D5-4A84-8BCC-608F06E31261@knauth.org>
 <54262A0B.2010901@memebeam.org>
Message-ID: <39AEBCAD-6138-4FE6-84A5-6AFF1A1A466C@knauth.org>

Like this, I think:  (but for some reason it seems to print it 3 times?)
#lang typed/racket

(struct foo () #:transparent
  #:property prop:custom-write
  (lambda (this out mode)
    (displayln "whatever")))

(print (foo))


On Sep 26, 2014, at 11:07 PM, Anthony Carrico <acarrico at memebeam.org> wrote:

> Dredging up an old thread: So how exactly do you define custom-write for
> a typed racket struct?
> 
> On 05/25/2014 03:30 PM, Alexander D. Knauth wrote:
>> Never mind I just found prop:dict.
>> 
>> By the way it would probably be helpful if instead of the value for
>> prop:dict being a vector, it was a hash table with the method names
>> as the keys, so then it would be a lot easier to tell which procedure
>> goes with which method, and also a lot easier to extend it while
>> maintaining backwards compatibility.
>> 
>> On May 24, 2014, at 10:38 PM, Alexander D. Knauth
>> <alexander at knauth.org> wrote:
>> 
>>> Do generic interfaces work using structure type properties, and if
>>> they do, is there a way to use generic interfaces through
>>> properties so that I can do it in typed racket?
>>> 
>>> Specifically I?m trying to use the gen:custom-write and gen:dict
>>> generic interfaces.  I can use prop:custom-write for the first one,
>>> but I don?t know what to do for gen:dict.
>>> 
>>> Otherwise I?ll just put it in an untyped submodule.
> 
> -- 
> Anthony Carrico
> 
> ____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users



From plragde at uwaterloo.ca  Sat Sep 27 10:33:15 2014
From: plragde at uwaterloo.ca (Prabhakar Ragde)
Date: Sat, 27 Sep 2014 10:33:15 -0400
Subject: [racket]  proof assistants, DrRacket and Bootstrap
In-Reply-To: <201409270738.s8R7cmcd022184@poisson.math.northwestern.edu>
Message-ID: <5426CAAB.2010400@uwaterloo.ca>

> I have a few questions that might be off-topic.  Are you interested
> in formal proofs?  Have you considered adapting DrRacket to give an
> integrated editor for a proof assistant?  The proof assistants Coq
> and Isabelle use jedit and ProofGeneral, which I think aren't nearly
> as nice as DrRacket.  I actually use HOL Light, which nobody uses an
> integrated editor for.  Here are the slides for a talk I gave at the
> Institut Henri Poincar? in the workshop ``Formalization of
> mathematics in proof assistants''
> http://www.math.northwestern.edu/~richter/RichterIHPslide.pdf

I am teaching a grad course using Coq right now, and using Proof General 
within Emacs for it. That is not bad, but I would love to have a 
DrRacket interface. I know something like this is possible, because at 
Brown they used DrRacket to prepare and run OCaml programs, a few years 
ago. I really don't have the time to work on this, though. (And if I 
did, I would first work on teaching language subsets of Haskell, which I 
also would like to have inside DrRacket.)

> HOL Light and Coq are written in OCaml, a dialect of ML, which is
> therefore similar to Scheme, but it has one difference that I wonder
> if anyone here's knows how to deal with.  Scheme is well-suited for
> writing a Scheme interpreter, because of the quote feature.  OCaml
> doesn't have a quote feature, so the question arises how to write an
> OCaml interpreter inside OCaml.  That's not quite what I want to do,
> but if anyone could explain how to do it, I'd be grateful.

In my senior undergraduate programming language course, I have them 
write various interpreters and typecheckers in various languages, 
including Racket, OCaml, and Haskell. In the latter two, you need to use 
algebraic datatypes to build an abstract syntax tree, and interpret 
that. So the expression (* (+ 1 2) (- 3 4)) might become

(Mult (Plus (Num 1) (Num 2)) (Minus (Num 3) (Num 4)))

Writing an interpreter for such an AST is even easier than in 
Racket/Scheme, due to concise pattern-matching syntax. But parsing an 
expression written in faux OCaml/Haskell is quite complicated. I don't 
ask students to do that, since it really is part of a separate course 
(both languages come with general parsing tools that can be used). 
Parsing is much easier in Racket/Scheme because of the similarity of 
code/data and the presence of 'read'.

> I have a 7th grade student I'm trying to teach Racket, and I'm a bit
> confused about Bootstrap and Program By Design.  Is there any reason
> for my student not to just read HtDP?

You should view word from the authors/developers as definitive, but I 
would say that if your student has good algebraic skills, then using 
HtDP/2e is the right path. If not, Bootstrap may be a way to help 
develop those skills. --PR

From alexander at knauth.org  Sat Sep 27 13:55:12 2014
From: alexander at knauth.org (Alexander D. Knauth)
Date: Sat, 27 Sep 2014 13:55:12 -0400
Subject: [racket] typed racket and generic interfaces,
	is there a workaround using properties?
In-Reply-To: <1411794462639.6720ba81@Nodemailer>
References: <2021E4C9-B660-444D-8E4B-7C41169D61F7@knauth.org>
 <1411794462639.6720ba81@Nodemailer>
Message-ID: <38E7C628-8FA0-4F07-ABAD-E271E2A66426@knauth.org>

Actually using prop:dict works (I hadn?t found prop:dict yet when I first asked this)
#lang typed/racket

(define-type Dict-Ref ([Dict Any] [Any] . ->* . Any))
(define-type Dict-Set! (Dict Any Any . -> . Void))
(define-type Dict-Set (Dict Any Any . -> . Dict))
(define-type Dict-Remove! (Dict Any . -> . Void))
(define-type Dict-Remove (Dict Any . -> . Dict))
(define-type Dict-Iterate-First (Dict . -> . Any))
(define-type Dict-Iterate-Next (Dict Any . -> . Any))
(define-type Dict-Iterate-Key (Dict Any . -> . Any))
(define-type Dict-Iterate-Value (Dict Any . -> . Any))
(define-type Dict-Count (Dict . -> . Natural))

(require/typed racket/dict
               [#:opaque Dict dict?]
               [dict-ref Dict-Ref]
               [dict-set! Dict-Set!]
               [dict-set Dict-Set]
               [dict-remove! Dict-Remove!]
               [dict-remove Dict-Remove]
               [dict-iterate-first Dict-Iterate-First]
               [dict-iterate-next Dict-Iterate-Next]
               [dict-iterate-key Dict-Iterate-Key]
               [dict-iterate-value Dict-Iterate-Value]
               [dict-count Dict-Count]
               [prop:dict Struct-Type-Property])

(define (make-dict-prop #:dict-ref [dict-ref : Dict-Ref]
                        #:dict-set! [dict-set! : (U Dict-Set! #f) #f]
                        #:dict-set [dict-set : (U Dict-Set #f) #f]
                        #:dict-remove! [dict-remove! : (U Dict-Remove! #f) #f]
                        #:dict-remove [dict-remove : (U Dict-Remove #f) #f]
                        #:dict-count [dict-count : Dict-Count]
                        #:dict-iterate-first [dict-iterate-first : Dict-Iterate-First]
                        #:dict-iterate-next [dict-iterate-next : Dict-Iterate-Next]
                        #:dict-iterate-key [dict-iterate-key : Dict-Iterate-Key]
                        #:dict-iterate-value [dict-iterate-value : Dict-Iterate-Value])
  (vector-immutable dict-ref
                    dict-set!
                    dict-set
                    dict-remove!
                    dict-remove
                    dict-count
                    dict-iterate-first
                    dict-iterate-next
                    dict-iterate-key
                    dict-iterate-value))

(struct foo () #:transparent
  #:property prop:dict
  (make-dict-prop #:dict-ref (lambda (this key [failure #f]) "whatever")
                  #:dict-count (lambda (this) 0)
                  #:dict-iterate-first (lambda (this) #f)
                  #:dict-iterate-next (lambda (this pos) 0)
                  #:dict-iterate-key (lambda (this pos) pos)
                  #:dict-iterate-value (lambda (this pos) "whatever")))

(dict-ref (assert (foo) dict?) "idontkare")


On Sep 27, 2014, at 1:07 AM, Spencer florence <spencerflorence at gmail.com> wrote:

> I don?t think you can. You would need define the struct in an untyped module then require it via require/typed. This is why dict?s don?t work in typed/racket either.
> 
> 
> 
> On Sat, May 24, 2014 at 9:39 PM, Alexander D. Knauth <alexander at knauth.org> wrote:
> 
> Do generic interfaces work using structure type properties, and if they do, is there a way to use generic interfaces through properties so that I can do it in typed racket? 
> 
> Specifically I?m trying to use the gen:custom-write and gen:dict generic interfaces. I can use prop:custom-write for the first one, but I don?t know what to do for gen:dict. 
> 
> Otherwise I?ll just put it in an untyped submodule. 
> 
> 
> ____________________ 
> Racket Users list: 
> http://lists.racket-lang.org/users 
> 
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140927/17fc9f08/attachment.html>

From ianj at ccs.neu.edu  Sat Sep 27 16:09:09 2014
From: ianj at ccs.neu.edu (J. Ian Johnson)
Date: Sat, 27 Sep 2014 16:09:09 -0400 (EDT)
Subject: [racket] proof assistants, DrRacket and Bootstrap
In-Reply-To: <201409270738.s8R7cmcd022184@poisson.math.northwestern.edu>
Message-ID: <28884853.169131411848549503.JavaMail.root@zimbra>

To answer the first part of your email, yes. Carl Eastlund developed Dracula for Rackety use of the ACL2 theorem prover.
-Ian
----- Original Message -----
From: Bill Richter <richter at math.northwestern.edu>
To: users at racket-lang.org
Sent: Sat, 27 Sep 2014 03:38:48 -0400 (EDT)
Subject: [racket] proof assistants, DrRacket and Bootstrap

I have a few questions that might be off-topic.  Are you interested in formal proofs?  Have you considered adapting DrRacket to give an integrated editor for a proof assistant?  The proof assistants Coq and Isabelle use jedit and ProofGeneral, which I think aren't nearly as nice as DrRacket.  I actually use HOL Light, which nobody uses an integrated editor for.  Here are the slides for a talk I gave at the Institut Henri Poincar? in the workshop ``Formalization of mathematics in proof assistants''
http://www.math.northwestern.edu/~richter/RichterIHPslide.pdf

HOL Light and Coq are written in OCaml, a dialect of ML, which is therefore similar to Scheme, but it has one difference that I wonder if anyone here's knows how to deal with.  Scheme is well-suited for writing a Scheme interpreter, because of the quote feature.  OCaml doesn't have a quote feature, so the question arises how to write an OCaml interpreter inside OCaml.  That's not quite what I want to do, but if anyone could explain how to do it, I'd be grateful.

I have a 7th grade student I'm trying to teach Racket, and I'm a bit confused about Bootstrap and Program By Design.  Is there any reason for my student not to just read HtDP?

-- 
Best,
Bill 
____________________
  Racket Users list:
  http://lists.racket-lang.org/users



From norman at astro.gla.ac.uk  Sat Sep 27 16:14:55 2014
From: norman at astro.gla.ac.uk (Norman Gray)
Date: Sat, 27 Sep 2014 21:14:55 +0100
Subject: [racket] aws/glacier: credential scope
Message-ID: <0CA916BA-DD1E-47E9-BFA1-7092EC3F17F5@astro.gla.ac.uk>


Greetings.

I'm trying to use the aws/glacier package, and running into a problem where I'm being told:

    Credential should be scoped to a valid region, not 'eu-west-1'

I'm following the instructions at <https://github.com/greghendershott/aws/blob/master/aws/manual.md>

My test code is:

    % cat glacier.rkt
    #lang racket/base
    
    (require aws/glacier
             aws/keys)
    
    (define vault "testvault")
    (region "eu-west-1")
    (read-keys "aws-zbu-credentials") ; local file
    
    (module+ main
      (printf "region=~a~%" (region))
      (printf "Vaults: ~s~%" (list-vaults))
      (printf "...specifically: ~s~%" (describe-vault vault)))

Running this produces:

    % racket glacier.rkt                
    region=eu-west-1
    aws: HTTP/1.1 403 Forbidden
    x-amzn-RequestId: Un3-L2zlaJBPyrIVKJrWuQcqtMMYQAr34gYUOSScg6Qepc4
    Content-Type: application/json
    Content-Length: 129
    Date: Sat, 27 Sep 2014 18:35:50 GMT
    
     {"message":"Credential should be scoped to a valid region, not 'eu-west-1'. ","code":"InvalidSignatureException","type":"Client"}
    HTTP 403 "Forbidden". AWS Code="InvalidSignatureException" Message="Credential should be scoped to a valid region, not 'eu-west-1'. "
      context...:
       check-response
       /Users/norman/Library/Racket/6.1/pkgs/aws/aws/glacier.rkt:97:22: temp68
       request/redirect/uri
       (submod /checkouts/me/code/zbu/glacier.rkt main): [running body]

Things I thought of:

  * Printing (public-key)/(private-key) indicates that the credentials are being read correctly.
  * When I change the argument of (region) to "us-west-1", that's the region that appears in the error message.
  * My "testvault" vault is in eu-west-1 (and this is indeed one of the valid regions for glacier, reported in <http://docs.aws.amazon.com/general/latest/gr/rande.html> and which does have a host at http://glacier.eu-west-1.amazonaws.com
  * As far as I can see, credentials are _not_ scoped, but are all at us-east-1.
  * <http://docs.aws.amazon.com/general/latest/gr/signature-v4-troubleshooting.html> says that "IAM [...] accepts only us-east-1 as its region specification", so I'm taking it that (region) is for setting the _vault_'s region.
  * I'm not a great AWS expert, so I could have something in my setup broken; but if so, I've no clue what.

If, however, I change the (region) argument to "us-east-1", I get a different error message "User: arn:aws:iam::786725553169:user/zbu is not authorized to perform: glacier:ListVaults on resource: arn:aws:glacier:us-east-1:786725553169:vaults/"  That makes sense, since there's no such vault, but it's interesting that it gets _further_ when the (region) matches the region for the IAM service.

I don't see any other (region) equivalents for the other services supported by the package.  Is that because all of the other services supported by the package are supported by all the AWS regions, or am I missing a configuration?

Thanks for any pointers.

All the best,

Norman


-- 
Norman Gray  :  http://nxg.me.uk
SUPA School of Physics and Astronomy, University of Glasgow, UK



From richter at math.northwestern.edu  Sat Sep 27 17:08:49 2014
From: richter at math.northwestern.edu (Bill Richter)
Date: Sat, 27 Sep 2014 16:08:49 -0500
Subject: [racket] proof assistants, DrRacket and Bootstrap
In-Reply-To: <28884853.169131411848549503.JavaMail.root@zimbra>
 (ianj@ccs.neu.edu)
References: <28884853.169131411848549503.JavaMail.root@zimbra>
Message-ID: <201409272108.s8RL8nqh030467@poisson.math.northwestern.edu>

   Carl Eastlund developed Dracula for Rackety use of the ACL2 theorem prover.

Thanks, Ian!  May I suggest that you try to handle Coq, Isabelle or HOL Light?  I believe these are the foremost proof assistants.  The fields medalist Vladimir Voevodsky uses Coq, in which the 4-color theorem and the Feit-Thompson theorem was formalized by Georges Gonthier.  Tom Hales just finished formalizing his proof of Kepler conjecture in HOL Light and Isabelle (which is HOL as well).  I'm no expert, but I think that ACL2 (also Prover9) is an FOL prover, and the bulk of activity in formal proofs uses type theories.  HOL is simple, it's Church's simple types (a version of the Lambda Calculus), and Coq uses a much more complicated type theory.

-- 
Best,
Bill 

From richter at math.northwestern.edu  Sat Sep 27 17:50:11 2014
From: richter at math.northwestern.edu (Bill Richter)
Date: Sat, 27 Sep 2014 16:50:11 -0500
Subject: [racket] proof assistants, DrRacket and Bootstrap
In-Reply-To: <5426CAAB.2010400@uwaterloo.ca> (message from Prabhakar Ragde on
 Sat, 27 Sep 2014 10:33:15 -0400)
References: <5426CAAB.2010400@uwaterloo.ca>
Message-ID: <201409272150.s8RLoBQW030663@poisson.math.northwestern.edu>

Thanks, Prabhakar, this is exactly what I wanted, and you know something about it (I was barely able to install Coq):

   I am teaching a grad course using Coq right now, and using Proof
   General within Emacs for it. That is not bad, but I would love to
   have a DrRacket interface.

   > OCaml doesn't have a quote feature, so the question arises how to
   > write an OCaml interpreter inside OCaml.  

   In my senior undergraduate programming language course, I have them
   write various interpreters and typecheckers in various languages,
   including Racket, OCaml, and Haskell. 
   [...]

I know this is way off-topic, but it seems to me that Schemers are the only people who want an elegant solution here: 

   But parsing an expression written in faux OCaml/Haskell is quite
   complicated. [...] Parsing is much easier in Racket/Scheme because
   of the similarity of code/data and the presence of 'read'.

I think it's a parsing question I'm stuck on.  I think I need to use camlp to avoid the following hack/kludge which I learned from the HOL Light experts, found in my file 
hol_light/RichterHilbertAxiomGeometry/readable.ml
which is part of the HOL Light distribution:

(* From update_database.ml: Execute any OCaml expression given as a string.  *)

let exec = ignore o Toploop.execute_phrase false Format.std_formatter
  o !Toploop.parse_toplevel_phrase o Lexing.from_string;;

(* Following miz3.ml, exec_thm returns the theorem representing a string, so *)
(* exec_thm "FORALL_PAIR_THM";; returns                                      *)
(* val it : thm = |- !P. (!p. P p) <=> (!p1 p2. P (p1,p2))                   *)

let thm_ref = ref TRUTH;;

let exec_thm s =
  if Str.string_match (Str.regexp "[^;]*;") s 0 then raise Noparse
  else
    try exec ("thm_ref := (("^ s ^"): thm);;");
      !thm_ref
    with _ -> raise Noparse;;

When I posted my question on the OCaml list, the experts said don't ever use Obj.magic, but I think they didn't think much of me using Toploop either.

-- 
Best,
Bill 

From plragde at uwaterloo.ca  Sat Sep 27 18:00:24 2014
From: plragde at uwaterloo.ca (Prabhakar Ragde)
Date: Sat, 27 Sep 2014 18:00:24 -0400
Subject: [racket] proof assistants, DrRacket and Bootstrap
Message-ID: <54273378.1070605@uwaterloo.ca>

Bill Richter wrote:

> >Carl Eastlund developed Dracula for Rackety use of the ACL2 theorem
> >prover.
>
> Thanks, Ian!  May I suggest that you try to handle Coq, Isabelle or
> HOL Light?  I believe these are the foremost proof assistants.  The
> fields medalist Vladimir Voevodsky uses Coq, in which the 4-color
> theorem and the Feit-Thompson theorem was formalized by Georges
> Gonthier.  Tom Hales just finished formalizing his proof of Kepler
> conjecture in HOL Light and Isabelle (which is HOL as well).  I'm no
> expert, but I think that ACL2 (also Prover9) is an FOL prover, and
> the bulk of activity in formal proofs uses type theories.  HOL is
> simple, it's Church's simple types (a version of the Lambda
> Calculus), and Coq uses a much more complicated type theory.

I'm sorry, I forgot about Carl's Dracula work when composing my earlier 
reply. ACL2 is more limited, but its advantage (besides being probably 
far and away the theorem prover that has had the most industrial impact) 
is that its language (Applicative Common Lisp, hence the acronym) is 
much closer to Racket, allowing for stronger integration both with 
DrRacket and into a Racket-based curriculum. A DrRacket interface to Coq 
or Isabelle would, in contrast, be more superficial, in the style of 
Proof General. --PR

From richter at math.northwestern.edu  Sat Sep 27 18:26:44 2014
From: richter at math.northwestern.edu (Bill Richter)
Date: Sat, 27 Sep 2014 17:26:44 -0500
Subject: [racket] proof assistants, DrRacket and Bootstrap
In-Reply-To: <54273378.1070605@uwaterloo.ca> (message from Prabhakar Ragde on
 Sat, 27 Sep 2014 18:00:24 -0400)
References: <54273378.1070605@uwaterloo.ca>
Message-ID: <201409272226.s8RMQiEM030897@poisson.math.northwestern.edu>

   ACL2 [is] probably far and away the theorem prover that has had the
   most industrial impact

Interesting, Prabhakar!  I had not heard that before.  

   A DrRacket interface to Coq or Isabelle would, in contrast, be more
   superficial, in the style of Proof General.

I think so, but interfaces are really important, and DrRacket has a beautiful interface. 

-- 
Best,
Bill 

From martindemello at gmail.com  Sat Sep 27 18:46:25 2014
From: martindemello at gmail.com (Martin DeMello)
Date: Sat, 27 Sep 2014 15:46:25 -0700
Subject: [racket] problem with pict->bitmap and mrlib/write-animated-gif
In-Reply-To: <CAFrFfuHsyYBhWB01PnOf8E5HzwDM_69n0nAKfLyNn2UY1+nkhQ@mail.gmail.com>
References: <CAFrFfuHsyYBhWB01PnOf8E5HzwDM_69n0nAKfLyNn2UY1+nkhQ@mail.gmail.com>
Message-ID: <CAFrFfuG9GXu+N6Tp6Ccg1-n1QX6KFybMRpyj9p_Drk2+SRcrKw@mail.gmail.com>

Solved, with help from Jens Soegaard; I had to modify mrlib/write-gifs to
take a disposal argument, and use it in all calls to gif-add-control.

martin

On Sat, Sep 27, 2014 at 12:27 AM, Martin DeMello <martindemello at gmail.com>
wrote:

> Can't figure it out, but something in the interaction between pict->bitmap
> and write-animated-gif is causing the frames to display one on top of the
> other when viewing the gif.
>
> #lang racket
>
> (require pict
>          mrlib/gif)
>
> (define (draw-frame i)
>   (pict->bitmap (circle (* 50 i))))
>
> (write-animated-gif
>  (map draw-frame (sequence->list (in-range 1 10)))
>  10
>  "test1.gif"
>  #:loop? true
>  #:one-at-a-time? true)
>
> martin
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140927/a42ef6eb/attachment.html>

From plragde at uwaterloo.ca  Sat Sep 27 18:54:56 2014
From: plragde at uwaterloo.ca (Prabhakar Ragde)
Date: Sat, 27 Sep 2014 18:54:56 -0400
Subject: [racket] proof assistants, DrRacket and Bootstrap
In-Reply-To: <201409272150.s8RLoBQW030663@poisson.math.northwestern.edu>
References: <5426CAAB.2010400@uwaterloo.ca>
 <201409272150.s8RLoBQW030663@poisson.math.northwestern.edu>
Message-ID: <54274040.5070902@uwaterloo.ca>

On 2014-09-27, 5:50 PM, Bill Richter wrote:
> Thanks, Prabhakar, this is exactly what I wanted, and you know something about it (I was barely able to install Coq):

I struggled with installing it also, especially since I'm using a Mac. 
Racket has spoiled us in that respect.

>     But parsing an expression written in faux OCaml/Haskell is quite
>     complicated. [...] Parsing is much easier in Racket/Scheme because
>     of the similarity of code/data and the presence of 'read'.
>
> I think it's a parsing question I'm stuck on.  I think I need to use camlp to avoid the following hack/kludge which I learned from the HOL Light experts, found in my file
> hol_light/RichterHilbertAxiomGeometry/readable.ml
> which is part of the HOL Light distribution:
>
> (* From update_database.ml: Execute any OCaml expression given as a string.  *)

I don't have enough practice in OCaml to grasp what your code does (and, 
yes, it is way off-topic), but I do know that about once every month or 
two, someone posts to this list about using 'eval' in Racket, and they 
are warned not to use it unless they really know what they are doing and 
have a suitable application (which is rare). That warning probably is 
doubled in a statically-typed language, in which 'eval' should not even 
be possible without subverting the type system (I gather this is what 
you are doing). Translating the gist of suggestions to such posters, I 
would say: do you really need to execute *any* expression? If not, use 
OCamllex and/or OCamlyacc (maybe Menhir, or a smaller and cleaner parser 
combinator library) to write a parser for your language subset, and then 
write an interpreter for the resulting ASTs. --PR

From acarrico at memebeam.org  Sat Sep 27 21:11:08 2014
From: acarrico at memebeam.org (Anthony Carrico)
Date: Sat, 27 Sep 2014 21:11:08 -0400
Subject: [racket] typed racket and generic interfaces,
 is there a workaround using properties?
In-Reply-To: <39AEBCAD-6138-4FE6-84A5-6AFF1A1A466C@knauth.org>
References: <2021E4C9-B660-444D-8E4B-7C41169D61F7@knauth.org>
 <B952C6DE-40D5-4A84-8BCC-608F06E31261@knauth.org>
 <54262A0B.2010901@memebeam.org>
 <39AEBCAD-6138-4FE6-84A5-6AFF1A1A466C@knauth.org>
Message-ID: <5427602C.5010101@memebeam.org>

On 09/27/2014 09:49 AM, Alexander D. Knauth wrote:
> Like this, I think:  (but for some reason it seems to print it 3 times?)
> #lang typed/racket
> 
> (struct foo () #:transparent
>   #:property prop:custom-write
>   (lambda (this out mode)
>     (displayln "whatever")))
> 
> (print (foo))

I didn't realize #:property was legal in typed struct. I guess there
should be a documentation update here:
  http://docs.racket-lang.org/ts-reference/special-forms.html

BTW change your program to use the given port: (displayln "whatever"
out) to avoid the double printing problem, although I'm not exactly sure
why it would have that effect.

Thanks!

-- 
Anthony Carrico


-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 181 bytes
Desc: OpenPGP digital signature
URL: <http://lists.racket-lang.org/users/archive/attachments/20140927/63d8e553/attachment-0001.sig>

From samth at cs.indiana.edu  Sat Sep 27 21:58:14 2014
From: samth at cs.indiana.edu (Sam Tobin-Hochstadt)
Date: Sat, 27 Sep 2014 21:58:14 -0400
Subject: [racket] typed racket and generic interfaces,
 is there a workaround using properties?
In-Reply-To: <5427602C.5010101@memebeam.org>
References: <2021E4C9-B660-444D-8E4B-7C41169D61F7@knauth.org>
 <B952C6DE-40D5-4A84-8BCC-608F06E31261@knauth.org>
 <54262A0B.2010901@memebeam.org>
 <39AEBCAD-6138-4FE6-84A5-6AFF1A1A466C@knauth.org>
 <5427602C.5010101@memebeam.org>
Message-ID: <CAK=HD+aVN_KNXZtf9_9WSFftehQN5m_xb7qGFEgHdpb1EuVXCQ@mail.gmail.com>

On Sat, Sep 27, 2014 at 9:11 PM, Anthony Carrico <acarrico at memebeam.org> wrote:
> On 09/27/2014 09:49 AM, Alexander D. Knauth wrote:
>> Like this, I think:  (but for some reason it seems to print it 3 times?)
>> #lang typed/racket
>>
>> (struct foo () #:transparent
>>   #:property prop:custom-write
>>   (lambda (this out mode)
>>     (displayln "whatever")))
>>
>> (print (foo))
>
> I didn't realize #:property was legal in typed struct. I guess there
> should be a documentation update here:
>   http://docs.racket-lang.org/ts-reference/special-forms.html

No, instead we need to change typed racket to reject those programs.
However, that would break a bunch of code with no easy fix that relies
on getting away with this, so we'll probably have to wait till we have
a real story here.

> BTW change your program to use the given port: (displayln "whatever"
> out) to avoid the double printing problem, although I'm not exactly sure
> why it would have that effect.

It prints multiple times (to hidden output ports) to detect cycles
among other things.

Sam

> Thanks!
>
> --
> Anthony Carrico
>
>
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>

From richter at math.northwestern.edu  Sun Sep 28 00:57:12 2014
From: richter at math.northwestern.edu (Bill Richter)
Date: Sat, 27 Sep 2014 23:57:12 -0500
Subject: [racket] proof assistants, DrRacket and Bootstrap
In-Reply-To: <54274040.5070902@uwaterloo.ca> (message from Prabhakar Ragde on
 Sat, 27 Sep 2014 18:54:56 -0400)
References: <5426CAAB.2010400@uwaterloo.ca>
 <201409272150.s8RLoBQW030663@poisson.math.northwestern.edu>
 <54274040.5070902@uwaterloo.ca>
Message-ID: <201409280457.s8S4vCrp000773@poisson.math.northwestern.edu>

   Translating the gist of suggestions to such posters, I would say:
   do you really need to execute *any* expression? If not, use
   OCamllex and/or OCamlyacc (maybe Menhir, or a smaller and cleaner
   parser combinator library) to write a parser for your language
   subset, and then write an interpreter for the resulting ASTs.

Thanks, Prabhakar!  I think something like what you're saying is true.  HOL Light is based on quite a lot of the OCaml parser camlp, and I've thought for a long time that I had to learn it myself.  

   I don't have enough practice in OCaml to grasp what your code does

I'm making a dialect ofHOL Light with different syntax by interpreting one of my programs as a string and then breaking the string up into the component pieces, but then I need this Toploop/exec hack to evaluate my variables.  That sounds like the Scheme eval, and in Scheme you can solve that problem nicely with the quote feature. 

-- 
Best,
Bill 

From matthias at ccs.neu.edu  Sat Sep 27 18:00:35 2014
From: matthias at ccs.neu.edu (Matthias Felleisen)
Date: Sat, 27 Sep 2014 16:00:35 -0600
Subject: [racket] proof assistants, DrRacket and Bootstrap
In-Reply-To: <201409270738.s8R7cmcd022184@poisson.math.northwestern.edu>
References: <201409270738.s8R7cmcd022184@poisson.math.northwestern.edu>
Message-ID: <B7FBF0FD-EA70-44BE-BB63-577E2E1B6CFF@ccs.neu.edu>


On Sep 27, 2014, at 1:38 AM, Bill Richter wrote:

> Scheme is well-suited for writing a Scheme interpreter, because of the quote feature.  OCaml doesn't have a quote feature, so the question arises how to write an OCaml interpreter inside OCaml.  That's not quite what I want to do, but if anyone could explain how to do it, I'd be grateful.

Decouple parsing from interpreting and your ML interpreter for ML won't look too different from your Lisp/Racket/Scheme interpreter for Lisp/Racket/Scheme. Indeed, almost everyone who does the latter actually does some form of decoupling usually. 

In general, 

 (1) pick a representation for your chosen programming language that is tree-oriented, call it AST
 (2) pick a representation for the values that your programs may return, call it VAL 
 (3) design a function 

	interpreter : AST -> VAL 

 (4) for a typed language, also design a function 
	
	type_check : AST -> Boolean 

In ML, you might try

  type ast = VAR of String | LAM of String * ast | APP of ast * ast | CONST of int | ADD of ast * ast 
  type val = BASIC of int | FUNCTION of val -> val | ...


> I have a 7th grade student I'm trying to teach Racket, and I'm a bit confused about Bootstrap and Program By Design.  Is there any reason for my student not to just read HtDP?


1. HtDP explains the above in a couple of sections. 

2. HtDP is for college freshmen (roughly) and for high school students who have the patience to read. The occasional 7th and 8th grader work their way thru HtDP with adults who understand HtDP or learn to understand HtDP that way. The basic 7th grader will benefit from the material at bootstrap-world.org, which is what we use with many middle school students around the country. 

-- Matthias







From richter at math.northwestern.edu  Sun Sep 28 01:39:20 2014
From: richter at math.northwestern.edu (Bill Richter)
Date: Sun, 28 Sep 2014 00:39:20 -0500
Subject: [racket] proof assistants, DrRacket and Bootstrap
In-Reply-To: <B7FBF0FD-EA70-44BE-BB63-577E2E1B6CFF@ccs.neu.edu> (message from
 Matthias Felleisen on Sat, 27 Sep 2014 16:00:35 -0600)
References: <201409270738.s8R7cmcd022184@poisson.math.northwestern.edu>
 <B7FBF0FD-EA70-44BE-BB63-577E2E1B6CFF@ccs.neu.edu>
Message-ID: <201409280539.s8S5dKcx001372@poisson.math.northwestern.edu>

     type ast = VAR of String | LAM of String * ast | APP of ast * ast | CONST of int | ADD of ast * ast 
     type val = BASIC of int | FUNCTION of val -> val | ...

Thanks, Matthias!  I know this is way off topic, but I think it's the sort of thing only Schemers understand, and I think HOL Light would be a lot better off my problem got solved.  There's a lot I didn't understand with your ML explanation.  Is this written down somewhere?  How to write a ML interpreter in ML using your type ideas?  Are you saying I don't need camlp parsing?  I would find that hard to believe.  Formal proofs in HOL Light are really just OCaml programs where a lot of code has already been evaluated.  Here's a simple Hilbert geometry program in my dialect:

let ExistsPointOffLine = theorem `;
  ?l. Line l  ?  ?Q. Q ? l

  proof
    intro_TAC ?l, H1;
    consider A B C such that
    ?(A = B) ? ?(A = C) ? ?(B = C) ? ?Collinear A B C     [Distinct] by fol I3;
    assume (A ? l) ? (B ? l) ? (C ? l)     [all_on] by fol ?;
    Collinear A B C     [] by fol H1 - Collinear_DEF;
    fol - Distinct;
  qed;
`;;

I think I need the `;[...]` construct, which at present just turns my proof into a string without any backslash or newline problems, which is then passed to my function theorem.  The usual HOL Light style is like this, using instead a function prove: 

let MULT_0 = prove
 (`!m. m * 0 = 0`,
  INDUCT_TAC THEN ASM_REWRITE_TAC[MULT; ADD_CLAUSES]);;

Notice the related `[...]` construct, which I think sends the ordered pair off to heavy camlp parsing.  

Thanks for the 7th and 8th advice.  I think my student is smart and hard-working, and I (finally!) learned how to program reading your book HtDP, so it sounds like you're saying that I should be able to teach it to my student.  I went to bootstrap-world.org and, uh, maybe I didn't look hard enough, but I didn't find anything that looked juicy.  

-- 
Best,
Bill 

From asumu at ccs.neu.edu  Sun Sep 28 11:17:15 2014
From: asumu at ccs.neu.edu (Asumu Takikawa)
Date: Sun, 28 Sep 2014 11:17:15 -0400
Subject: [racket] (fourth RacketCon) videos
Message-ID: <20140928151714.GH30398@localhost>

Hi all,

The hi-res talk recordings for (fourth RacketCon) are now up on Youtube.
You can watch all of them via this playlist:

  https://www.youtube.com/playlist?list=PLXr4KViVC0qI9t3lizitiFJ1cFIeN2Gdh

Let me know if you find any glitches.

Cheers,
Asumu

From plragde at uwaterloo.ca  Sun Sep 28 11:29:32 2014
From: plragde at uwaterloo.ca (Prabhakar Ragde)
Date: Sun, 28 Sep 2014 11:29:32 -0400
Subject: [racket] proof assistants, DrRacket and Bootstrap
In-Reply-To: <201409280457.s8S4vCrp000773@poisson.math.northwestern.edu>
References: <5426CAAB.2010400@uwaterloo.ca>
 <201409272150.s8RLoBQW030663@poisson.math.northwestern.edu>
 <54274040.5070902@uwaterloo.ca>
 <201409280457.s8S4vCrp000773@poisson.math.northwestern.edu>
Message-ID: <5428295C.5050006@uwaterloo.ca>

On 2014-09-28, 12:57 AM, Bill Richter wrote:

> I'm making a dialect of HOL Light with different syntax by
> interpreting one of my programs as a string and then breaking the
> string up into the component pieces, but then I need this
> Toploop/exec hack to evaluate my variables.  That sounds like the
> Scheme eval, and in Scheme you can solve that problem nicely with the
> quote feature.

A Racketeer would probably tell you to use macros to define your 
dialect, instead of using an eval hack. The corresponding tool for OCaml 
is camlp4, and it sounds as if it would be worth your time to learn it 
thoroughly. --PR

From samth at cs.indiana.edu  Sun Sep 28 11:48:12 2014
From: samth at cs.indiana.edu (Sam Tobin-Hochstadt)
Date: Sun, 28 Sep 2014 11:48:12 -0400
Subject: [racket] typed racket, filters, and polymorphism
In-Reply-To: <8CF041CE-65AD-488D-B2AE-FAE6C8C1C352@knauth.org>
References: <D276D7BD-D6AF-4F26-A79B-C1016641EF06@knauth.org>
 <CAK=HD+ZpPuvYTMZJ1_LXrnPGjcPamZ73Z8aiPizCussR+UxYsw@mail.gmail.com>
 <F16C1892-F09C-4A9F-94E2-16D5DE72F57D@knauth.org>
 <8CF041CE-65AD-488D-B2AE-FAE6C8C1C352@knauth.org>
Message-ID: <CAK=HD+avDfszLgz=1rZL0Jf1tz7gmnF_wadSF0oN_NRfAyJ7ng@mail.gmail.com>

Why not do this with the type, instead of making this polymorphic?

Sam

On Fri, Sep 26, 2014 at 7:35 PM, Alexander D. Knauth
<alexander at knauth.org> wrote:
> Is it possible to have a struct that does certain things according to the
> guard?
> #lang typed/racket
>
> (struct (a) foo ([a : a]) #:transparent
>   #:guard (lambda (a _)
>             (unless (exact-integer? a)
>               (error 'foo "expected Integer, given ~v" a))
>             a))
>
> (ann (foo (ann 1 Any)) (foo Integer))
>
> (: x : (foo Any))
> (define x (foo 1))
>
> (ann (foo-a x) Integer)
>
> ;. Type Checker: Polymorphic function `foo1' could not be applied to
> arguments:
> ;Argument 1:
> ;  Expected: a
> ;  Given:    Any
> ;
> ;Result type:     (foo a)
> ;Expected result: (foo Integer)
> ; in: (foo (ann 1 Any))
> ;. Type Checker: Polymorphic function `foo-a' could not be applied to
> arguments:
> ;Argument 1:
> ;  Expected: (foo a)
> ;  Given:    (foo Any)
> ;
> ;Result type:     (a : ....)
> ;Expected result: Integer
> ; in: (foo-a x)
>
> On Sep 25, 2014, at 9:42 PM, Alexander D. Knauth <alexander at knauth.org>
> wrote:
>
> What I?m trying to accomplish is something more like this:
> #lang typed/racket
>
> (require "dimensions.rkt")
>
> (struct (d) unit ([name : Any] [scalar : Positive-Real] [dimension : d])
> #:transparent
>   #:guard (lambda (name scalar dimension _)
>             (unless (dimension? dimension)
>               (error 'unit "expected Dimension, given ~v" dimension))
>             (values name scalar dimension)))
>
> (define-type (Unitof d) (unit d))
>
> (define-type Unit (Unitof Dimension))
>
> (define Unit? (make-predicate Unit))
>
> (define-type Unitish
>   (U (Unitof Any)
>      Dimension
>      Positive-Real))
>
> (: ->unit : (All (d) (case-> [(Unitof d) -> (Unitof d)]
>                              [Unitish -> Unit])))
> (define (->unit u)
>   (cond [(unit? u)
>          (unless (Unit? u) ; this should never happen anyway because of the
> guard
>            (error '->unit "expected (Unitof Dimension), given ~v" u))
>          u]
>         [(dimension? u) (unit u 1 u)]
>         [(positive-real? u) (unit u u dimensionless-dimension)]))
>
>
> On Sep 25, 2014, at 6:19 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu>
> wrote:
>
> No, I don't think you can do this. Can you say more about what you're
> trying to accomplish?
>
> Sam
>
> On Thu, Sep 25, 2014 at 6:15 PM, Alexander D. Knauth
> <alexander at knauth.org> wrote:
>
> Do any of you have any advice for getting a function like this to
> type-check?
> #lang typed/racket
>
> (: check-int : (All (a) (case-> [a -> a]
>                                [Any -> Integer])))
> (define (check-int int)
>  (unless (exact-integer? int)
>    (error 'check-int "expected Integer, given ~v" int))
>  int)
>
> ;. Type Checker: type mismatch
> ;  expected: a
> ;  given: Integer in: int
>
>
>
> ____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users
>
>
> ____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users
>
>


From alexander at knauth.org  Sun Sep 28 12:02:24 2014
From: alexander at knauth.org (Alexander D. Knauth)
Date: Sun, 28 Sep 2014 12:02:24 -0400
Subject: [racket] typed racket, filters, and polymorphism
In-Reply-To: <CAK=HD+avDfszLgz=1rZL0Jf1tz7gmnF_wadSF0oN_NRfAyJ7ng@mail.gmail.com>
References: <D276D7BD-D6AF-4F26-A79B-C1016641EF06@knauth.org>
 <CAK=HD+ZpPuvYTMZJ1_LXrnPGjcPamZ73Z8aiPizCussR+UxYsw@mail.gmail.com>
 <F16C1892-F09C-4A9F-94E2-16D5DE72F57D@knauth.org>
 <8CF041CE-65AD-488D-B2AE-FAE6C8C1C352@knauth.org>
 <CAK=HD+avDfszLgz=1rZL0Jf1tz7gmnF_wadSF0oN_NRfAyJ7ng@mail.gmail.com>
Message-ID: <1FB59E1F-0D4F-4276-BAB7-C6328866CE70@knauth.org>

Because the struct is representing a unit (kilograms, meters, seconds, etc.), and a unit has a dimension (mass, length, time, etc.) and I want the type-checker to be able to know what the dimension of a unit is so that the types of functions can specify the dimension that something should have.  
The real solution to this would probably be bounded polymorphism, but I was wondering if there was some other way to do it with occurrence typing in the guard or something like that.  

On Sep 28, 2014, at 11:48 AM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> wrote:

> Why not do this with the type, instead of making this polymorphic?
> 
> Sam
> 
> On Fri, Sep 26, 2014 at 7:35 PM, Alexander D. Knauth
> <alexander at knauth.org> wrote:
>> Is it possible to have a struct that does certain things according to the
>> guard?
>> #lang typed/racket
>> 
>> (struct (a) foo ([a : a]) #:transparent
>>  #:guard (lambda (a _)
>>            (unless (exact-integer? a)
>>              (error 'foo "expected Integer, given ~v" a))
>>            a))
>> 
>> (ann (foo (ann 1 Any)) (foo Integer))
>> 
>> (: x : (foo Any))
>> (define x (foo 1))
>> 
>> (ann (foo-a x) Integer)
>> 
>> ;. Type Checker: Polymorphic function `foo1' could not be applied to
>> arguments:
>> ;Argument 1:
>> ;  Expected: a
>> ;  Given:    Any
>> ;
>> ;Result type:     (foo a)
>> ;Expected result: (foo Integer)
>> ; in: (foo (ann 1 Any))
>> ;. Type Checker: Polymorphic function `foo-a' could not be applied to
>> arguments:
>> ;Argument 1:
>> ;  Expected: (foo a)
>> ;  Given:    (foo Any)
>> ;
>> ;Result type:     (a : ....)
>> ;Expected result: Integer
>> ; in: (foo-a x)
>> 
>> On Sep 25, 2014, at 9:42 PM, Alexander D. Knauth <alexander at knauth.org>
>> wrote:
>> 
>> What I?m trying to accomplish is something more like this:
>> #lang typed/racket
>> 
>> (require "dimensions.rkt")
>> 
>> (struct (d) unit ([name : Any] [scalar : Positive-Real] [dimension : d])
>> #:transparent
>>  #:guard (lambda (name scalar dimension _)
>>            (unless (dimension? dimension)
>>              (error 'unit "expected Dimension, given ~v" dimension))
>>            (values name scalar dimension)))
>> 
>> (define-type (Unitof d) (unit d))
>> 
>> (define-type Unit (Unitof Dimension))
>> 
>> (define Unit? (make-predicate Unit))
>> 
>> (define-type Unitish
>>  (U (Unitof Any)
>>     Dimension
>>     Positive-Real))
>> 
>> (: ->unit : (All (d) (case-> [(Unitof d) -> (Unitof d)]
>>                             [Unitish -> Unit])))
>> (define (->unit u)
>>  (cond [(unit? u)
>>         (unless (Unit? u) ; this should never happen anyway because of the
>> guard
>>           (error '->unit "expected (Unitof Dimension), given ~v" u))
>>         u]
>>        [(dimension? u) (unit u 1 u)]
>>        [(positive-real? u) (unit u u dimensionless-dimension)]))
>> 
>> 
>> On Sep 25, 2014, at 6:19 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu>
>> wrote:
>> 
>> No, I don't think you can do this. Can you say more about what you're
>> trying to accomplish?
>> 
>> Sam
>> 
>> On Thu, Sep 25, 2014 at 6:15 PM, Alexander D. Knauth
>> <alexander at knauth.org> wrote:
>> 
>> Do any of you have any advice for getting a function like this to
>> type-check?
>> #lang typed/racket
>> 
>> (: check-int : (All (a) (case-> [a -> a]
>>                               [Any -> Integer])))
>> (define (check-int int)
>> (unless (exact-integer? int)
>>   (error 'check-int "expected Integer, given ~v" int))
>> int)
>> 
>> ;. Type Checker: type mismatch
>> ;  expected: a
>> ;  given: Integer in: int
>> 
>> 
>> 
>> ____________________
>> Racket Users list:
>> http://lists.racket-lang.org/users
>> 
>> 
>> ____________________
>> Racket Users list:
>> http://lists.racket-lang.org/users
>> 
>> 



From samth at cs.indiana.edu  Sun Sep 28 12:08:43 2014
From: samth at cs.indiana.edu (Sam Tobin-Hochstadt)
Date: Sun, 28 Sep 2014 12:08:43 -0400
Subject: [racket] typed racket, filters, and polymorphism
In-Reply-To: <1FB59E1F-0D4F-4276-BAB7-C6328866CE70@knauth.org>
References: <D276D7BD-D6AF-4F26-A79B-C1016641EF06@knauth.org>
 <CAK=HD+ZpPuvYTMZJ1_LXrnPGjcPamZ73Z8aiPizCussR+UxYsw@mail.gmail.com>
 <F16C1892-F09C-4A9F-94E2-16D5DE72F57D@knauth.org>
 <8CF041CE-65AD-488D-B2AE-FAE6C8C1C352@knauth.org>
 <CAK=HD+avDfszLgz=1rZL0Jf1tz7gmnF_wadSF0oN_NRfAyJ7ng@mail.gmail.com>
 <1FB59E1F-0D4F-4276-BAB7-C6328866CE70@knauth.org>
Message-ID: <CAK=HD+Y16z72xD2jrxPMojS3jQqDs5ZDb9TCR1n5zby5dMHSew@mail.gmail.com>

No, I think we just need to add bounded polymorphism to TR.

Sam

On Sun, Sep 28, 2014 at 12:02 PM, Alexander D. Knauth
<alexander at knauth.org> wrote:
> Because the struct is representing a unit (kilograms, meters, seconds, etc.), and a unit has a dimension (mass, length, time, etc.) and I want the type-checker to be able to know what the dimension of a unit is so that the types of functions can specify the dimension that something should have.
> The real solution to this would probably be bounded polymorphism, but I was wondering if there was some other way to do it with occurrence typing in the guard or something like that.
>
> On Sep 28, 2014, at 11:48 AM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> wrote:
>
>> Why not do this with the type, instead of making this polymorphic?
>>
>> Sam
>>
>> On Fri, Sep 26, 2014 at 7:35 PM, Alexander D. Knauth
>> <alexander at knauth.org> wrote:
>>> Is it possible to have a struct that does certain things according to the
>>> guard?
>>> #lang typed/racket
>>>
>>> (struct (a) foo ([a : a]) #:transparent
>>>  #:guard (lambda (a _)
>>>            (unless (exact-integer? a)
>>>              (error 'foo "expected Integer, given ~v" a))
>>>            a))
>>>
>>> (ann (foo (ann 1 Any)) (foo Integer))
>>>
>>> (: x : (foo Any))
>>> (define x (foo 1))
>>>
>>> (ann (foo-a x) Integer)
>>>
>>> ;. Type Checker: Polymorphic function `foo1' could not be applied to
>>> arguments:
>>> ;Argument 1:
>>> ;  Expected: a
>>> ;  Given:    Any
>>> ;
>>> ;Result type:     (foo a)
>>> ;Expected result: (foo Integer)
>>> ; in: (foo (ann 1 Any))
>>> ;. Type Checker: Polymorphic function `foo-a' could not be applied to
>>> arguments:
>>> ;Argument 1:
>>> ;  Expected: (foo a)
>>> ;  Given:    (foo Any)
>>> ;
>>> ;Result type:     (a : ....)
>>> ;Expected result: Integer
>>> ; in: (foo-a x)
>>>
>>> On Sep 25, 2014, at 9:42 PM, Alexander D. Knauth <alexander at knauth.org>
>>> wrote:
>>>
>>> What I?m trying to accomplish is something more like this:
>>> #lang typed/racket
>>>
>>> (require "dimensions.rkt")
>>>
>>> (struct (d) unit ([name : Any] [scalar : Positive-Real] [dimension : d])
>>> #:transparent
>>>  #:guard (lambda (name scalar dimension _)
>>>            (unless (dimension? dimension)
>>>              (error 'unit "expected Dimension, given ~v" dimension))
>>>            (values name scalar dimension)))
>>>
>>> (define-type (Unitof d) (unit d))
>>>
>>> (define-type Unit (Unitof Dimension))
>>>
>>> (define Unit? (make-predicate Unit))
>>>
>>> (define-type Unitish
>>>  (U (Unitof Any)
>>>     Dimension
>>>     Positive-Real))
>>>
>>> (: ->unit : (All (d) (case-> [(Unitof d) -> (Unitof d)]
>>>                             [Unitish -> Unit])))
>>> (define (->unit u)
>>>  (cond [(unit? u)
>>>         (unless (Unit? u) ; this should never happen anyway because of the
>>> guard
>>>           (error '->unit "expected (Unitof Dimension), given ~v" u))
>>>         u]
>>>        [(dimension? u) (unit u 1 u)]
>>>        [(positive-real? u) (unit u u dimensionless-dimension)]))
>>>
>>>
>>> On Sep 25, 2014, at 6:19 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu>
>>> wrote:
>>>
>>> No, I don't think you can do this. Can you say more about what you're
>>> trying to accomplish?
>>>
>>> Sam
>>>
>>> On Thu, Sep 25, 2014 at 6:15 PM, Alexander D. Knauth
>>> <alexander at knauth.org> wrote:
>>>
>>> Do any of you have any advice for getting a function like this to
>>> type-check?
>>> #lang typed/racket
>>>
>>> (: check-int : (All (a) (case-> [a -> a]
>>>                               [Any -> Integer])))
>>> (define (check-int int)
>>> (unless (exact-integer? int)
>>>   (error 'check-int "expected Integer, given ~v" int))
>>> int)
>>>
>>> ;. Type Checker: type mismatch
>>> ;  expected: a
>>> ;  given: Integer in: int
>>>
>>>
>>>
>>> ____________________
>>> Racket Users list:
>>> http://lists.racket-lang.org/users
>>>
>>>
>>> ____________________
>>> Racket Users list:
>>> http://lists.racket-lang.org/users
>>>
>>>
>


From spencerflorence at gmail.com  Sun Sep 28 12:13:22 2014
From: spencerflorence at gmail.com (Spencer florence)
Date: Sun, 28 Sep 2014 09:13:22 -0700 (PDT)
Subject: [racket] typed racket, filters, and polymorphism
In-Reply-To: <1FB59E1F-0D4F-4276-BAB7-C6328866CE70@knauth.org>
References: <1FB59E1F-0D4F-4276-BAB7-C6328866CE70@knauth.org>
Message-ID: <1411920801296.55e3aa5@Nodemailer>

would something like this work?





#lang typed/racket




(struct (U) number+unit ([amount : Real] [unit : U]))




(define-type Weight-Unit (U 'kg 'g 'mg '?g))

(define-type Weight (number+unit Weight-Unit))

(define-predicate weight? Weight)




(: make-weight : Real Weight-Unit -> Weight)

(define (make-weight n u)

? (number+unit n u))




(: +/weight : Weight Weight -> Weight)

;; something something needs unit conversion

(define (+/weight w1 w2)

? (number+unit (+ (number+unit-amount w1)

? ? ? ? ? ? ? ? ? (number+unit-amount w1))

?? ? ? ? ? ? ? (number+unit-unit w1)))




(+/weight (make-weight 1 'kg) (make-weight 1 'kg))

On Sun, Sep 28, 2014 at 11:03 AM, Alexander D. Knauth
<alexander at knauth.org> wrote:

> Because the struct is representing a unit (kilograms, meters, seconds, etc.), and a unit has a dimension (mass, length, time, etc.) and I want the type-checker to be able to know what the dimension of a unit is so that the types of functions can specify the dimension that something should have.  
> The real solution to this would probably be bounded polymorphism, but I was wondering if there was some other way to do it with occurrence typing in the guard or something like that.  
> On Sep 28, 2014, at 11:48 AM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> wrote:
>> Why not do this with the type, instead of making this polymorphic?
>> 
>> Sam
>> 
>> On Fri, Sep 26, 2014 at 7:35 PM, Alexander D. Knauth
>> <alexander at knauth.org> wrote:
>>> Is it possible to have a struct that does certain things according to the
>>> guard?
>>> #lang typed/racket
>>> 
>>> (struct (a) foo ([a : a]) #:transparent
>>>  #:guard (lambda (a _)
>>>            (unless (exact-integer? a)
>>>              (error 'foo "expected Integer, given ~v" a))
>>>            a))
>>> 
>>> (ann (foo (ann 1 Any)) (foo Integer))
>>> 
>>> (: x : (foo Any))
>>> (define x (foo 1))
>>> 
>>> (ann (foo-a x) Integer)
>>> 
>>> ;. Type Checker: Polymorphic function `foo1' could not be applied to
>>> arguments:
>>> ;Argument 1:
>>> ;  Expected: a
>>> ;  Given:    Any
>>> ;
>>> ;Result type:     (foo a)
>>> ;Expected result: (foo Integer)
>>> ; in: (foo (ann 1 Any))
>>> ;. Type Checker: Polymorphic function `foo-a' could not be applied to
>>> arguments:
>>> ;Argument 1:
>>> ;  Expected: (foo a)
>>> ;  Given:    (foo Any)
>>> ;
>>> ;Result type:     (a : ....)
>>> ;Expected result: Integer
>>> ; in: (foo-a x)
>>> 
>>> On Sep 25, 2014, at 9:42 PM, Alexander D. Knauth <alexander at knauth.org>
>>> wrote:
>>> 
>>> What I?m trying to accomplish is something more like this:
>>> #lang typed/racket
>>> 
>>> (require "dimensions.rkt")
>>> 
>>> (struct (d) unit ([name : Any] [scalar : Positive-Real] [dimension : d])
>>> #:transparent
>>>  #:guard (lambda (name scalar dimension _)
>>>            (unless (dimension? dimension)
>>>              (error 'unit "expected Dimension, given ~v" dimension))
>>>            (values name scalar dimension)))
>>> 
>>> (define-type (Unitof d) (unit d))
>>> 
>>> (define-type Unit (Unitof Dimension))
>>> 
>>> (define Unit? (make-predicate Unit))
>>> 
>>> (define-type Unitish
>>>  (U (Unitof Any)
>>>     Dimension
>>>     Positive-Real))
>>> 
>>> (: ->unit : (All (d) (case-> [(Unitof d) -> (Unitof d)]
>>>                             [Unitish -> Unit])))
>>> (define (->unit u)
>>>  (cond [(unit? u)
>>>         (unless (Unit? u) ; this should never happen anyway because of the
>>> guard
>>>           (error '->unit "expected (Unitof Dimension), given ~v" u))
>>>         u]
>>>        [(dimension? u) (unit u 1 u)]
>>>        [(positive-real? u) (unit u u dimensionless-dimension)]))
>>> 
>>> 
>>> On Sep 25, 2014, at 6:19 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu>
>>> wrote:
>>> 
>>> No, I don't think you can do this. Can you say more about what you're
>>> trying to accomplish?
>>> 
>>> Sam
>>> 
>>> On Thu, Sep 25, 2014 at 6:15 PM, Alexander D. Knauth
>>> <alexander at knauth.org> wrote:
>>> 
>>> Do any of you have any advice for getting a function like this to
>>> type-check?
>>> #lang typed/racket
>>> 
>>> (: check-int : (All (a) (case-> [a -> a]
>>>                               [Any -> Integer])))
>>> (define (check-int int)
>>> (unless (exact-integer? int)
>>>   (error 'check-int "expected Integer, given ~v" int))
>>> int)
>>> 
>>> ;. Type Checker: type mismatch
>>> ;  expected: a
>>> ;  given: Integer in: int
>>> 
>>> 
>>> 
>>> ____________________
>>> Racket Users list:
>>> http://lists.racket-lang.org/users
>>> 
>>> 
>>> ____________________
>>> Racket Users list:
>>> http://lists.racket-lang.org/users
>>> 
>>> 
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140928/784e05a1/attachment-0001.html>

From alexander at knauth.org  Sun Sep 28 13:37:30 2014
From: alexander at knauth.org (Alexander D. Knauth)
Date: Sun, 28 Sep 2014 13:37:30 -0400
Subject: [racket] typed racket, filters, and polymorphism
In-Reply-To: <1411920801296.55e3aa5@Nodemailer>
References: <1FB59E1F-0D4F-4276-BAB7-C6328866CE70@knauth.org>
 <1411920801296.55e3aa5@Nodemailer>
Message-ID: <B849BCC0-7F5C-402E-B98C-01F97E1762D8@knauth.org>

No because I want the unit to be a struct that has a dimension field, not a symbol with various dimensions defined as unions of units.
I want the unit to be based on the dimension, not the other way around, so that new units can be made that have the same dimension.

I have something like the number+unit struct (I called it measure), but I?ll work on more that after I have the unit struct figured out.  

On Sep 28, 2014, at 12:13 PM, Spencer florence <spencerflorence at gmail.com> wrote:

> would something like this work?
> 
> #lang typed/racket
> 
> (struct (U) number+unit ([amount : Real] [unit : U]))
> 
> (define-type Weight-Unit (U 'kg 'g 'mg '?g))
> (define-type Weight (number+unit Weight-Unit))
> (define-predicate weight? Weight)
> 
> (: make-weight : Real Weight-Unit -> Weight)
> (define (make-weight n u)
>   (number+unit n u))
> 
> (: +/weight : Weight Weight -> Weight)
> ;; something something needs unit conversion
> (define (+/weight w1 w2)
>   (number+unit (+ (number+unit-amount w1)
>                   (number+unit-amount w1))
>                (number+unit-unit w1)))
> 
> (+/weight (make-weight 1 'kg) (make-weight 1 'kg))
> 
> 
> 
> On Sun, Sep 28, 2014 at 11:03 AM, Alexander D. Knauth <alexander at knauth.org> wrote:
> 
> Because the struct is representing a unit (kilograms, meters, seconds, etc.), and a unit has a dimension (mass, length, time, etc.) and I want the type-checker to be able to know what the dimension of a unit is so that the types of functions can specify the dimension that something should have. 
> The real solution to this would probably be bounded polymorphism, but I was wondering if there was some other way to do it with occurrence typing in the guard or something like that. 
> 
> On Sep 28, 2014, at 11:48 AM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> wrote: 
> 
> > Why not do this with the type, instead of making this polymorphic? 
> > 
> > Sam 
> > 
> > On Fri, Sep 26, 2014 at 7:35 PM, Alexander D. Knauth 
> > <alexander at knauth.org> wrote: 
> >> Is it possible to have a struct that does certain things according to the 
> >> guard? 
> >> #lang typed/racket 
> >> 
> >> (struct (a) foo ([a : a]) #:transparent 
> >> #:guard (lambda (a _) 
> >> (unless (exact-integer? a) 
> >> (error 'foo "expected Integer, given ~v" a)) 
> >> a)) 
> >> 
> >> (ann (foo (ann 1 Any)) (foo Integer)) 
> >> 
> >> (: x : (foo Any)) 
> >> (define x (foo 1)) 
> >> 
> >> (ann (foo-a x) Integer) 
> >> 
> >> ;. Type Checker: Polymorphic function `foo1' could not be applied to 
> >> arguments: 
> >> ;Argument 1: 
> >> ; Expected: a 
> >> ; Given: Any 
> >> ; 
> >> ;Result type: (foo a) 
> >> ;Expected result: (foo Integer) 
> >> ; in: (foo (ann 1 Any)) 
> >> ;. Type Checker: Polymorphic function `foo-a' could not be applied to 
> >> arguments: 
> >> ;Argument 1: 
> >> ; Expected: (foo a) 
> >> ; Given: (foo Any) 
> >> ; 
> >> ;Result type: (a : ....) 
> >> ;Expected result: Integer 
> >> ; in: (foo-a x) 
> >> 
> >> On Sep 25, 2014, at 9:42 PM, Alexander D. Knauth <alexander at knauth.org> 
> >> wrote: 
> >> 
> >> What I?m trying to accomplish is something more like this: 
> >> #lang typed/racket 
> >> 
> >> (require "dimensions.rkt") 
> >> 
> >> (struct (d) unit ([name : Any] [scalar : Positive-Real] [dimension : d]) 
> >> #:transparent 
> >> #:guard (lambda (name scalar dimension _) 
> >> (unless (dimension? dimension) 
> >> (error 'unit "expected Dimension, given ~v" dimension)) 
> >> (values name scalar dimension))) 
> >> 
> >> (define-type (Unitof d) (unit d)) 
> >> 
> >> (define-type Unit (Unitof Dimension)) 
> >> 
> >> (define Unit? (make-predicate Unit)) 
> >> 
> >> (define-type Unitish 
> >> (U (Unitof Any) 
> >> Dimension 
> >> Positive-Real)) 
> >> 
> >> (: ->unit : (All (d) (case-> [(Unitof d) -> (Unitof d)] 
> >> [Unitish -> Unit]))) 
> >> (define (->unit u) 
> >> (cond [(unit? u) 
> >> (unless (Unit? u) ; this should never happen anyway because of the 
> >> guard 
> >> (error '->unit "expected (Unitof Dimension), given ~v" u)) 
> >> u] 
> >> [(dimension? u) (unit u 1 u)] 
> >> [(positive-real? u) (unit u u dimensionless-dimension)])) 
> >> 
> >> 
> >> On Sep 25, 2014, at 6:19 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> 
> >> wrote: 
> >> 
> >> No, I don't think you can do this. Can you say more about what you're 
> >> trying to accomplish? 
> >> 
> >> Sam 
> >> 
> >> On Thu, Sep 25, 2014 at 6:15 PM, Alexander D. Knauth 
> >> <alexander at knauth.org> wrote: 
> >> 
> >> Do any of you have any advice for getting a function like this to 
> >> type-check? 
> >> #lang typed/racket 
> >> 
> >> (: check-int : (All (a) (case-> [a -> a] 
> >> [Any -> Integer]))) 
> >> (define (check-int int) 
> >> (unless (exact-integer? int) 
> >> (error 'check-int "expected Integer, given ~v" int)) 
> >> int) 
> >> 
> >> ;. Type Checker: type mismatch 
> >> ; expected: a 
> >> ; given: Integer in: int 
> >> 
> >> 
> >> 
> >> ____________________ 
> >> Racket Users list: 
> >> http://lists.racket-lang.org/users 
> >> 
> >> 
> >> ____________________ 
> >> Racket Users list: 
> >> http://lists.racket-lang.org/users 
> >> 
> >> 
> 
> 
> ____________________ 
> Racket Users list: 
> http://lists.racket-lang.org/users 
> 
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140928/56742b6b/attachment.html>

From eric.n.dobson at gmail.com  Sun Sep 28 13:48:06 2014
From: eric.n.dobson at gmail.com (Eric Dobson)
Date: Sun, 28 Sep 2014 10:48:06 -0700
Subject: [racket] Help debugging a ffi crash
Message-ID: <CAAEHQ5uRU5V4i1rvNf6sdfcbQsD087ntCwftjRVhKroNErmUXg@mail.gmail.com>

I'm trying to debug an FFI crash that I'm seeing, and because it is
dealing with C code the error just presents as a segfault. I believe I
have tracked down what is causing the problem, but don't understand
how it could be doing so.

I have two racket functions which take a "cursor" (the foreign
libraries object) and return a string representation of it, which I'm
trying to use for debugging.

(define raw-clang-get-cstring
  (get-ffi-obj "clang_getCString" lib-clang
    (_fun _pointer -> _string)))

(define raw-cursor-spelling
 (get-ffi-obj "clang_getCursorSpelling" lib-clang
   (_fun _CXCursor -> _pointer)))

(define (cursor-spelling c)
  (raw-clang-get-cstring (raw-cursor-spelling c)))

(define cursor-spelling2
 (get-ffi-obj "clang_getCursorSpelling" lib-clang
   (_fun _CXCursor -> (make-ctype _pointer values (? (v)
(raw-clang-get-cstring v))))))

If I use cursor-spelling, I have not been able to trigger a crash. But
if I use cursor-spelling2 I can reliably trigger a crash.

Is there anything obvious on how these functions are different?
Because they look to me like they should be doing the same thing. If
it would be helpful I can try to get my code in a portable enough
shape so that it will work/crash on another machine.


From adrian.klaver at aklaver.com  Thu Sep 25 20:25:15 2014
From: adrian.klaver at aklaver.com (Adrian Klaver)
Date: Thu, 25 Sep 2014 17:25:15 -0700
Subject: [racket] [GENERAL] Off Topic: Anybody reading this via
	news.gmane.org?
In-Reply-To: <5424AFC2.6020302@comcast.net>
References: <CAJYbDam0maAG7ZmCBjbha1uqhJyCpX9J+L3=kxqAL4GjdKUvkg@mail.gmail.com>
 <54248425.90509@comcast.net> <5424A05A.9020809@aklaver.com>
 <5424AFC2.6020302@comcast.net>
Message-ID: <5424B26B.9080504@aklaver.com>

On 09/25/2014 05:13 PM, George Neuner wrote:
>
> On 9/25/2014 5:26 PM, Neil Van Dyke wrote:
>> You can check your IP addr against the list at
>> "http://gmane.org/denied.php".
>>
>> Neil V.
>
> On 9/25/2014 7:08 PM, Adrian Klaver wrote:
>>
>> Take a look here:
>>
>> http://gmane.org/denied.php
>>
>> My guess is you are the fourth one from the bottom.
>>
>>
>> Might want to take a look at this thread to see what your options are
>> and what the turn around time is on your request:
>>
>> http://thread.gmane.org/gmane.discuss/16309
>
> Thanks to both of you - I missed seeing the entry about the denied list
> in the FAQ.
>
> Adrian you were right - it seems that I have been blocked for some
> reason.  Based on the description of infractions I really don't think I
> did anything to warrant it ... but there it is.  I'll have to follow my
> lists by email until I can unblocked.

I think the relevant part at gmane.org/denied.php is:

"Since there is no way to identify users, news readers are denied on a 
domain/IP basis. So if somebody on a machine close to you downloaded the 
news spool yesterday, you're likely to be denied today"

>
> Thanks again,
> George


-- 
Adrian Klaver
adrian.klaver at aklaver.com

From mflatt at cs.utah.edu  Sun Sep 28 14:08:23 2014
From: mflatt at cs.utah.edu (Matthew Flatt)
Date: Sun, 28 Sep 2014 12:08:23 -0600
Subject: [racket] Help debugging a ffi crash
In-Reply-To: <CAAEHQ5uRU5V4i1rvNf6sdfcbQsD087ntCwftjRVhKroNErmUXg@mail.gmail.com>
References: <CAAEHQ5uRU5V4i1rvNf6sdfcbQsD087ntCwftjRVhKroNErmUXg@mail.gmail.com>
Message-ID: <20140928180825.80D74650196@mail-svr1.cs.utah.edu>

Looking at 

  http://clang.llvm.org/doxygen/CXString_8h_source.html

it seems that CXString as returned by clang_getCursorSpelling() is not
a pointer:

 typedef struct {
   const void *data;
   unsigned private_flags;
 } CXString;

If that's right, I'm a little surprised that `cursor-spelling` works
--- but when you get representation wrong, strange things can happen,
including something working when it shouldn't.

Am I looking at the right library/definitions?

At Sun, 28 Sep 2014 10:48:06 -0700, Eric Dobson wrote:
> I'm trying to debug an FFI crash that I'm seeing, and because it is
> dealing with C code the error just presents as a segfault. I believe I
> have tracked down what is causing the problem, but don't understand
> how it could be doing so.
> 
> I have two racket functions which take a "cursor" (the foreign
> libraries object) and return a string representation of it, which I'm
> trying to use for debugging.
> 
> (define raw-clang-get-cstring
>   (get-ffi-obj "clang_getCString" lib-clang
>     (_fun _pointer -> _string)))
> 
> (define raw-cursor-spelling
>  (get-ffi-obj "clang_getCursorSpelling" lib-clang
>    (_fun _CXCursor -> _pointer)))
> 
> (define (cursor-spelling c)
>   (raw-clang-get-cstring (raw-cursor-spelling c)))
> 
> (define cursor-spelling2
>  (get-ffi-obj "clang_getCursorSpelling" lib-clang
>    (_fun _CXCursor -> (make-ctype _pointer values (? (v)
> (raw-clang-get-cstring v))))))
> 
> If I use cursor-spelling, I have not been able to trigger a crash. But
> if I use cursor-spelling2 I can reliably trigger a crash.
> 
> Is there anything obvious on how these functions are different?
> Because they look to me like they should be doing the same thing. If
> it would be helpful I can try to get my code in a portable enough
> shape so that it will work/crash on another machine.
> 
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users


From eric.n.dobson at gmail.com  Sun Sep 28 14:15:33 2014
From: eric.n.dobson at gmail.com (Eric Dobson)
Date: Sun, 28 Sep 2014 11:15:33 -0700
Subject: [racket] Help debugging a ffi crash
In-Reply-To: <20140928180825.80D74650196@mail-svr1.cs.utah.edu>
References: <CAAEHQ5uRU5V4i1rvNf6sdfcbQsD087ntCwftjRVhKroNErmUXg@mail.gmail.com>
 <20140928180825.80D74650196@mail-svr1.cs.utah.edu>
Message-ID: <CAAEHQ5tWAKzi5AizADUH9smVsNXpYTbmxtO3teGm9Zj8ZDUnLg@mail.gmail.com>

That is almost surely it, thanks for the second pair of eyes. I had
similar issues with a different type that I thought was a pointer but
was actually a struct, but that one I couldn't even get a single call
to work so it was much more obvious something was up.

On Sun, Sep 28, 2014 at 11:08 AM, Matthew Flatt <mflatt at cs.utah.edu> wrote:
> Looking at
>
>   http://clang.llvm.org/doxygen/CXString_8h_source.html
>
> it seems that CXString as returned by clang_getCursorSpelling() is not
> a pointer:
>
>  typedef struct {
>    const void *data;
>    unsigned private_flags;
>  } CXString;
>
> If that's right, I'm a little surprised that `cursor-spelling` works
> --- but when you get representation wrong, strange things can happen,
> including something working when it shouldn't.
>
> Am I looking at the right library/definitions?
>
> At Sun, 28 Sep 2014 10:48:06 -0700, Eric Dobson wrote:
>> I'm trying to debug an FFI crash that I'm seeing, and because it is
>> dealing with C code the error just presents as a segfault. I believe I
>> have tracked down what is causing the problem, but don't understand
>> how it could be doing so.
>>
>> I have two racket functions which take a "cursor" (the foreign
>> libraries object) and return a string representation of it, which I'm
>> trying to use for debugging.
>>
>> (define raw-clang-get-cstring
>>   (get-ffi-obj "clang_getCString" lib-clang
>>     (_fun _pointer -> _string)))
>>
>> (define raw-cursor-spelling
>>  (get-ffi-obj "clang_getCursorSpelling" lib-clang
>>    (_fun _CXCursor -> _pointer)))
>>
>> (define (cursor-spelling c)
>>   (raw-clang-get-cstring (raw-cursor-spelling c)))
>>
>> (define cursor-spelling2
>>  (get-ffi-obj "clang_getCursorSpelling" lib-clang
>>    (_fun _CXCursor -> (make-ctype _pointer values (? (v)
>> (raw-clang-get-cstring v))))))
>>
>> If I use cursor-spelling, I have not been able to trigger a crash. But
>> if I use cursor-spelling2 I can reliably trigger a crash.
>>
>> Is there anything obvious on how these functions are different?
>> Because they look to me like they should be doing the same thing. If
>> it would be helpful I can try to get my code in a portable enough
>> shape so that it will work/crash on another machine.
>>
>> ____________________
>>   Racket Users list:
>>   http://lists.racket-lang.org/users


From kalimehtar at mail.ru  Sun Sep 28 14:39:54 2014
From: kalimehtar at mail.ru (=?UTF-8?B?Um9tYW4gS2xvY2hrb3Y=?=)
Date: Sun, 28 Sep 2014 22:39:54 +0400
Subject: [racket] =?utf-8?q?How_to_document_a_field=3F?=
Message-ID: <1411929594.923492574@f67.i.mail.ru>

 At  http://docs.racket-lang.org/scribble/doc-classes.html I see classes, interfaces, methods, but don't see fields.

How to document them? Via @defthing or is there some syntax with (get-field ..) (set-field! ...) template like @defparam ?


-- 
Roman Klochkov
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140928/ac478807/attachment.html>

From richter at math.northwestern.edu  Sun Sep 28 14:51:26 2014
From: richter at math.northwestern.edu (Bill Richter)
Date: Sun, 28 Sep 2014 13:51:26 -0500
Subject: [racket] proof assistants, DrRacket and Bootstrap
In-Reply-To: <B7FBF0FD-EA70-44BE-BB63-577E2E1B6CFF@ccs.neu.edu> (message from
 Matthias Felleisen on Sat, 27 Sep 2014 16:00:35 -0600)
References: <201409270738.s8R7cmcd022184@poisson.math.northwestern.edu>
 <B7FBF0FD-EA70-44BE-BB63-577E2E1B6CFF@ccs.neu.edu>
Message-ID: <201409281851.s8SIpQMx006764@poisson.math.northwestern.edu>

Matthias & Prabhakar, I think I see what you're saying now.  I need the camlp parser to turn my dialect-expressions into trees.  Then I turn my  tree into standard expressions just as one does in Scheme.  The only differences is that in Scheme we have the quote function that skips the parsing step.  So your ML code

MF>   type ast = VAR of String | LAM of String * ast | APP of ast * ast | CONST of int | ADD of ast * ast 
MF>   type val = BASIC of int | FUNCTION of val -> val | ...

looks a lot like the HOL Light definition of preterms: 

type preterm = Varp of string * pretype       (* Variable           - v      *)
             | Constp of string * pretype     (* Constant           - c      *)
             | Combp of preterm * preterm     (* Combination        - f x    *)
             | Absp of preterm * preterm      (* Lambda-abstraction - \x. t  *)
             | Typing of preterm * pretype;;  (* Type constraint    - t : ty *)

and so we can see one of these preterms in a standard term: 

# preterm_of_term `x + y`;;
val it : preterm =
  Combp
   (Combp
     (Constp ("+",
       Ptycon ("fun",
        [Ptycon ("num", []);
         Ptycon ("fun", [Ptycon ("num", []); Ptycon ("num", [])])])),
     Varp ("x", Ptycon ("num", []))),
   Varp ("y", Ptycon ("num", [])))

So the Scheme story works fine except that I haven't understood camlp or how HOL Light turns expressions `[...]` into preterms.  But I feel better about investing the time into learning camlp now that I see the Scheme story remains intact in HOL Light. And you're right  about my ignorance:

MF> 1. HtDP explains the above in a couple of sections. 

Right, there's something I seem not to have understood in either Scheme or ML, how when we create and evaluate our trees we retain the values of our variables.  I can figure that out, and HtDP sounds like a good place to start. 

PR> A Racketeer would probably tell you to use macros to define your
PR> dialect, instead of using an eval hack. The corresponding tool for
PR> OCaml is camlp4, and it sounds as if it would be worth your time
PR> to learn it thoroughly.

That's an interesting comparison, and since macros, maybe I can learn to like camlp.  In fact, maybe I ought to learn something about racket macros first. 

-- 
Best,
Bill 

From greghendershott at gmail.com  Sun Sep 28 16:01:20 2014
From: greghendershott at gmail.com (Greg Hendershott)
Date: Sun, 28 Sep 2014 16:01:20 -0400
Subject: [racket] aws/glacier: credential scope
In-Reply-To: <0CA916BA-DD1E-47E9-BFA1-7092EC3F17F5@astro.gla.ac.uk>
References: <0CA916BA-DD1E-47E9-BFA1-7092EC3F17F5@astro.gla.ac.uk>
Message-ID: <CAGspUn0tqgSVUaDuk1C3Wb6hm6s6nQ2ps-j42pkJJSY2UESL0A@mail.gmail.com>

Hi, Norman.

I logged an issue for this:

  https://github.com/greghendershott/aws/issues/32

I see the problem (or at least the main problem) and will push a fix.

The bug is embarrassing, not just because it's such a silly mistake,
but it's something a unit test could have caught. (I could say that
doing unit tests for Glacier is challenging, because the retrieval
process can take hours. Although that's true, I could have done more
to test _some_ operations working among various regions.)

On Sat, Sep 27, 2014 at 4:14 PM, Norman Gray <norman at astro.gla.ac.uk> wrote:
>
> Greetings.
>
> I'm trying to use the aws/glacier package, and running into a problem where I'm being told:
>
>     Credential should be scoped to a valid region, not 'eu-west-1'
>
> I'm following the instructions at <https://github.com/greghendershott/aws/blob/master/aws/manual.md>
>
> My test code is:
>
>     % cat glacier.rkt
>     #lang racket/base
>
>     (require aws/glacier
>              aws/keys)
>
>     (define vault "testvault")
>     (region "eu-west-1")
>     (read-keys "aws-zbu-credentials") ; local file
>
>     (module+ main
>       (printf "region=~a~%" (region))
>       (printf "Vaults: ~s~%" (list-vaults))
>       (printf "...specifically: ~s~%" (describe-vault vault)))
>
> Running this produces:
>
>     % racket glacier.rkt
>     region=eu-west-1
>     aws: HTTP/1.1 403 Forbidden
>     x-amzn-RequestId: Un3-L2zlaJBPyrIVKJrWuQcqtMMYQAr34gYUOSScg6Qepc4
>     Content-Type: application/json
>     Content-Length: 129
>     Date: Sat, 27 Sep 2014 18:35:50 GMT
>
>      {"message":"Credential should be scoped to a valid region, not 'eu-west-1'. ","code":"InvalidSignatureException","type":"Client"}
>     HTTP 403 "Forbidden". AWS Code="InvalidSignatureException" Message="Credential should be scoped to a valid region, not 'eu-west-1'. "
>       context...:
>        check-response
>        /Users/norman/Library/Racket/6.1/pkgs/aws/aws/glacier.rkt:97:22: temp68
>        request/redirect/uri
>        (submod /checkouts/me/code/zbu/glacier.rkt main): [running body]
>
> Things I thought of:
>
>   * Printing (public-key)/(private-key) indicates that the credentials are being read correctly.
>   * When I change the argument of (region) to "us-west-1", that's the region that appears in the error message.
>   * My "testvault" vault is in eu-west-1 (and this is indeed one of the valid regions for glacier, reported in <http://docs.aws.amazon.com/general/latest/gr/rande.html> and which does have a host at http://glacier.eu-west-1.amazonaws.com
>   * As far as I can see, credentials are _not_ scoped, but are all at us-east-1.
>   * <http://docs.aws.amazon.com/general/latest/gr/signature-v4-troubleshooting.html> says that "IAM [...] accepts only us-east-1 as its region specification", so I'm taking it that (region) is for setting the _vault_'s region.
>   * I'm not a great AWS expert, so I could have something in my setup broken; but if so, I've no clue what.
>
> If, however, I change the (region) argument to "us-east-1", I get a different error message "User: arn:aws:iam::786725553169:user/zbu is not authorized to perform: glacier:ListVaults on resource: arn:aws:glacier:us-east-1:786725553169:vaults/"  That makes sense, since there's no such vault, but it's interesting that it gets _further_ when the (region) matches the region for the IAM service.
>
> I don't see any other (region) equivalents for the other services supported by the package.  Is that because all of the other services supported by the package are supported by all the AWS regions, or am I missing a configuration?
>
> Thanks for any pointers.
>
> All the best,
>
> Norman
>
>
> --
> Norman Gray  :  http://nxg.me.uk
> SUPA School of Physics and Astronomy, University of Glasgow, UK
>
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users


From norman at astro.gla.ac.uk  Sun Sep 28 16:29:03 2014
From: norman at astro.gla.ac.uk (Norman Gray)
Date: Sun, 28 Sep 2014 21:29:03 +0100
Subject: [racket] aws/glacier: credential scope
In-Reply-To: <CAGspUn0tqgSVUaDuk1C3Wb6hm6s6nQ2ps-j42pkJJSY2UESL0A@mail.gmail.com>
References: <0CA916BA-DD1E-47E9-BFA1-7092EC3F17F5@astro.gla.ac.uk>
 <CAGspUn0tqgSVUaDuk1C3Wb6hm6s6nQ2ps-j42pkJJSY2UESL0A@mail.gmail.com>
Message-ID: <FE320180-07AE-43E8-AE51-3AA566329E67@astro.gla.ac.uk>


Greg, hello.

On 2014 Sep 28, at 21:01, Greg Hendershott <greghendershott at gmail.com> wrote:

> I logged an issue for this:
> 
>  https://github.com/greghendershott/aws/issues/32

Ah: I wondered if the problem might be related to that.

> The bug is embarrassing, not just because it's such a silly mistake,
> but it's something a unit test could have caught. (I could say that
> doing unit tests for Glacier is challenging, because the retrieval
> process can take hours.

I can imagine how protracted that would be ... *shudder*.

Thanks for looking at this.

Also, Frank:

> What happens if you read-keys before setting the region?

Thanks for this suggestion: I did try that just a little earlier, but it didn't make any difference. 

All the best,

Norman


-- 
Norman Gray  :  http://nxg.me.uk
SUPA School of Physics and Astronomy, University of Glasgow, UK



From greghendershott at gmail.com  Sun Sep 28 16:29:12 2014
From: greghendershott at gmail.com (Greg Hendershott)
Date: Sun, 28 Sep 2014 16:29:12 -0400
Subject: [racket] aws/glacier: credential scope
In-Reply-To: <CAGspUn0tqgSVUaDuk1C3Wb6hm6s6nQ2ps-j42pkJJSY2UESL0A@mail.gmail.com>
References: <0CA916BA-DD1E-47E9-BFA1-7092EC3F17F5@astro.gla.ac.uk>
 <CAGspUn0tqgSVUaDuk1C3Wb6hm6s6nQ2ps-j42pkJJSY2UESL0A@mail.gmail.com>
Message-ID: <CAGspUn2au5xydsH+o9E2Y3JTA1iQbPRX58zRcitywAPWh5Djng@mail.gmail.com>

I pushed a fix, including running the tests across a few different AWS
regions. They pass.

I clicked "update" on pkgs.racket-lang.org. But it seems to be slower
than usual to refresh. After it does, you can `raco pkg update aws` to
get the fix.

On Sun, Sep 28, 2014 at 4:01 PM, Greg Hendershott
<greghendershott at gmail.com> wrote:
> Hi, Norman.
>
> I logged an issue for this:
>
>   https://github.com/greghendershott/aws/issues/32
>
> I see the problem (or at least the main problem) and will push a fix.
>
> The bug is embarrassing, not just because it's such a silly mistake,
> but it's something a unit test could have caught. (I could say that
> doing unit tests for Glacier is challenging, because the retrieval
> process can take hours. Although that's true, I could have done more
> to test _some_ operations working among various regions.)
>
> On Sat, Sep 27, 2014 at 4:14 PM, Norman Gray <norman at astro.gla.ac.uk> wrote:
>>
>> Greetings.
>>
>> I'm trying to use the aws/glacier package, and running into a problem where I'm being told:
>>
>>     Credential should be scoped to a valid region, not 'eu-west-1'
>>
>> I'm following the instructions at <https://github.com/greghendershott/aws/blob/master/aws/manual.md>
>>
>> My test code is:
>>
>>     % cat glacier.rkt
>>     #lang racket/base
>>
>>     (require aws/glacier
>>              aws/keys)
>>
>>     (define vault "testvault")
>>     (region "eu-west-1")
>>     (read-keys "aws-zbu-credentials") ; local file
>>
>>     (module+ main
>>       (printf "region=~a~%" (region))
>>       (printf "Vaults: ~s~%" (list-vaults))
>>       (printf "...specifically: ~s~%" (describe-vault vault)))
>>
>> Running this produces:
>>
>>     % racket glacier.rkt
>>     region=eu-west-1
>>     aws: HTTP/1.1 403 Forbidden
>>     x-amzn-RequestId: Un3-L2zlaJBPyrIVKJrWuQcqtMMYQAr34gYUOSScg6Qepc4
>>     Content-Type: application/json
>>     Content-Length: 129
>>     Date: Sat, 27 Sep 2014 18:35:50 GMT
>>
>>      {"message":"Credential should be scoped to a valid region, not 'eu-west-1'. ","code":"InvalidSignatureException","type":"Client"}
>>     HTTP 403 "Forbidden". AWS Code="InvalidSignatureException" Message="Credential should be scoped to a valid region, not 'eu-west-1'. "
>>       context...:
>>        check-response
>>        /Users/norman/Library/Racket/6.1/pkgs/aws/aws/glacier.rkt:97:22: temp68
>>        request/redirect/uri
>>        (submod /checkouts/me/code/zbu/glacier.rkt main): [running body]
>>
>> Things I thought of:
>>
>>   * Printing (public-key)/(private-key) indicates that the credentials are being read correctly.
>>   * When I change the argument of (region) to "us-west-1", that's the region that appears in the error message.
>>   * My "testvault" vault is in eu-west-1 (and this is indeed one of the valid regions for glacier, reported in <http://docs.aws.amazon.com/general/latest/gr/rande.html> and which does have a host at http://glacier.eu-west-1.amazonaws.com
>>   * As far as I can see, credentials are _not_ scoped, but are all at us-east-1.
>>   * <http://docs.aws.amazon.com/general/latest/gr/signature-v4-troubleshooting.html> says that "IAM [...] accepts only us-east-1 as its region specification", so I'm taking it that (region) is for setting the _vault_'s region.
>>   * I'm not a great AWS expert, so I could have something in my setup broken; but if so, I've no clue what.
>>
>> If, however, I change the (region) argument to "us-east-1", I get a different error message "User: arn:aws:iam::786725553169:user/zbu is not authorized to perform: glacier:ListVaults on resource: arn:aws:glacier:us-east-1:786725553169:vaults/"  That makes sense, since there's no such vault, but it's interesting that it gets _further_ when the (region) matches the region for the IAM service.
>>
>> I don't see any other (region) equivalents for the other services supported by the package.  Is that because all of the other services supported by the package are supported by all the AWS regions, or am I missing a configuration?
>>
>> Thanks for any pointers.
>>
>> All the best,
>>
>> Norman
>>
>>
>> --
>> Norman Gray  :  http://nxg.me.uk
>> SUPA School of Physics and Astronomy, University of Glasgow, UK
>>
>>
>> ____________________
>>   Racket Users list:
>>   http://lists.racket-lang.org/users


From greghendershott at gmail.com  Sun Sep 28 17:15:21 2014
From: greghendershott at gmail.com (Greg Hendershott)
Date: Sun, 28 Sep 2014 17:15:21 -0400
Subject: [racket] aws/glacier: credential scope
In-Reply-To: <CAGspUn2au5xydsH+o9E2Y3JTA1iQbPRX58zRcitywAPWh5Djng@mail.gmail.com>
References: <0CA916BA-DD1E-47E9-BFA1-7092EC3F17F5@astro.gla.ac.uk>
 <CAGspUn0tqgSVUaDuk1C3Wb6hm6s6nQ2ps-j42pkJJSY2UESL0A@mail.gmail.com>
 <CAGspUn2au5xydsH+o9E2Y3JTA1iQbPRX58zRcitywAPWh5Djng@mail.gmail.com>
Message-ID: <CAGspUn2t_ZOaEAgM2PDvSJDrTZbw2aX0VO66QdqmXi6REjJcMQ@mail.gmail.com>

> I clicked "update" on pkgs.racket-lang.org. But it seems to be slower
> than usual to refresh. After it does, you can `raco pkg update aws` to
> get the fix.

Something seems wrong/stuck on pkgs.racket-lang.org. The top of the page says:

  "update upload in progress: there may be inconsistencies below"

It's been saying that for nearly an hour.

In the past, clicking "update" would cause a successful refresh -- the
package would get a red star indicating it's updated -- within a
minute or two.

From greghendershott at gmail.com  Sun Sep 28 17:17:32 2014
From: greghendershott at gmail.com (Greg Hendershott)
Date: Sun, 28 Sep 2014 17:17:32 -0400
Subject: [racket] aws/glacier: credential scope
In-Reply-To: <CAGspUn2t_ZOaEAgM2PDvSJDrTZbw2aX0VO66QdqmXi6REjJcMQ@mail.gmail.com>
References: <0CA916BA-DD1E-47E9-BFA1-7092EC3F17F5@astro.gla.ac.uk>
 <CAGspUn0tqgSVUaDuk1C3Wb6hm6s6nQ2ps-j42pkJJSY2UESL0A@mail.gmail.com>
 <CAGspUn2au5xydsH+o9E2Y3JTA1iQbPRX58zRcitywAPWh5Djng@mail.gmail.com>
 <CAGspUn2t_ZOaEAgM2PDvSJDrTZbw2aX0VO66QdqmXi6REjJcMQ@mail.gmail.com>
Message-ID: <CAGspUn1tMRkaZjK30rvxg+aUS7wXG5L_XT=_toU5o0YBg0pV+A@mail.gmail.com>

>> I clicked "update" on pkgs.racket-lang.org. But it seems to be slower
>> than usual to refresh. After it does, you can `raco pkg update aws` to
>> get the fix.
>
> Something seems wrong/stuck on pkgs.racket-lang.org. The top of the page says:
>
>   "update upload in progress: there may be inconsistencies below"
>
> It's been saying that for nearly an hour.

Sometimes instead it says:

  "update upload being computed: the information below may not
represent all recent changes and updates"

It has switched between those two messages at various intervals over
the last hour.

From norman at astro.gla.ac.uk  Sun Sep 28 17:40:19 2014
From: norman at astro.gla.ac.uk (Norman Gray)
Date: Sun, 28 Sep 2014 22:40:19 +0100
Subject: [racket] aws/glacier: credential scope
In-Reply-To: <CAGspUn1tMRkaZjK30rvxg+aUS7wXG5L_XT=_toU5o0YBg0pV+A@mail.gmail.com>
References: <0CA916BA-DD1E-47E9-BFA1-7092EC3F17F5@astro.gla.ac.uk>
 <CAGspUn0tqgSVUaDuk1C3Wb6hm6s6nQ2ps-j42pkJJSY2UESL0A@mail.gmail.com>
 <CAGspUn2au5xydsH+o9E2Y3JTA1iQbPRX58zRcitywAPWh5Djng@mail.gmail.com>
 <CAGspUn2t_ZOaEAgM2PDvSJDrTZbw2aX0VO66QdqmXi6REjJcMQ@mail.gmail.com>
 <CAGspUn1tMRkaZjK30rvxg+aUS7wXG5L_XT=_toU5o0YBg0pV+A@mail.gmail.com>
Message-ID: <73617243-E0D5-43C6-872B-EA0121AD5142@astro.gla.ac.uk>


On 2014 Sep 28, at 22:17, Greg Hendershott <greghendershott at gmail.com> wrote:

>>> I clicked "update" on pkgs.racket-lang.org. But it seems to be slower
>>> than usual to refresh. After it does, you can `raco pkg update aws` to
>>> get the fix.
>> 
>> Something seems wrong/stuck on pkgs.racket-lang.org. The top of the page says:
>> 
>>  "update upload in progress: there may be inconsistencies below"
>> 
>> It's been saying that for nearly an hour.
> 
> Sometimes instead it says:
> 
>  "update upload being computed: the information below may not
> represent all recent changes and updates"
> 
> It has switched between those two messages at various intervals over
> the last hour.

And in case it's a useful data point, I'm getting

% raco pkg update aws
Resolving "aws" via http://download.racket-lang.org/releases/6.1/catalog/
Resolving "aws" via http://pkgs.racket-lang.org
No updates available
%

And Greg, your fix is close enough to what I guessed might be the problem that I'm now kicking myself for not just diving in and trying a fix myself.  Garhhhh...

Norman


-- 
Norman Gray  :  http://nxg.me.uk
SUPA School of Physics and Astronomy, University of Glasgow, UK



From greghendershott at gmail.com  Sun Sep 28 18:10:55 2014
From: greghendershott at gmail.com (Greg Hendershott)
Date: Sun, 28 Sep 2014 18:10:55 -0400
Subject: [racket] aws/glacier: credential scope
In-Reply-To: <73617243-E0D5-43C6-872B-EA0121AD5142@astro.gla.ac.uk>
References: <0CA916BA-DD1E-47E9-BFA1-7092EC3F17F5@astro.gla.ac.uk>
 <CAGspUn0tqgSVUaDuk1C3Wb6hm6s6nQ2ps-j42pkJJSY2UESL0A@mail.gmail.com>
 <CAGspUn2au5xydsH+o9E2Y3JTA1iQbPRX58zRcitywAPWh5Djng@mail.gmail.com>
 <CAGspUn2t_ZOaEAgM2PDvSJDrTZbw2aX0VO66QdqmXi6REjJcMQ@mail.gmail.com>
 <CAGspUn1tMRkaZjK30rvxg+aUS7wXG5L_XT=_toU5o0YBg0pV+A@mail.gmail.com>
 <73617243-E0D5-43C6-872B-EA0121AD5142@astro.gla.ac.uk>
Message-ID: <CAGspUn2cc44ouZpTmvu8e6nW_EmtjWNccco6WoH5p2QuU=4zTQ@mail.gmail.com>

If you're in a hurry you could remove and re-install directly from GitHub:

 $ raco pkg remove aws
 $ raco pkg install git://github.com/greghendershott/aws

On Sun, Sep 28, 2014 at 5:40 PM, Norman Gray <norman at astro.gla.ac.uk> wrote:
>
> On 2014 Sep 28, at 22:17, Greg Hendershott <greghendershott at gmail.com> wrote:
>
>>>> I clicked "update" on pkgs.racket-lang.org. But it seems to be slower
>>>> than usual to refresh. After it does, you can `raco pkg update aws` to
>>>> get the fix.
>>>
>>> Something seems wrong/stuck on pkgs.racket-lang.org. The top of the page says:
>>>
>>>  "update upload in progress: there may be inconsistencies below"
>>>
>>> It's been saying that for nearly an hour.
>>
>> Sometimes instead it says:
>>
>>  "update upload being computed: the information below may not
>> represent all recent changes and updates"
>>
>> It has switched between those two messages at various intervals over
>> the last hour.
>
> And in case it's a useful data point, I'm getting
>
> % raco pkg update aws
> Resolving "aws" via http://download.racket-lang.org/releases/6.1/catalog/
> Resolving "aws" via http://pkgs.racket-lang.org
> No updates available
> %
>
> And Greg, your fix is close enough to what I guessed might be the problem that I'm now kicking myself for not just diving in and trying a fix myself.  Garhhhh...
>
> Norman
>
>
> --
> Norman Gray  :  http://nxg.me.uk
> SUPA School of Physics and Astronomy, University of Glasgow, UK
>

From matthias at ccs.neu.edu  Sun Sep 28 18:58:20 2014
From: matthias at ccs.neu.edu (Matthias Felleisen)
Date: Sun, 28 Sep 2014 18:58:20 -0400
Subject: [racket] typed racket, filters, and polymorphism
In-Reply-To: <B849BCC0-7F5C-402E-B98C-01F97E1762D8@knauth.org>
References: <1FB59E1F-0D4F-4276-BAB7-C6328866CE70@knauth.org>
 <1411920801296.55e3aa5@Nodemailer>
 <B849BCC0-7F5C-402E-B98C-01F97E1762D8@knauth.org>
Message-ID: <2B2B37BB-377F-4D28-8E69-EFEFCF8181A2@ccs.neu.edu>


If you want the type checker to ensure your program is unit-correct, I assume you also want no run-time residual of the dimensions but that is in conflict with wanting a structure because it imposes a run-time cost. Are you sure you want a unit? -- Matthias






On Sep 28, 2014, at 1:37 PM, Alexander D. Knauth wrote:

> No because I want the unit to be a struct that has a dimension field, not a symbol with various dimensions defined as unions of units.
> I want the unit to be based on the dimension, not the other way around, so that new units can be made that have the same dimension.
> 
> I have something like the number+unit struct (I called it measure), but I?ll work on more that after I have the unit struct figured out.  
> 
> On Sep 28, 2014, at 12:13 PM, Spencer florence <spencerflorence at gmail.com> wrote:
> 
>> would something like this work?
>> 
>> #lang typed/racket
>> 
>> (struct (U) number+unit ([amount : Real] [unit : U]))
>> 
>> (define-type Weight-Unit (U 'kg 'g 'mg '?g))
>> (define-type Weight (number+unit Weight-Unit))
>> (define-predicate weight? Weight)
>> 
>> (: make-weight : Real Weight-Unit -> Weight)
>> (define (make-weight n u)
>>   (number+unit n u))
>> 
>> (: +/weight : Weight Weight -> Weight)
>> ;; something something needs unit conversion
>> (define (+/weight w1 w2)
>>   (number+unit (+ (number+unit-amount w1)
>>                   (number+unit-amount w1))
>>                (number+unit-unit w1)))
>> 
>> (+/weight (make-weight 1 'kg) (make-weight 1 'kg))
>> 
>> 
>> 
>> On Sun, Sep 28, 2014 at 11:03 AM, Alexander D. Knauth <alexander at knauth.org> wrote:
>> 
>> Because the struct is representing a unit (kilograms, meters, seconds, etc.), and a unit has a dimension (mass, length, time, etc.) and I want the type-checker to be able to know what the dimension of a unit is so that the types of functions can specify the dimension that something should have. 
>> The real solution to this would probably be bounded polymorphism, but I was wondering if there was some other way to do it with occurrence typing in the guard or something like that. 
>> 
>> On Sep 28, 2014, at 11:48 AM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> wrote: 
>> 
>> > Why not do this with the type, instead of making this polymorphic? 
>> > 
>> > Sam 
>> > 
>> > On Fri, Sep 26, 2014 at 7:35 PM, Alexander D. Knauth 
>> > <alexander at knauth.org> wrote: 
>> >> Is it possible to have a struct that does certain things according to the 
>> >> guard? 
>> >> #lang typed/racket 
>> >> 
>> >> (struct (a) foo ([a : a]) #:transparent 
>> >> #:guard (lambda (a _) 
>> >> (unless (exact-integer? a) 
>> >> (error 'foo "expected Integer, given ~v" a)) 
>> >> a)) 
>> >> 
>> >> (ann (foo (ann 1 Any)) (foo Integer)) 
>> >> 
>> >> (: x : (foo Any)) 
>> >> (define x (foo 1)) 
>> >> 
>> >> (ann (foo-a x) Integer) 
>> >> 
>> >> ;. Type Checker: Polymorphic function `foo1' could not be applied to 
>> >> arguments: 
>> >> ;Argument 1: 
>> >> ; Expected: a 
>> >> ; Given: Any 
>> >> ; 
>> >> ;Result type: (foo a) 
>> >> ;Expected result: (foo Integer) 
>> >> ; in: (foo (ann 1 Any)) 
>> >> ;. Type Checker: Polymorphic function `foo-a' could not be applied to 
>> >> arguments: 
>> >> ;Argument 1: 
>> >> ; Expected: (foo a) 
>> >> ; Given: (foo Any) 
>> >> ; 
>> >> ;Result type: (a : ....) 
>> >> ;Expected result: Integer 
>> >> ; in: (foo-a x) 
>> >> 
>> >> On Sep 25, 2014, at 9:42 PM, Alexander D. Knauth <alexander at knauth.org> 
>> >> wrote: 
>> >> 
>> >> What I?m trying to accomplish is something more like this: 
>> >> #lang typed/racket 
>> >> 
>> >> (require "dimensions.rkt") 
>> >> 
>> >> (struct (d) unit ([name : Any] [scalar : Positive-Real] [dimension : d]) 
>> >> #:transparent 
>> >> #:guard (lambda (name scalar dimension _) 
>> >> (unless (dimension? dimension) 
>> >> (error 'unit "expected Dimension, given ~v" dimension)) 
>> >> (values name scalar dimension))) 
>> >> 
>> >> (define-type (Unitof d) (unit d)) 
>> >> 
>> >> (define-type Unit (Unitof Dimension)) 
>> >> 
>> >> (define Unit? (make-predicate Unit)) 
>> >> 
>> >> (define-type Unitish 
>> >> (U (Unitof Any) 
>> >> Dimension 
>> >> Positive-Real)) 
>> >> 
>> >> (: ->unit : (All (d) (case-> [(Unitof d) -> (Unitof d)] 
>> >> [Unitish -> Unit]))) 
>> >> (define (->unit u) 
>> >> (cond [(unit? u) 
>> >> (unless (Unit? u) ; this should never happen anyway because of the 
>> >> guard 
>> >> (error '->unit "expected (Unitof Dimension), given ~v" u)) 
>> >> u] 
>> >> [(dimension? u) (unit u 1 u)] 
>> >> [(positive-real? u) (unit u u dimensionless-dimension)])) 
>> >> 
>> >> 
>> >> On Sep 25, 2014, at 6:19 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> 
>> >> wrote: 
>> >> 
>> >> No, I don't think you can do this. Can you say more about what you're 
>> >> trying to accomplish? 
>> >> 
>> >> Sam 
>> >> 
>> >> On Thu, Sep 25, 2014 at 6:15 PM, Alexander D. Knauth 
>> >> <alexander at knauth.org> wrote: 
>> >> 
>> >> Do any of you have any advice for getting a function like this to 
>> >> type-check? 
>> >> #lang typed/racket 
>> >> 
>> >> (: check-int : (All (a) (case-> [a -> a] 
>> >> [Any -> Integer]))) 
>> >> (define (check-int int) 
>> >> (unless (exact-integer? int) 
>> >> (error 'check-int "expected Integer, given ~v" int)) 
>> >> int) 
>> >> 
>> >> ;. Type Checker: type mismatch 
>> >> ; expected: a 
>> >> ; given: Integer in: int 
>> >> 
>> >> 
>> >> 
>> >> ____________________ 
>> >> Racket Users list: 
>> >> http://lists.racket-lang.org/users 
>> >> 
>> >> 
>> >> ____________________ 
>> >> Racket Users list: 
>> >> http://lists.racket-lang.org/users 
>> >> 
>> >> 
>> 
>> 
>> ____________________ 
>> Racket Users list: 
>> http://lists.racket-lang.org/users 
>> 
>> 
> 
> ____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140928/0821b16d/attachment-0001.html>

From alexander at knauth.org  Sun Sep 28 20:14:43 2014
From: alexander at knauth.org (Alexander D. Knauth)
Date: Sun, 28 Sep 2014 20:14:43 -0400
Subject: [racket] typed racket, filters, and polymorphism
In-Reply-To: <2B2B37BB-377F-4D28-8E69-EFEFCF8181A2@ccs.neu.edu>
References: <1FB59E1F-0D4F-4276-BAB7-C6328866CE70@knauth.org>
 <1411920801296.55e3aa5@Nodemailer>
 <B849BCC0-7F5C-402E-B98C-01F97E1762D8@knauth.org>
 <2B2B37BB-377F-4D28-8E69-EFEFCF8181A2@ccs.neu.edu>
Message-ID: <B52D6D46-C9C1-46D6-9F1E-46800DB7B23C@knauth.org>

Yes I do want run-time residual of the dimensions, and yes I want a unit struct.  

Also if there?s run-time residual, then I can use predicates and occurrence typing, and it can also be used in untyped code with contracts instead of types.  

And also the type-checker would check if it is dimension-correct, but at run-time it would would still convert between different units of the same dimension, multiply and divide units, etc.  

On Sep 28, 2014, at 6:58 PM, Matthias Felleisen <matthias at ccs.neu.edu> wrote:

> 
> If you want the type checker to ensure your program is unit-correct, I assume you also want no run-time residual of the dimensions but that is in conflict with wanting a structure because it imposes a run-time cost. Are you sure you want a unit? -- Matthias
> 
> 
> 
> 
> 
> 
> On Sep 28, 2014, at 1:37 PM, Alexander D. Knauth wrote:
> 
>> No because I want the unit to be a struct that has a dimension field, not a symbol with various dimensions defined as unions of units.
>> I want the unit to be based on the dimension, not the other way around, so that new units can be made that have the same dimension.
>> 
>> I have something like the number+unit struct (I called it measure), but I?ll work on more that after I have the unit struct figured out.  
>> 
>> On Sep 28, 2014, at 12:13 PM, Spencer florence <spencerflorence at gmail.com> wrote:
>> 
>>> would something like this work?
>>> 
>>> #lang typed/racket
>>> 
>>> (struct (U) number+unit ([amount : Real] [unit : U]))
>>> 
>>> (define-type Weight-Unit (U 'kg 'g 'mg '?g))
>>> (define-type Weight (number+unit Weight-Unit))
>>> (define-predicate weight? Weight)
>>> 
>>> (: make-weight : Real Weight-Unit -> Weight)
>>> (define (make-weight n u)
>>>   (number+unit n u))
>>> 
>>> (: +/weight : Weight Weight -> Weight)
>>> ;; something something needs unit conversion
>>> (define (+/weight w1 w2)
>>>   (number+unit (+ (number+unit-amount w1)
>>>                   (number+unit-amount w1))
>>>                (number+unit-unit w1)))
>>> 
>>> (+/weight (make-weight 1 'kg) (make-weight 1 'kg))
>>> 
>>> 
>>> 
>>> On Sun, Sep 28, 2014 at 11:03 AM, Alexander D. Knauth <alexander at knauth.org> wrote:
>>> 
>>> Because the struct is representing a unit (kilograms, meters, seconds, etc.), and a unit has a dimension (mass, length, time, etc.) and I want the type-checker to be able to know what the dimension of a unit is so that the types of functions can specify the dimension that something should have. 
>>> The real solution to this would probably be bounded polymorphism, but I was wondering if there was some other way to do it with occurrence typing in the guard or something like that. 
>>> 
>>> On Sep 28, 2014, at 11:48 AM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> wrote: 
>>> 
>>> > Why not do this with the type, instead of making this polymorphic? 
>>> > 
>>> > Sam 
>>> > 
>>> > On Fri, Sep 26, 2014 at 7:35 PM, Alexander D. Knauth 
>>> > <alexander at knauth.org> wrote: 
>>> >> Is it possible to have a struct that does certain things according to the 
>>> >> guard? 
>>> >> #lang typed/racket 
>>> >> 
>>> >> (struct (a) foo ([a : a]) #:transparent 
>>> >> #:guard (lambda (a _) 
>>> >> (unless (exact-integer? a) 
>>> >> (error 'foo "expected Integer, given ~v" a)) 
>>> >> a)) 
>>> >> 
>>> >> (ann (foo (ann 1 Any)) (foo Integer)) 
>>> >> 
>>> >> (: x : (foo Any)) 
>>> >> (define x (foo 1)) 
>>> >> 
>>> >> (ann (foo-a x) Integer) 
>>> >> 
>>> >> ;. Type Checker: Polymorphic function `foo1' could not be applied to 
>>> >> arguments: 
>>> >> ;Argument 1: 
>>> >> ; Expected: a 
>>> >> ; Given: Any 
>>> >> ; 
>>> >> ;Result type: (foo a) 
>>> >> ;Expected result: (foo Integer) 
>>> >> ; in: (foo (ann 1 Any)) 
>>> >> ;. Type Checker: Polymorphic function `foo-a' could not be applied to 
>>> >> arguments: 
>>> >> ;Argument 1: 
>>> >> ; Expected: (foo a) 
>>> >> ; Given: (foo Any) 
>>> >> ; 
>>> >> ;Result type: (a : ....) 
>>> >> ;Expected result: Integer 
>>> >> ; in: (foo-a x) 
>>> >> 
>>> >> On Sep 25, 2014, at 9:42 PM, Alexander D. Knauth <alexander at knauth.org> 
>>> >> wrote: 
>>> >> 
>>> >> What I?m trying to accomplish is something more like this: 
>>> >> #lang typed/racket 
>>> >> 
>>> >> (require "dimensions.rkt") 
>>> >> 
>>> >> (struct (d) unit ([name : Any] [scalar : Positive-Real] [dimension : d]) 
>>> >> #:transparent 
>>> >> #:guard (lambda (name scalar dimension _) 
>>> >> (unless (dimension? dimension) 
>>> >> (error 'unit "expected Dimension, given ~v" dimension)) 
>>> >> (values name scalar dimension))) 
>>> >> 
>>> >> (define-type (Unitof d) (unit d)) 
>>> >> 
>>> >> (define-type Unit (Unitof Dimension)) 
>>> >> 
>>> >> (define Unit? (make-predicate Unit)) 
>>> >> 
>>> >> (define-type Unitish 
>>> >> (U (Unitof Any) 
>>> >> Dimension 
>>> >> Positive-Real)) 
>>> >> 
>>> >> (: ->unit : (All (d) (case-> [(Unitof d) -> (Unitof d)] 
>>> >> [Unitish -> Unit]))) 
>>> >> (define (->unit u) 
>>> >> (cond [(unit? u) 
>>> >> (unless (Unit? u) ; this should never happen anyway because of the 
>>> >> guard 
>>> >> (error '->unit "expected (Unitof Dimension), given ~v" u)) 
>>> >> u] 
>>> >> [(dimension? u) (unit u 1 u)] 
>>> >> [(positive-real? u) (unit u u dimensionless-dimension)])) 
>>> >> 
>>> >> 
>>> >> On Sep 25, 2014, at 6:19 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> 
>>> >> wrote: 
>>> >> 
>>> >> No, I don't think you can do this. Can you say more about what you're 
>>> >> trying to accomplish? 
>>> >> 
>>> >> Sam 
>>> >> 
>>> >> On Thu, Sep 25, 2014 at 6:15 PM, Alexander D. Knauth 
>>> >> <alexander at knauth.org> wrote: 
>>> >> 
>>> >> Do any of you have any advice for getting a function like this to 
>>> >> type-check? 
>>> >> #lang typed/racket 
>>> >> 
>>> >> (: check-int : (All (a) (case-> [a -> a] 
>>> >> [Any -> Integer]))) 
>>> >> (define (check-int int) 
>>> >> (unless (exact-integer? int) 
>>> >> (error 'check-int "expected Integer, given ~v" int)) 
>>> >> int) 
>>> >> 
>>> >> ;. Type Checker: type mismatch 
>>> >> ; expected: a 
>>> >> ; given: Integer in: int 
>>> >> 
>>> >> 
>>> >> 
>>> >> ____________________ 
>>> >> Racket Users list: 
>>> >> http://lists.racket-lang.org/users 
>>> >> 
>>> >> 
>>> >> ____________________ 
>>> >> Racket Users list: 
>>> >> http://lists.racket-lang.org/users 
>>> >> 
>>> >> 
>>> 
>>> 
>>> ____________________ 
>>> Racket Users list: 
>>> http://lists.racket-lang.org/users 
>>> 
>>> 
>> 
>> ____________________
>>  Racket Users list:
>>  http://lists.racket-lang.org/users
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140928/f86df9a7/attachment.html>

From matthias at ccs.neu.edu  Sun Sep 28 20:25:45 2014
From: matthias at ccs.neu.edu (Matthias Felleisen)
Date: Sun, 28 Sep 2014 20:25:45 -0400
Subject: [racket] typed racket, filters, and polymorphism
In-Reply-To: <B52D6D46-C9C1-46D6-9F1E-46800DB7B23C@knauth.org>
References: <1FB59E1F-0D4F-4276-BAB7-C6328866CE70@knauth.org>
 <1411920801296.55e3aa5@Nodemailer>
 <B849BCC0-7F5C-402E-B98C-01F97E1762D8@knauth.org>
 <2B2B37BB-377F-4D28-8E69-EFEFCF8181A2@ccs.neu.edu>
 <B52D6D46-C9C1-46D6-9F1E-46800DB7B23C@knauth.org>
Message-ID: <C3486009-3F00-4717-BD42-E18FF074FD4B@ccs.neu.edu>

	
Your message points out where gradual typing badly fails. 

Type systems combine two nearly orthogonal ideas: (1) proving theorems about your program and (2) bridging the gap between abstract linguistic constructs and concrete machine implementations. In a sense, the design of a type system should always have the goal to eliminate as much run-time overhead as possible. 

In this specific case, you have two aspects of dimensionality: dimension per se (I am sure there is a word for it) and the chosen unit to represent it. So if you know that a number denotes a distance, you can choose mm, m, and km to represent it. If you want to represent an area, you choose mm^2, m^2, km^2. Now if you want to compute the area of a rectangle, you write

 area-of-rectangle :  Real [Distance in m] * Real [Distance in m] -> Real [Area in m^2]  

If someone writes (area-of-rectangle 1 [mm] 1 [km]), there is nothing wrong with it -- except that the type checker should insert a coercion from mm to m and from km to m (multiplying/dividing by 1,000). That is, a type system for this application ought to use the type elaboration process idea and translate code as it goes. 

This is quite comparable to the idea that a machine-oriented type system needs to insert a coercion from integer to real (for example) as Algol 60 and its degraded successors do. 

Our gradual type system lacks some (intensiona) expressive power, which is why you can't get close to this ideal. 

-- Matthias








On Sep 28, 2014, at 8:14 PM, Alexander D. Knauth wrote:

> Yes I do want run-time residual of the dimensions, and yes I want a unit struct.  
> 
> Also if there?s run-time residual, then I can use predicates and occurrence typing, and it can also be used in untyped code with contracts instead of types.  
> 
> And also the type-checker would check if it is dimension-correct, but at run-time it would would still convert between different units of the same dimension, multiply and divide units, etc.  
> 
> On Sep 28, 2014, at 6:58 PM, Matthias Felleisen <matthias at ccs.neu.edu> wrote:
> 
>> 
>> If you want the type checker to ensure your program is unit-correct, I assume you also want no run-time residual of the dimensions but that is in conflict with wanting a structure because it imposes a run-time cost. Are you sure you want a unit? -- Matthias
>> 
>> 
>> 
>> 
>> 
>> 
>> On Sep 28, 2014, at 1:37 PM, Alexander D. Knauth wrote:
>> 
>>> No because I want the unit to be a struct that has a dimension field, not a symbol with various dimensions defined as unions of units.
>>> I want the unit to be based on the dimension, not the other way around, so that new units can be made that have the same dimension.
>>> 
>>> I have something like the number+unit struct (I called it measure), but I?ll work on more that after I have the unit struct figured out.  
>>> 
>>> On Sep 28, 2014, at 12:13 PM, Spencer florence <spencerflorence at gmail.com> wrote:
>>> 
>>>> would something like this work?
>>>> 
>>>> #lang typed/racket
>>>> 
>>>> (struct (U) number+unit ([amount : Real] [unit : U]))
>>>> 
>>>> (define-type Weight-Unit (U 'kg 'g 'mg '?g))
>>>> (define-type Weight (number+unit Weight-Unit))
>>>> (define-predicate weight? Weight)
>>>> 
>>>> (: make-weight : Real Weight-Unit -> Weight)
>>>> (define (make-weight n u)
>>>>   (number+unit n u))
>>>> 
>>>> (: +/weight : Weight Weight -> Weight)
>>>> ;; something something needs unit conversion
>>>> (define (+/weight w1 w2)
>>>>   (number+unit (+ (number+unit-amount w1)
>>>>                   (number+unit-amount w1))
>>>>                (number+unit-unit w1)))
>>>> 
>>>> (+/weight (make-weight 1 'kg) (make-weight 1 'kg))
>>>> 
>>>> 
>>>> 
>>>> On Sun, Sep 28, 2014 at 11:03 AM, Alexander D. Knauth <alexander at knauth.org> wrote:
>>>> 
>>>> Because the struct is representing a unit (kilograms, meters, seconds, etc.), and a unit has a dimension (mass, length, time, etc.) and I want the type-checker to be able to know what the dimension of a unit is so that the types of functions can specify the dimension that something should have. 
>>>> The real solution to this would probably be bounded polymorphism, but I was wondering if there was some other way to do it with occurrence typing in the guard or something like that. 
>>>> 
>>>> On Sep 28, 2014, at 11:48 AM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> wrote: 
>>>> 
>>>> > Why not do this with the type, instead of making this polymorphic? 
>>>> > 
>>>> > Sam 
>>>> > 
>>>> > On Fri, Sep 26, 2014 at 7:35 PM, Alexander D. Knauth 
>>>> > <alexander at knauth.org> wrote: 
>>>> >> Is it possible to have a struct that does certain things according to the 
>>>> >> guard? 
>>>> >> #lang typed/racket 
>>>> >> 
>>>> >> (struct (a) foo ([a : a]) #:transparent 
>>>> >> #:guard (lambda (a _) 
>>>> >> (unless (exact-integer? a) 
>>>> >> (error 'foo "expected Integer, given ~v" a)) 
>>>> >> a)) 
>>>> >> 
>>>> >> (ann (foo (ann 1 Any)) (foo Integer)) 
>>>> >> 
>>>> >> (: x : (foo Any)) 
>>>> >> (define x (foo 1)) 
>>>> >> 
>>>> >> (ann (foo-a x) Integer) 
>>>> >> 
>>>> >> ;. Type Checker: Polymorphic function `foo1' could not be applied to 
>>>> >> arguments: 
>>>> >> ;Argument 1: 
>>>> >> ; Expected: a 
>>>> >> ; Given: Any 
>>>> >> ; 
>>>> >> ;Result type: (foo a) 
>>>> >> ;Expected result: (foo Integer) 
>>>> >> ; in: (foo (ann 1 Any)) 
>>>> >> ;. Type Checker: Polymorphic function `foo-a' could not be applied to 
>>>> >> arguments: 
>>>> >> ;Argument 1: 
>>>> >> ; Expected: (foo a) 
>>>> >> ; Given: (foo Any) 
>>>> >> ; 
>>>> >> ;Result type: (a : ....) 
>>>> >> ;Expected result: Integer 
>>>> >> ; in: (foo-a x) 
>>>> >> 
>>>> >> On Sep 25, 2014, at 9:42 PM, Alexander D. Knauth <alexander at knauth.org> 
>>>> >> wrote: 
>>>> >> 
>>>> >> What I?m trying to accomplish is something more like this: 
>>>> >> #lang typed/racket 
>>>> >> 
>>>> >> (require "dimensions.rkt") 
>>>> >> 
>>>> >> (struct (d) unit ([name : Any] [scalar : Positive-Real] [dimension : d]) 
>>>> >> #:transparent 
>>>> >> #:guard (lambda (name scalar dimension _) 
>>>> >> (unless (dimension? dimension) 
>>>> >> (error 'unit "expected Dimension, given ~v" dimension)) 
>>>> >> (values name scalar dimension))) 
>>>> >> 
>>>> >> (define-type (Unitof d) (unit d)) 
>>>> >> 
>>>> >> (define-type Unit (Unitof Dimension)) 
>>>> >> 
>>>> >> (define Unit? (make-predicate Unit)) 
>>>> >> 
>>>> >> (define-type Unitish 
>>>> >> (U (Unitof Any) 
>>>> >> Dimension 
>>>> >> Positive-Real)) 
>>>> >> 
>>>> >> (: ->unit : (All (d) (case-> [(Unitof d) -> (Unitof d)] 
>>>> >> [Unitish -> Unit]))) 
>>>> >> (define (->unit u) 
>>>> >> (cond [(unit? u) 
>>>> >> (unless (Unit? u) ; this should never happen anyway because of the 
>>>> >> guard 
>>>> >> (error '->unit "expected (Unitof Dimension), given ~v" u)) 
>>>> >> u] 
>>>> >> [(dimension? u) (unit u 1 u)] 
>>>> >> [(positive-real? u) (unit u u dimensionless-dimension)])) 
>>>> >> 
>>>> >> 
>>>> >> On Sep 25, 2014, at 6:19 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> 
>>>> >> wrote: 
>>>> >> 
>>>> >> No, I don't think you can do this. Can you say more about what you're 
>>>> >> trying to accomplish? 
>>>> >> 
>>>> >> Sam 
>>>> >> 
>>>> >> On Thu, Sep 25, 2014 at 6:15 PM, Alexander D. Knauth 
>>>> >> <alexander at knauth.org> wrote: 
>>>> >> 
>>>> >> Do any of you have any advice for getting a function like this to 
>>>> >> type-check? 
>>>> >> #lang typed/racket 
>>>> >> 
>>>> >> (: check-int : (All (a) (case-> [a -> a] 
>>>> >> [Any -> Integer]))) 
>>>> >> (define (check-int int) 
>>>> >> (unless (exact-integer? int) 
>>>> >> (error 'check-int "expected Integer, given ~v" int)) 
>>>> >> int) 
>>>> >> 
>>>> >> ;. Type Checker: type mismatch 
>>>> >> ; expected: a 
>>>> >> ; given: Integer in: int 
>>>> >> 
>>>> >> 
>>>> >> 
>>>> >> ____________________ 
>>>> >> Racket Users list: 
>>>> >> http://lists.racket-lang.org/users 
>>>> >> 
>>>> >> 
>>>> >> ____________________ 
>>>> >> Racket Users list: 
>>>> >> http://lists.racket-lang.org/users 
>>>> >> 
>>>> >> 
>>>> 
>>>> 
>>>> ____________________ 
>>>> Racket Users list: 
>>>> http://lists.racket-lang.org/users 
>>>> 
>>>> 
>>> 
>>> ____________________
>>>  Racket Users list:
>>>  http://lists.racket-lang.org/users
>> 
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140928/d777b137/attachment-0001.html>

From alexander at knauth.org  Sun Sep 28 22:03:02 2014
From: alexander at knauth.org (Alexander D. Knauth)
Date: Sun, 28 Sep 2014 22:03:02 -0400
Subject: [racket] typed racket, filters, and polymorphism
In-Reply-To: <C3486009-3F00-4717-BD42-E18FF074FD4B@ccs.neu.edu>
References: <1FB59E1F-0D4F-4276-BAB7-C6328866CE70@knauth.org>
 <1411920801296.55e3aa5@Nodemailer>
 <B849BCC0-7F5C-402E-B98C-01F97E1762D8@knauth.org>
 <2B2B37BB-377F-4D28-8E69-EFEFCF8181A2@ccs.neu.edu>
 <B52D6D46-C9C1-46D6-9F1E-46800DB7B23C@knauth.org>
 <C3486009-3F00-4717-BD42-E18FF074FD4B@ccs.neu.edu>
Message-ID: <6237F562-9E68-4D2D-A307-49668BB8E5DF@knauth.org>



On Sep 28, 2014, at 8:25 PM, Matthias Felleisen <matthias at ccs.neu.edu> wrote:

> 	
> Your message points out where gradual typing badly fails. 
> 
> Type systems combine two nearly orthogonal ideas: (1) proving theorems about your program and (2) bridging the gap between abstract linguistic constructs and concrete machine implementations. In a sense, the design of a type system should always have the goal to eliminate as much run-time overhead as possible. 
> 
> In this specific case, you have two aspects of dimensionality: dimension per se (I am sure there is a word for it) and the chosen unit to represent it. So if you know that a number denotes a distance, you can choose mm, m, and km to represent it. If you want to represent an area, you choose mm^2, m^2, km^2. Now if you want to compute the area of a rectangle, you write
> 
>  area-of-rectangle :  Real [Distance in m] * Real [Distance in m] -> Real [Area in m^2]  
> 
> If someone writes (area-of-rectangle 1 [mm] 1 [km]), there is nothing wrong with it -- except that the type checker should insert a coercion from mm to m and from km to m (multiplying/dividing by 1,000). That is, a type system for this application ought to use the type elaboration process idea and translate code as it goes. 
> 
> This is quite comparable to the idea that a machine-oriented type system needs to insert a coercion from integer to real (for example) as Algol 60 and its degraded successors do. 

What I had in mind was for the structs to be available at run-time, but that ideally the optimizer could take them out for the intermediate operations and put them back for the final result.  

And then if this value goes into untyped code, then it knows at run-time that it has the unit (u* millimeter kilometer), which is unit=? to square-meter.  

Is this sort of like what happens with flonum unboxing?  



> 
> Our gradual type system lacks some (intensiona) expressive power, which is why you can't get close to this ideal. 
> 
> -- Matthias
> 
> 
> 
> 
> 
> 
> 
> 
> On Sep 28, 2014, at 8:14 PM, Alexander D. Knauth wrote:
> 
>> Yes I do want run-time residual of the dimensions, and yes I want a unit struct.  
>> 
>> Also if there?s run-time residual, then I can use predicates and occurrence typing, and it can also be used in untyped code with contracts instead of types.  
>> 
>> And also the type-checker would check if it is dimension-correct, but at run-time it would would still convert between different units of the same dimension, multiply and divide units, etc.  
>> 
>> On Sep 28, 2014, at 6:58 PM, Matthias Felleisen <matthias at ccs.neu.edu> wrote:
>> 
>>> 
>>> If you want the type checker to ensure your program is unit-correct, I assume you also want no run-time residual of the dimensions but that is in conflict with wanting a structure because it imposes a run-time cost. Are you sure you want a unit? -- Matthias
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> On Sep 28, 2014, at 1:37 PM, Alexander D. Knauth wrote:
>>> 
>>>> No because I want the unit to be a struct that has a dimension field, not a symbol with various dimensions defined as unions of units.
>>>> I want the unit to be based on the dimension, not the other way around, so that new units can be made that have the same dimension.
>>>> 
>>>> I have something like the number+unit struct (I called it measure), but I?ll work on more that after I have the unit struct figured out.  
>>>> 
>>>> On Sep 28, 2014, at 12:13 PM, Spencer florence <spencerflorence at gmail.com> wrote:
>>>> 
>>>>> would something like this work?
>>>>> 
>>>>> #lang typed/racket
>>>>> 
>>>>> (struct (U) number+unit ([amount : Real] [unit : U]))
>>>>> 
>>>>> (define-type Weight-Unit (U 'kg 'g 'mg '?g))
>>>>> (define-type Weight (number+unit Weight-Unit))
>>>>> (define-predicate weight? Weight)
>>>>> 
>>>>> (: make-weight : Real Weight-Unit -> Weight)
>>>>> (define (make-weight n u)
>>>>>   (number+unit n u))
>>>>> 
>>>>> (: +/weight : Weight Weight -> Weight)
>>>>> ;; something something needs unit conversion
>>>>> (define (+/weight w1 w2)
>>>>>   (number+unit (+ (number+unit-amount w1)
>>>>>                   (number+unit-amount w1))
>>>>>                (number+unit-unit w1)))
>>>>> 
>>>>> (+/weight (make-weight 1 'kg) (make-weight 1 'kg))
>>>>> 
>>>>> 
>>>>> 
>>>>> On Sun, Sep 28, 2014 at 11:03 AM, Alexander D. Knauth <alexander at knauth.org> wrote:
>>>>> 
>>>>> Because the struct is representing a unit (kilograms, meters, seconds, etc.), and a unit has a dimension (mass, length, time, etc.) and I want the type-checker to be able to know what the dimension of a unit is so that the types of functions can specify the dimension that something should have. 
>>>>> The real solution to this would probably be bounded polymorphism, but I was wondering if there was some other way to do it with occurrence typing in the guard or something like that. 
>>>>> 
>>>>> On Sep 28, 2014, at 11:48 AM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> wrote: 
>>>>> 
>>>>> > Why not do this with the type, instead of making this polymorphic? 
>>>>> > 
>>>>> > Sam 
>>>>> > 
>>>>> > On Fri, Sep 26, 2014 at 7:35 PM, Alexander D. Knauth 
>>>>> > <alexander at knauth.org> wrote: 
>>>>> >> Is it possible to have a struct that does certain things according to the 
>>>>> >> guard? 
>>>>> >> #lang typed/racket 
>>>>> >> 
>>>>> >> (struct (a) foo ([a : a]) #:transparent 
>>>>> >> #:guard (lambda (a _) 
>>>>> >> (unless (exact-integer? a) 
>>>>> >> (error 'foo "expected Integer, given ~v" a)) 
>>>>> >> a)) 
>>>>> >> 
>>>>> >> (ann (foo (ann 1 Any)) (foo Integer)) 
>>>>> >> 
>>>>> >> (: x : (foo Any)) 
>>>>> >> (define x (foo 1)) 
>>>>> >> 
>>>>> >> (ann (foo-a x) Integer) 
>>>>> >> 
>>>>> >> ;. Type Checker: Polymorphic function `foo1' could not be applied to 
>>>>> >> arguments: 
>>>>> >> ;Argument 1: 
>>>>> >> ; Expected: a 
>>>>> >> ; Given: Any 
>>>>> >> ; 
>>>>> >> ;Result type: (foo a) 
>>>>> >> ;Expected result: (foo Integer) 
>>>>> >> ; in: (foo (ann 1 Any)) 
>>>>> >> ;. Type Checker: Polymorphic function `foo-a' could not be applied to 
>>>>> >> arguments: 
>>>>> >> ;Argument 1: 
>>>>> >> ; Expected: (foo a) 
>>>>> >> ; Given: (foo Any) 
>>>>> >> ; 
>>>>> >> ;Result type: (a : ....) 
>>>>> >> ;Expected result: Integer 
>>>>> >> ; in: (foo-a x) 
>>>>> >> 
>>>>> >> On Sep 25, 2014, at 9:42 PM, Alexander D. Knauth <alexander at knauth.org> 
>>>>> >> wrote: 
>>>>> >> 
>>>>> >> What I?m trying to accomplish is something more like this: 
>>>>> >> #lang typed/racket 
>>>>> >> 
>>>>> >> (require "dimensions.rkt") 
>>>>> >> 
>>>>> >> (struct (d) unit ([name : Any] [scalar : Positive-Real] [dimension : d]) 
>>>>> >> #:transparent 
>>>>> >> #:guard (lambda (name scalar dimension _) 
>>>>> >> (unless (dimension? dimension) 
>>>>> >> (error 'unit "expected Dimension, given ~v" dimension)) 
>>>>> >> (values name scalar dimension))) 
>>>>> >> 
>>>>> >> (define-type (Unitof d) (unit d)) 
>>>>> >> 
>>>>> >> (define-type Unit (Unitof Dimension)) 
>>>>> >> 
>>>>> >> (define Unit? (make-predicate Unit)) 
>>>>> >> 
>>>>> >> (define-type Unitish 
>>>>> >> (U (Unitof Any) 
>>>>> >> Dimension 
>>>>> >> Positive-Real)) 
>>>>> >> 
>>>>> >> (: ->unit : (All (d) (case-> [(Unitof d) -> (Unitof d)] 
>>>>> >> [Unitish -> Unit]))) 
>>>>> >> (define (->unit u) 
>>>>> >> (cond [(unit? u) 
>>>>> >> (unless (Unit? u) ; this should never happen anyway because of the 
>>>>> >> guard 
>>>>> >> (error '->unit "expected (Unitof Dimension), given ~v" u)) 
>>>>> >> u] 
>>>>> >> [(dimension? u) (unit u 1 u)] 
>>>>> >> [(positive-real? u) (unit u u dimensionless-dimension)])) 
>>>>> >> 
>>>>> >> 
>>>>> >> On Sep 25, 2014, at 6:19 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> 
>>>>> >> wrote: 
>>>>> >> 
>>>>> >> No, I don't think you can do this. Can you say more about what you're 
>>>>> >> trying to accomplish? 
>>>>> >> 
>>>>> >> Sam 
>>>>> >> 
>>>>> >> On Thu, Sep 25, 2014 at 6:15 PM, Alexander D. Knauth 
>>>>> >> <alexander at knauth.org> wrote: 
>>>>> >> 
>>>>> >> Do any of you have any advice for getting a function like this to 
>>>>> >> type-check? 
>>>>> >> #lang typed/racket 
>>>>> >> 
>>>>> >> (: check-int : (All (a) (case-> [a -> a] 
>>>>> >> [Any -> Integer]))) 
>>>>> >> (define (check-int int) 
>>>>> >> (unless (exact-integer? int) 
>>>>> >> (error 'check-int "expected Integer, given ~v" int)) 
>>>>> >> int) 
>>>>> >> 
>>>>> >> ;. Type Checker: type mismatch 
>>>>> >> ; expected: a 
>>>>> >> ; given: Integer in: int 
>>>>> >> 
>>>>> >> 
>>>>> >> 
>>>>> >> ____________________ 
>>>>> >> Racket Users list: 
>>>>> >> http://lists.racket-lang.org/users 
>>>>> >> 
>>>>> >> 
>>>>> >> ____________________ 
>>>>> >> Racket Users list: 
>>>>> >> http://lists.racket-lang.org/users 
>>>>> >> 
>>>>> >> 
>>>>> 
>>>>> 
>>>>> ____________________ 
>>>>> Racket Users list: 
>>>>> http://lists.racket-lang.org/users 
>>>>> 
>>>>> 
>>>> 
>>>> ____________________
>>>>  Racket Users list:
>>>>  http://lists.racket-lang.org/users
>>> 
>> 
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140928/ca23ed92/attachment.html>

From matthias at ccs.neu.edu  Sun Sep 28 22:29:31 2014
From: matthias at ccs.neu.edu (Matthias Felleisen)
Date: Sun, 28 Sep 2014 22:29:31 -0400
Subject: [racket] typed racket, filters, and polymorphism
In-Reply-To: <6237F562-9E68-4D2D-A307-49668BB8E5DF@knauth.org>
References: <1FB59E1F-0D4F-4276-BAB7-C6328866CE70@knauth.org>
 <1411920801296.55e3aa5@Nodemailer>
 <B849BCC0-7F5C-402E-B98C-01F97E1762D8@knauth.org>
 <2B2B37BB-377F-4D28-8E69-EFEFCF8181A2@ccs.neu.edu>
 <B52D6D46-C9C1-46D6-9F1E-46800DB7B23C@knauth.org>
 <C3486009-3F00-4717-BD42-E18FF074FD4B@ccs.neu.edu>
 <6237F562-9E68-4D2D-A307-49668BB8E5DF@knauth.org>
Message-ID: <6ADD4F6D-6B79-4F32-AD11-379CD2FF87AB@ccs.neu.edu>


You might be able to unwrap such things for a local context, but then you need to wrap it again on the way out. That cost might become quite high if it happens in a loop. Then again Racket programs rely on this a lot too (as do many dynamically typed programs) so it might not be a cost that kills performance completely. But do be aware of how others deal with this issue -- Matthias







On Sep 28, 2014, at 10:03 PM, Alexander D. Knauth wrote:

> 
> 
> On Sep 28, 2014, at 8:25 PM, Matthias Felleisen <matthias at ccs.neu.edu> wrote:
> 
>> 	
>> Your message points out where gradual typing badly fails. 
>> 
>> Type systems combine two nearly orthogonal ideas: (1) proving theorems about your program and (2) bridging the gap between abstract linguistic constructs and concrete machine implementations. In a sense, the design of a type system should always have the goal to eliminate as much run-time overhead as possible. 
>> 
>> In this specific case, you have two aspects of dimensionality: dimension per se (I am sure there is a word for it) and the chosen unit to represent it. So if you know that a number denotes a distance, you can choose mm, m, and km to represent it. If you want to represent an area, you choose mm^2, m^2, km^2. Now if you want to compute the area of a rectangle, you write
>> 
>>  area-of-rectangle :  Real [Distance in m] * Real [Distance in m] -> Real [Area in m^2]  
>> 
>> If someone writes (area-of-rectangle 1 [mm] 1 [km]), there is nothing wrong with it -- except that the type checker should insert a coercion from mm to m and from km to m (multiplying/dividing by 1,000). That is, a type system for this application ought to use the type elaboration process idea and translate code as it goes. 
>> 
>> This is quite comparable to the idea that a machine-oriented type system needs to insert a coercion from integer to real (for example) as Algol 60 and its degraded successors do. 
> 
> What I had in mind was for the structs to be available at run-time, but that ideally the optimizer could take them out for the intermediate operations and put them back for the final result.  
> 
> And then if this value goes into untyped code, then it knows at run-time that it has the unit (u* millimeter kilometer), which is unit=? to square-meter.  
> 
> Is this sort of like what happens with flonum unboxing?  
> 
> 
> 
>> 
>> Our gradual type system lacks some (intensiona) expressive power, which is why you can't get close to this ideal. 
>> 
>> -- Matthias
>> 
>> 
>> 
>> 
>> 
>> 
>> 
>> 
>> On Sep 28, 2014, at 8:14 PM, Alexander D. Knauth wrote:
>> 
>>> Yes I do want run-time residual of the dimensions, and yes I want a unit struct.  
>>> 
>>> Also if there?s run-time residual, then I can use predicates and occurrence typing, and it can also be used in untyped code with contracts instead of types.  
>>> 
>>> And also the type-checker would check if it is dimension-correct, but at run-time it would would still convert between different units of the same dimension, multiply and divide units, etc.  
>>> 
>>> On Sep 28, 2014, at 6:58 PM, Matthias Felleisen <matthias at ccs.neu.edu> wrote:
>>> 
>>>> 
>>>> If you want the type checker to ensure your program is unit-correct, I assume you also want no run-time residual of the dimensions but that is in conflict with wanting a structure because it imposes a run-time cost. Are you sure you want a unit? -- Matthias
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>>> 
>>>> On Sep 28, 2014, at 1:37 PM, Alexander D. Knauth wrote:
>>>> 
>>>>> No because I want the unit to be a struct that has a dimension field, not a symbol with various dimensions defined as unions of units.
>>>>> I want the unit to be based on the dimension, not the other way around, so that new units can be made that have the same dimension.
>>>>> 
>>>>> I have something like the number+unit struct (I called it measure), but I?ll work on more that after I have the unit struct figured out.  
>>>>> 
>>>>> On Sep 28, 2014, at 12:13 PM, Spencer florence <spencerflorence at gmail.com> wrote:
>>>>> 
>>>>>> would something like this work?
>>>>>> 
>>>>>> #lang typed/racket
>>>>>> 
>>>>>> (struct (U) number+unit ([amount : Real] [unit : U]))
>>>>>> 
>>>>>> (define-type Weight-Unit (U 'kg 'g 'mg '?g))
>>>>>> (define-type Weight (number+unit Weight-Unit))
>>>>>> (define-predicate weight? Weight)
>>>>>> 
>>>>>> (: make-weight : Real Weight-Unit -> Weight)
>>>>>> (define (make-weight n u)
>>>>>>   (number+unit n u))
>>>>>> 
>>>>>> (: +/weight : Weight Weight -> Weight)
>>>>>> ;; something something needs unit conversion
>>>>>> (define (+/weight w1 w2)
>>>>>>   (number+unit (+ (number+unit-amount w1)
>>>>>>                   (number+unit-amount w1))
>>>>>>                (number+unit-unit w1)))
>>>>>> 
>>>>>> (+/weight (make-weight 1 'kg) (make-weight 1 'kg))
>>>>>> 
>>>>>> 
>>>>>> 
>>>>>> On Sun, Sep 28, 2014 at 11:03 AM, Alexander D. Knauth <alexander at knauth.org> wrote:
>>>>>> 
>>>>>> Because the struct is representing a unit (kilograms, meters, seconds, etc.), and a unit has a dimension (mass, length, time, etc.) and I want the type-checker to be able to know what the dimension of a unit is so that the types of functions can specify the dimension that something should have. 
>>>>>> The real solution to this would probably be bounded polymorphism, but I was wondering if there was some other way to do it with occurrence typing in the guard or something like that. 
>>>>>> 
>>>>>> On Sep 28, 2014, at 11:48 AM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> wrote: 
>>>>>> 
>>>>>> > Why not do this with the type, instead of making this polymorphic? 
>>>>>> > 
>>>>>> > Sam 
>>>>>> > 
>>>>>> > On Fri, Sep 26, 2014 at 7:35 PM, Alexander D. Knauth 
>>>>>> > <alexander at knauth.org> wrote: 
>>>>>> >> Is it possible to have a struct that does certain things according to the 
>>>>>> >> guard? 
>>>>>> >> #lang typed/racket 
>>>>>> >> 
>>>>>> >> (struct (a) foo ([a : a]) #:transparent 
>>>>>> >> #:guard (lambda (a _) 
>>>>>> >> (unless (exact-integer? a) 
>>>>>> >> (error 'foo "expected Integer, given ~v" a)) 
>>>>>> >> a)) 
>>>>>> >> 
>>>>>> >> (ann (foo (ann 1 Any)) (foo Integer)) 
>>>>>> >> 
>>>>>> >> (: x : (foo Any)) 
>>>>>> >> (define x (foo 1)) 
>>>>>> >> 
>>>>>> >> (ann (foo-a x) Integer) 
>>>>>> >> 
>>>>>> >> ;. Type Checker: Polymorphic function `foo1' could not be applied to 
>>>>>> >> arguments: 
>>>>>> >> ;Argument 1: 
>>>>>> >> ; Expected: a 
>>>>>> >> ; Given: Any 
>>>>>> >> ; 
>>>>>> >> ;Result type: (foo a) 
>>>>>> >> ;Expected result: (foo Integer) 
>>>>>> >> ; in: (foo (ann 1 Any)) 
>>>>>> >> ;. Type Checker: Polymorphic function `foo-a' could not be applied to 
>>>>>> >> arguments: 
>>>>>> >> ;Argument 1: 
>>>>>> >> ; Expected: (foo a) 
>>>>>> >> ; Given: (foo Any) 
>>>>>> >> ; 
>>>>>> >> ;Result type: (a : ....) 
>>>>>> >> ;Expected result: Integer 
>>>>>> >> ; in: (foo-a x) 
>>>>>> >> 
>>>>>> >> On Sep 25, 2014, at 9:42 PM, Alexander D. Knauth <alexander at knauth.org> 
>>>>>> >> wrote: 
>>>>>> >> 
>>>>>> >> What I?m trying to accomplish is something more like this: 
>>>>>> >> #lang typed/racket 
>>>>>> >> 
>>>>>> >> (require "dimensions.rkt") 
>>>>>> >> 
>>>>>> >> (struct (d) unit ([name : Any] [scalar : Positive-Real] [dimension : d]) 
>>>>>> >> #:transparent 
>>>>>> >> #:guard (lambda (name scalar dimension _) 
>>>>>> >> (unless (dimension? dimension) 
>>>>>> >> (error 'unit "expected Dimension, given ~v" dimension)) 
>>>>>> >> (values name scalar dimension))) 
>>>>>> >> 
>>>>>> >> (define-type (Unitof d) (unit d)) 
>>>>>> >> 
>>>>>> >> (define-type Unit (Unitof Dimension)) 
>>>>>> >> 
>>>>>> >> (define Unit? (make-predicate Unit)) 
>>>>>> >> 
>>>>>> >> (define-type Unitish 
>>>>>> >> (U (Unitof Any) 
>>>>>> >> Dimension 
>>>>>> >> Positive-Real)) 
>>>>>> >> 
>>>>>> >> (: ->unit : (All (d) (case-> [(Unitof d) -> (Unitof d)] 
>>>>>> >> [Unitish -> Unit]))) 
>>>>>> >> (define (->unit u) 
>>>>>> >> (cond [(unit? u) 
>>>>>> >> (unless (Unit? u) ; this should never happen anyway because of the 
>>>>>> >> guard 
>>>>>> >> (error '->unit "expected (Unitof Dimension), given ~v" u)) 
>>>>>> >> u] 
>>>>>> >> [(dimension? u) (unit u 1 u)] 
>>>>>> >> [(positive-real? u) (unit u u dimensionless-dimension)])) 
>>>>>> >> 
>>>>>> >> 
>>>>>> >> On Sep 25, 2014, at 6:19 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> 
>>>>>> >> wrote: 
>>>>>> >> 
>>>>>> >> No, I don't think you can do this. Can you say more about what you're 
>>>>>> >> trying to accomplish? 
>>>>>> >> 
>>>>>> >> Sam 
>>>>>> >> 
>>>>>> >> On Thu, Sep 25, 2014 at 6:15 PM, Alexander D. Knauth 
>>>>>> >> <alexander at knauth.org> wrote: 
>>>>>> >> 
>>>>>> >> Do any of you have any advice for getting a function like this to 
>>>>>> >> type-check? 
>>>>>> >> #lang typed/racket 
>>>>>> >> 
>>>>>> >> (: check-int : (All (a) (case-> [a -> a] 
>>>>>> >> [Any -> Integer]))) 
>>>>>> >> (define (check-int int) 
>>>>>> >> (unless (exact-integer? int) 
>>>>>> >> (error 'check-int "expected Integer, given ~v" int)) 
>>>>>> >> int) 
>>>>>> >> 
>>>>>> >> ;. Type Checker: type mismatch 
>>>>>> >> ; expected: a 
>>>>>> >> ; given: Integer in: int 
>>>>>> >> 
>>>>>> >> 
>>>>>> >> 
>>>>>> >> ____________________ 
>>>>>> >> Racket Users list: 
>>>>>> >> http://lists.racket-lang.org/users 
>>>>>> >> 
>>>>>> >> 
>>>>>> >> ____________________ 
>>>>>> >> Racket Users list: 
>>>>>> >> http://lists.racket-lang.org/users 
>>>>>> >> 
>>>>>> >> 
>>>>>> 
>>>>>> 
>>>>>> ____________________ 
>>>>>> Racket Users list: 
>>>>>> http://lists.racket-lang.org/users 
>>>>>> 
>>>>>> 
>>>>> 
>>>>> ____________________
>>>>>  Racket Users list:
>>>>>  http://lists.racket-lang.org/users
>>>> 
>>> 
>> 
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140928/d95c74e9/attachment-0001.html>

From alexander at knauth.org  Sun Sep 28 23:00:17 2014
From: alexander at knauth.org (Alexander D. Knauth)
Date: Sun, 28 Sep 2014 23:00:17 -0400
Subject: [racket] typed racket, filters, and polymorphism
In-Reply-To: <CAAtAoRp4grT_UdeJGxDpw83uBr9BMpM4gs8cZPk+bYb9ab=P+Q@mail.gmail.com>
References: <1FB59E1F-0D4F-4276-BAB7-C6328866CE70@knauth.org>
 <1411920801296.55e3aa5@Nodemailer>
 <B849BCC0-7F5C-402E-B98C-01F97E1762D8@knauth.org>
 <2B2B37BB-377F-4D28-8E69-EFEFCF8181A2@ccs.neu.edu>
 <B52D6D46-C9C1-46D6-9F1E-46800DB7B23C@knauth.org>
 <C3486009-3F00-4717-BD42-E18FF074FD4B@ccs.neu.edu>
 <6237F562-9E68-4D2D-A307-49668BB8E5DF@knauth.org>
 <6ADD4F6D-6B79-4F32-AD11-379CD2FF87AB@ccs.neu.edu>
 <CAAtAoRp4grT_UdeJGxDpw83uBr9BMpM4gs8cZPk+bYb9ab=P+Q@mail.gmail.com>
Message-ID: <0BDCBA09-5AA1-410A-A954-D2B6A8B30465@knauth.org>

I just realized that typed racket seems to ignore the guard anyway, so I could do something like this and convince the type-checker that any value I want has any type that I want:

(struct (a) foo ([a : a])
  #:guard (lambda (a _) "this value"))

(: x : (foo 1))
(define x (foo 1))

> (foo-a x)
- : Integer [more precisely: One]
"this value?


On Sep 28, 2014, at 10:41 PM, Benjamin Greenman <blg59 at cornell.edu> wrote:

> I was able to get something very similar to your example program to compile.
> 
> (struct (a) foo ([a : a]) #:transparent 
>         #:guard (lambda (a _)
>                   (define a* a)  ;; a* not "polluted" by occurrence typing
>                   (unless (exact-integer? a) 
>                     (error 'foo "expected Integer, given ~v" a)) 
>                   a*))  ;; we return a value of type a, so the typechecker is happy
> 
> From here, you can safely cast a foo-a into an Integer. It's not type safe to ann because we (deliberately) don't have the occurrence typing information, but we the programmers can be sure that the casts will never fail.
> 
> (cast (foo (ann 1 Any)) (foo Integer))
> > (foo 1)
> 
> 
> 
> On Sun, Sep 28, 2014 at 10:29 PM, Matthias Felleisen <matthias at ccs.neu.edu> wrote:
> 
> You might be able to unwrap such things for a local context, but then you need to wrap it again on the way out. That cost might become quite high if it happens in a loop. Then again Racket programs rely on this a lot too (as do many dynamically typed programs) so it might not be a cost that kills performance completely. But do be aware of how others deal with this issue -- Matthias
> 
> 
> 
> 
> 
> 
> 
> On Sep 28, 2014, at 10:03 PM, Alexander D. Knauth wrote:
> 
>> 
>> 
>> On Sep 28, 2014, at 8:25 PM, Matthias Felleisen <matthias at ccs.neu.edu> wrote:
>> 
>>> 	
>>> Your message points out where gradual typing badly fails. 
>>> 
>>> Type systems combine two nearly orthogonal ideas: (1) proving theorems about your program and (2) bridging the gap between abstract linguistic constructs and concrete machine implementations. In a sense, the design of a type system should always have the goal to eliminate as much run-time overhead as possible. 
>>> 
>>> In this specific case, you have two aspects of dimensionality: dimension per se (I am sure there is a word for it) and the chosen unit to represent it. So if you know that a number denotes a distance, you can choose mm, m, and km to represent it. If you want to represent an area, you choose mm^2, m^2, km^2. Now if you want to compute the area of a rectangle, you write
>>> 
>>>  area-of-rectangle :  Real [Distance in m] * Real [Distance in m] -> Real [Area in m^2]  
>>> 
>>> If someone writes (area-of-rectangle 1 [mm] 1 [km]), there is nothing wrong with it -- except that the type checker should insert a coercion from mm to m and from km to m (multiplying/dividing by 1,000). That is, a type system for this application ought to use the type elaboration process idea and translate code as it goes. 
>>> 
>>> This is quite comparable to the idea that a machine-oriented type system needs to insert a coercion from integer to real (for example) as Algol 60 and its degraded successors do. 
>> 
>> What I had in mind was for the structs to be available at run-time, but that ideally the optimizer could take them out for the intermediate operations and put them back for the final result.  
>> 
>> And then if this value goes into untyped code, then it knows at run-time that it has the unit (u* millimeter kilometer), which is unit=? to square-meter.  
>> 
>> Is this sort of like what happens with flonum unboxing?  
>> 
>> 
>> 
>>> 
>>> Our gradual type system lacks some (intensiona) expressive power, which is why you can't get close to this ideal. 
>>> 
>>> -- Matthias
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> On Sep 28, 2014, at 8:14 PM, Alexander D. Knauth wrote:
>>> 
>>>> Yes I do want run-time residual of the dimensions, and yes I want a unit struct.  
>>>> 
>>>> Also if there?s run-time residual, then I can use predicates and occurrence typing, and it can also be used in untyped code with contracts instead of types.  
>>>> 
>>>> And also the type-checker would check if it is dimension-correct, but at run-time it would would still convert between different units of the same dimension, multiply and divide units, etc.  
>>>> 
>>>> On Sep 28, 2014, at 6:58 PM, Matthias Felleisen <matthias at ccs.neu.edu> wrote:
>>>> 
>>>>> 
>>>>> If you want the type checker to ensure your program is unit-correct, I assume you also want no run-time residual of the dimensions but that is in conflict with wanting a structure because it imposes a run-time cost. Are you sure you want a unit? -- Matthias
>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>> 
>>>>> On Sep 28, 2014, at 1:37 PM, Alexander D. Knauth wrote:
>>>>> 
>>>>>> No because I want the unit to be a struct that has a dimension field, not a symbol with various dimensions defined as unions of units.
>>>>>> I want the unit to be based on the dimension, not the other way around, so that new units can be made that have the same dimension.
>>>>>> 
>>>>>> I have something like the number+unit struct (I called it measure), but I?ll work on more that after I have the unit struct figured out.  
>>>>>> 
>>>>>> On Sep 28, 2014, at 12:13 PM, Spencer florence <spencerflorence at gmail.com> wrote:
>>>>>> 
>>>>>>> would something like this work?
>>>>>>> 
>>>>>>> #lang typed/racket
>>>>>>> 
>>>>>>> (struct (U) number+unit ([amount : Real] [unit : U]))
>>>>>>> 
>>>>>>> (define-type Weight-Unit (U 'kg 'g 'mg '?g))
>>>>>>> (define-type Weight (number+unit Weight-Unit))
>>>>>>> (define-predicate weight? Weight)
>>>>>>> 
>>>>>>> (: make-weight : Real Weight-Unit -> Weight)
>>>>>>> (define (make-weight n u)
>>>>>>>   (number+unit n u))
>>>>>>> 
>>>>>>> (: +/weight : Weight Weight -> Weight)
>>>>>>> ;; something something needs unit conversion
>>>>>>> (define (+/weight w1 w2)
>>>>>>>   (number+unit (+ (number+unit-amount w1)
>>>>>>>                   (number+unit-amount w1))
>>>>>>>                (number+unit-unit w1)))
>>>>>>> 
>>>>>>> (+/weight (make-weight 1 'kg) (make-weight 1 'kg))
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>>> On Sun, Sep 28, 2014 at 11:03 AM, Alexander D. Knauth <alexander at knauth.org> wrote:
>>>>>>> 
>>>>>>> Because the struct is representing a unit (kilograms, meters, seconds, etc.), and a unit has a dimension (mass, length, time, etc.) and I want the type-checker to be able to know what the dimension of a unit is so that the types of functions can specify the dimension that something should have. 
>>>>>>> The real solution to this would probably be bounded polymorphism, but I was wondering if there was some other way to do it with occurrence typing in the guard or something like that. 
>>>>>>> 
>>>>>>> On Sep 28, 2014, at 11:48 AM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> wrote: 
>>>>>>> 
>>>>>>> > Why not do this with the type, instead of making this polymorphic? 
>>>>>>> > 
>>>>>>> > Sam 
>>>>>>> > 
>>>>>>> > On Fri, Sep 26, 2014 at 7:35 PM, Alexander D. Knauth 
>>>>>>> > <alexander at knauth.org> wrote: 
>>>>>>> >> Is it possible to have a struct that does certain things according to the 
>>>>>>> >> guard? 
>>>>>>> >> #lang typed/racket 
>>>>>>> >> 
>>>>>>> >> (struct (a) foo ([a : a]) #:transparent 
>>>>>>> >> #:guard (lambda (a _) 
>>>>>>> >> (unless (exact-integer? a) 
>>>>>>> >> (error 'foo "expected Integer, given ~v" a)) 
>>>>>>> >> a)) 
>>>>>>> >> 
>>>>>>> >> (ann (foo (ann 1 Any)) (foo Integer)) 
>>>>>>> >> 
>>>>>>> >> (: x : (foo Any)) 
>>>>>>> >> (define x (foo 1)) 
>>>>>>> >> 
>>>>>>> >> (ann (foo-a x) Integer) 
>>>>>>> >> 
>>>>>>> >> ;. Type Checker: Polymorphic function `foo1' could not be applied to 
>>>>>>> >> arguments: 
>>>>>>> >> ;Argument 1: 
>>>>>>> >> ; Expected: a 
>>>>>>> >> ; Given: Any 
>>>>>>> >> ; 
>>>>>>> >> ;Result type: (foo a) 
>>>>>>> >> ;Expected result: (foo Integer) 
>>>>>>> >> ; in: (foo (ann 1 Any)) 
>>>>>>> >> ;. Type Checker: Polymorphic function `foo-a' could not be applied to 
>>>>>>> >> arguments: 
>>>>>>> >> ;Argument 1: 
>>>>>>> >> ; Expected: (foo a) 
>>>>>>> >> ; Given: (foo Any) 
>>>>>>> >> ; 
>>>>>>> >> ;Result type: (a : ....) 
>>>>>>> >> ;Expected result: Integer 
>>>>>>> >> ; in: (foo-a x) 
>>>>>>> >> 
>>>>>>> >> On Sep 25, 2014, at 9:42 PM, Alexander D. Knauth <alexander at knauth.org> 
>>>>>>> >> wrote: 
>>>>>>> >> 
>>>>>>> >> What I?m trying to accomplish is something more like this: 
>>>>>>> >> #lang typed/racket 
>>>>>>> >> 
>>>>>>> >> (require "dimensions.rkt") 
>>>>>>> >> 
>>>>>>> >> (struct (d) unit ([name : Any] [scalar : Positive-Real] [dimension : d]) 
>>>>>>> >> #:transparent 
>>>>>>> >> #:guard (lambda (name scalar dimension _) 
>>>>>>> >> (unless (dimension? dimension) 
>>>>>>> >> (error 'unit "expected Dimension, given ~v" dimension)) 
>>>>>>> >> (values name scalar dimension))) 
>>>>>>> >> 
>>>>>>> >> (define-type (Unitof d) (unit d)) 
>>>>>>> >> 
>>>>>>> >> (define-type Unit (Unitof Dimension)) 
>>>>>>> >> 
>>>>>>> >> (define Unit? (make-predicate Unit)) 
>>>>>>> >> 
>>>>>>> >> (define-type Unitish 
>>>>>>> >> (U (Unitof Any) 
>>>>>>> >> Dimension 
>>>>>>> >> Positive-Real)) 
>>>>>>> >> 
>>>>>>> >> (: ->unit : (All (d) (case-> [(Unitof d) -> (Unitof d)] 
>>>>>>> >> [Unitish -> Unit]))) 
>>>>>>> >> (define (->unit u) 
>>>>>>> >> (cond [(unit? u) 
>>>>>>> >> (unless (Unit? u) ; this should never happen anyway because of the 
>>>>>>> >> guard 
>>>>>>> >> (error '->unit "expected (Unitof Dimension), given ~v" u)) 
>>>>>>> >> u] 
>>>>>>> >> [(dimension? u) (unit u 1 u)] 
>>>>>>> >> [(positive-real? u) (unit u u dimensionless-dimension)])) 
>>>>>>> >> 
>>>>>>> >> 
>>>>>>> >> On Sep 25, 2014, at 6:19 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> 
>>>>>>> >> wrote: 
>>>>>>> >> 
>>>>>>> >> No, I don't think you can do this. Can you say more about what you're 
>>>>>>> >> trying to accomplish? 
>>>>>>> >> 
>>>>>>> >> Sam 
>>>>>>> >> 
>>>>>>> >> On Thu, Sep 25, 2014 at 6:15 PM, Alexander D. Knauth 
>>>>>>> >> <alexander at knauth.org> wrote: 
>>>>>>> >> 
>>>>>>> >> Do any of you have any advice for getting a function like this to 
>>>>>>> >> type-check? 
>>>>>>> >> #lang typed/racket 
>>>>>>> >> 
>>>>>>> >> (: check-int : (All (a) (case-> [a -> a] 
>>>>>>> >> [Any -> Integer]))) 
>>>>>>> >> (define (check-int int) 
>>>>>>> >> (unless (exact-integer? int) 
>>>>>>> >> (error 'check-int "expected Integer, given ~v" int)) 
>>>>>>> >> int) 
>>>>>>> >> 
>>>>>>> >> ;. Type Checker: type mismatch 
>>>>>>> >> ; expected: a 
>>>>>>> >> ; given: Integer in: int 
>>>>>>> >> 
>>>>>>> >> 
>>>>>>> >> 
>>>>>>> >> ____________________ 
>>>>>>> >> Racket Users list: 
>>>>>>> >> http://lists.racket-lang.org/users 
>>>>>>> >> 
>>>>>>> >> 
>>>>>>> >> ____________________ 
>>>>>>> >> Racket Users list: 
>>>>>>> >> http://lists.racket-lang.org/users 
>>>>>>> >> 
>>>>>>> >> 
>>>>>>> 
>>>>>>> 
>>>>>>> ____________________ 
>>>>>>> Racket Users list: 
>>>>>>> http://lists.racket-lang.org/users 
>>>>>>> 
>>>>>>> 
>>>>>> 
>>>>>> ____________________
>>>>>>  Racket Users list:
>>>>>>  http://lists.racket-lang.org/users
>>>>> 
>>>> 
>>> 
>> 
> 
> 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140928/baf5ed5d/attachment-0001.html>

From jab at math.brown.edu  Sun Sep 28 23:36:05 2014
From: jab at math.brown.edu (jab at math.brown.edu)
Date: Sun, 28 Sep 2014 23:36:05 -0400
Subject: [racket] canonical index of Racket courses? [was: Summer
 programs learning Racket for a student]
In-Reply-To: <CAK=HD+Yj==FD+N01Dr_hAzGfm+jieKO9k=F=2VnRpg0opgmV5Q@mail.gmail.com>
References: <CANwREeXB1EWu2SjHBHuzhJKba4dHJFhdGHdg7SXafJqJ=v-nsw@mail.gmail.com>
 <CANwREeXVeAtEirzyXOGQedt8uU4EK--qYU--ZkNRGKWQ5hn8Ag@mail.gmail.com>
 <CADBC08E-D847-4AAE-87BE-D62368E91936@ccs.neu.edu>
 <CAK=HD+Yj==FD+N01Dr_hAzGfm+jieKO9k=F=2VnRpg0opgmV5Q@mail.gmail.com>
Message-ID: <CANwREeXtCqExVYCchvg4pvqFWs06tb4b=jtmU5zpdwTsAaM4Hg@mail.gmail.com>

On Wed, Sep 17, 2014 at 10:51 AM, Sam Tobin-Hochstadt <samth at cs.indiana.edu>
wrote:

> I've now created a wiki page for this, with some initial content:
> https://github.com/plt/racket/wiki/Courses-using-Racket
>

And now it's up to 22 revisions! Thanks for creating, Sam, and to everyone
who added to it.

On Tue, Sep 16, 2014 at 8:43 PM, Matthias Felleisen
> <matthias at ccs.neu.edu> wrote:
> > Do you imagine listing college courses such as Brown's 17, which uses
> > DrRacket and the teaching languages? Or do you want Coursera courses that
> > everyone can access?
>

I think linking to any class that makes educational materials publicly
available is valuable, but they should be divided into one section for
classes that don't require being admitted into a larger institution to
participate in (e.g. Bootstrap World, Coursera, summer programs for teens,
etc.), and another section for those that do.


> > And yes, we should probably create a wiki like thing for such an effort
> or
> > perhaps something like packages.racket-lang.org. I'll bring it up as we
> meet
> > in St Louis.
>

Thanks, Matthias, and hope you all had a great time in St Louis. Looks like
it was a wonderful conference.

Can the wiki page above be directly linked off racket-lang.org, perhaps a
bit more prominently? I hadn't even noticed
https://github.com/plt/racket/wiki before; now I see the link to it kind of
buried under the Contributing column of the Community section. Perhaps it's
not easy to find for others too.

Thanks again for following up on this.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140928/ac267b87/attachment.html>

From konrad.hinsen at fastmail.net  Mon Sep 29 04:44:56 2014
From: konrad.hinsen at fastmail.net (Konrad Hinsen)
Date: Mon, 29 Sep 2014 10:44:56 +0200
Subject: [racket] typed racket, filters, and polymorphism
In-Reply-To: <C3486009-3F00-4717-BD42-E18FF074FD4B@ccs.neu.edu>
References: <1FB59E1F-0D4F-4276-BAB7-C6328866CE70@knauth.org>
 <1411920801296.55e3aa5@Nodemailer>
 <B849BCC0-7F5C-402E-B98C-01F97E1762D8@knauth.org>
 <2B2B37BB-377F-4D28-8E69-EFEFCF8181A2@ccs.neu.edu>
 <B52D6D46-C9C1-46D6-9F1E-46800DB7B23C@knauth.org>
 <C3486009-3F00-4717-BD42-E18FF074FD4B@ccs.neu.edu>
Message-ID: <21545.7176.252047.613404@Ordinateur-de-Catherine-Konrad.local>

Matthias Felleisen writes:

 > Your message points out where gradual typing badly fails. 

Not just gradual typing. I am not aware of any good unit
implementation in any type system, with the exception of F# and Frink
whose type systems were explicitly modified to add unit checking as a
special case. Dependent types should permit a good solution, but I
haven't seen it done yet.

The reason you need dependent types is that the product of two
measures is another measure with a new unit, so you must be able to
construct new types (representing units) from values. For example, a
length divided by a time yields a speed. Once you can do that, scaling
a unit (e.g. from meters to kilometers) becomes a special case.

 > In this specific case, you have two aspects of dimensionality:
 > dimension per se (I am sure there is a word for it) and the chosen
 > unit to represent it.

The two terms used for this distinction are in fact "dimension" and
"unit".

 > If someone writes (area-of-rectangle 1 [mm] 1 [km]), there is
 > nothing wrong with it -- except that the type checker should insert
 > a coercion from mm to m and from km to m (multiplying/dividing by
 > 1,000).

That's a topic of hot debate in the scientific computing community.
Some argue that the type checker should consider your example as an
error, and not do any implicit conversion. The motivation is that the
expression you used is more likely to be a mistake than the meaningful
use of expressive language, since best practices recommend to use a
minimal set of units inside any piece of code.

Anyway, that's a minor detail. If your type system can handle
automatic conversion, then it can also handle its absence.


Alexander D. Knauth writes:

 > What I had in mind was for the structs to be available at run-time,
 > but that ideally the optimizer could take them out for the
 > intermediate operations and put them back for the final result.

You are actually adding a very Racket-specific requirement to the
already difficult units-as-types problem: the interplay between a
typed and an untyped dialect of the same language. I agree this would
be nice to have.

Konrad.

From greghendershott at gmail.com  Mon Sep 29 09:13:11 2014
From: greghendershott at gmail.com (Greg Hendershott)
Date: Mon, 29 Sep 2014 09:13:11 -0400
Subject: [racket] aws/glacier: credential scope
In-Reply-To: <CAGspUn2cc44ouZpTmvu8e6nW_EmtjWNccco6WoH5p2QuU=4zTQ@mail.gmail.com>
References: <0CA916BA-DD1E-47E9-BFA1-7092EC3F17F5@astro.gla.ac.uk>
 <CAGspUn0tqgSVUaDuk1C3Wb6hm6s6nQ2ps-j42pkJJSY2UESL0A@mail.gmail.com>
 <CAGspUn2au5xydsH+o9E2Y3JTA1iQbPRX58zRcitywAPWh5Djng@mail.gmail.com>
 <CAGspUn2t_ZOaEAgM2PDvSJDrTZbw2aX0VO66QdqmXi6REjJcMQ@mail.gmail.com>
 <CAGspUn1tMRkaZjK30rvxg+aUS7wXG5L_XT=_toU5o0YBg0pV+A@mail.gmail.com>
 <73617243-E0D5-43C6-872B-EA0121AD5142@astro.gla.ac.uk>
 <CAGspUn2cc44ouZpTmvu8e6nW_EmtjWNccco6WoH5p2QuU=4zTQ@mail.gmail.com>
Message-ID: <CAGspUn1Q5MvFq4i6q3KuoxBcszOpTh-RdjL=QWy+JbwHUUrXSA@mail.gmail.com>

It looks like the package server did eventually refresh a few hours ago:

    Last Updated:  9/28/2014, 4:21:56 PM  ;change on GitHub
    Last Checked:  9/29/2014, 7:51:37 AM  ;pkgs.r-l.org refresh

(I believe those times are US MT, -0600.)

On Sun, Sep 28, 2014 at 6:10 PM, Greg Hendershott
<greghendershott at gmail.com> wrote:
> If you're in a hurry you could remove and re-install directly from GitHub:
>
>  $ raco pkg remove aws
>  $ raco pkg install git://github.com/greghendershott/aws
>
> On Sun, Sep 28, 2014 at 5:40 PM, Norman Gray <norman at astro.gla.ac.uk> wrote:
>>
>> On 2014 Sep 28, at 22:17, Greg Hendershott <greghendershott at gmail.com> wrote:
>>
>>>>> I clicked "update" on pkgs.racket-lang.org. But it seems to be slower
>>>>> than usual to refresh. After it does, you can `raco pkg update aws` to
>>>>> get the fix.
>>>>
>>>> Something seems wrong/stuck on pkgs.racket-lang.org. The top of the page says:
>>>>
>>>>  "update upload in progress: there may be inconsistencies below"
>>>>
>>>> It's been saying that for nearly an hour.
>>>
>>> Sometimes instead it says:
>>>
>>>  "update upload being computed: the information below may not
>>> represent all recent changes and updates"
>>>
>>> It has switched between those two messages at various intervals over
>>> the last hour.
>>
>> And in case it's a useful data point, I'm getting
>>
>> % raco pkg update aws
>> Resolving "aws" via http://download.racket-lang.org/releases/6.1/catalog/
>> Resolving "aws" via http://pkgs.racket-lang.org
>> No updates available
>> %
>>
>> And Greg, your fix is close enough to what I guessed might be the problem that I'm now kicking myself for not just diving in and trying a fix myself.  Garhhhh...
>>
>> Norman
>>
>>
>> --
>> Norman Gray  :  http://nxg.me.uk
>> SUPA School of Physics and Astronomy, University of Glasgow, UK
>>

From ckkashyap at gmail.com  Mon Sep 29 09:39:17 2014
From: ckkashyap at gmail.com (C K Kashyap)
Date: Mon, 29 Sep 2014 19:09:17 +0530
Subject: [racket] rsound docs
Message-ID: <CAGdT1gq=E_-O9JcnaMnOEQkEK51vWxdPsFHHNSudzRoJCwHm2g@mail.gmail.com>

Hi,

I saw this nice racketcon video (
https://www.youtube.com/watch?v=DkIVzHNjNEA&list=PLXr4KViVC0qI9t3lizitiFJ1cFIeN2Gdh&index=9)
on rsound. I've installed the rsound package and have been able to run the
example from the video.

Is there a place that has mode documentation/tutorial on rsound?

Regards,
Kashyap
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140929/7ce26b08/attachment.html>

From robby at eecs.northwestern.edu  Mon Sep 29 10:32:40 2014
From: robby at eecs.northwestern.edu (Robby Findler)
Date: Mon, 29 Sep 2014 09:32:40 -0500
Subject: [racket] typed racket, filters, and polymorphism
In-Reply-To: <21545.7176.252047.613404@Ordinateur-de-Catherine-Konrad.local>
References: <1FB59E1F-0D4F-4276-BAB7-C6328866CE70@knauth.org>
 <1411920801296.55e3aa5@Nodemailer>
 <B849BCC0-7F5C-402E-B98C-01F97E1762D8@knauth.org>
 <2B2B37BB-377F-4D28-8E69-EFEFCF8181A2@ccs.neu.edu>
 <B52D6D46-C9C1-46D6-9F1E-46800DB7B23C@knauth.org>
 <C3486009-3F00-4717-BD42-E18FF074FD4B@ccs.neu.edu>
 <21545.7176.252047.613404@Ordinateur-de-Catherine-Konrad.local>
Message-ID: <CAL3TdONRKwUJ0pV3WhXDQvRgMT6mhSp6D6QGr4=W4y7g7DsK4w@mail.gmail.com>

Just in case, here is one piece of related work from the PL world and
is probably a reasonable starting point for finding others' attempts
at this problem:

http://research.microsoft.com/en-us/um/people/akenn/units/

Robby


On Mon, Sep 29, 2014 at 3:44 AM, Konrad Hinsen
<konrad.hinsen at fastmail.net> wrote:
> Matthias Felleisen writes:
>
>  > Your message points out where gradual typing badly fails.
>
> Not just gradual typing. I am not aware of any good unit
> implementation in any type system, with the exception of F# and Frink
> whose type systems were explicitly modified to add unit checking as a
> special case. Dependent types should permit a good solution, but I
> haven't seen it done yet.
>
> The reason you need dependent types is that the product of two
> measures is another measure with a new unit, so you must be able to
> construct new types (representing units) from values. For example, a
> length divided by a time yields a speed. Once you can do that, scaling
> a unit (e.g. from meters to kilometers) becomes a special case.
>
>  > In this specific case, you have two aspects of dimensionality:
>  > dimension per se (I am sure there is a word for it) and the chosen
>  > unit to represent it.
>
> The two terms used for this distinction are in fact "dimension" and
> "unit".
>
>  > If someone writes (area-of-rectangle 1 [mm] 1 [km]), there is
>  > nothing wrong with it -- except that the type checker should insert
>  > a coercion from mm to m and from km to m (multiplying/dividing by
>  > 1,000).
>
> That's a topic of hot debate in the scientific computing community.
> Some argue that the type checker should consider your example as an
> error, and not do any implicit conversion. The motivation is that the
> expression you used is more likely to be a mistake than the meaningful
> use of expressive language, since best practices recommend to use a
> minimal set of units inside any piece of code.
>
> Anyway, that's a minor detail. If your type system can handle
> automatic conversion, then it can also handle its absence.
>
>
> Alexander D. Knauth writes:
>
>  > What I had in mind was for the structs to be available at run-time,
>  > but that ideally the optimizer could take them out for the
>  > intermediate operations and put them back for the final result.
>
> You are actually adding a very Racket-specific requirement to the
> already difficult units-as-types problem: the interplay between a
> typed and an untyped dialect of the same language. I agree this would
> be nice to have.
>
> Konrad.
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users

From emanonhere at gmx.com  Mon Sep 29 11:14:43 2014
From: emanonhere at gmx.com (rrandom)
Date: Mon, 29 Sep 2014 23:14:43 +0800
Subject: [racket]  Is there any way to get the bit depth of a bitmap?
Message-ID: <54297763.3020407@gmx.com>

Hi,I am recently working in the book Digital Image Processing with
racket. I want to know is there any way to get the bit depth or color
depth of a input bitmap in Racket? I found the get-depth method in
racket/draw,bitmap%, but it only return 1 and 32, that's not what i
really want.I input a bitmap whose bit depth is 8 but i got 32.
following is the code:

> (send (read-bitmap "Fig0305(a)(DFT_no_log).tif")
get-depth)
32

Is there anything I am wrong? Thank you!

From mflatt at cs.utah.edu  Mon Sep 29 11:24:07 2014
From: mflatt at cs.utah.edu (Matthew Flatt)
Date: Mon, 29 Sep 2014 09:24:07 -0600
Subject: [racket] Is there any way to get the bit depth of a bitmap?
In-Reply-To: <54297763.3020407@gmx.com>
References: <54297763.3020407@gmx.com>
Message-ID: <20140929152408.F0E42650192@mail-svr1.cs.utah.edu>

Racket libraries can read and write images of various formats and
pixels depths, but a `bitmap%` object always represents an image using
1 or 32 bits per pixel in memory.

At Mon, 29 Sep 2014 23:14:43 +0800, rrandom wrote:
> Hi,I am recently working in the book Digital Image Processing with
> racket. I want to know is there any way to get the bit depth or color
> depth of a input bitmap in Racket? I found the get-depth method in
> racket/draw,bitmap%, but it only return 1 and 32, that's not what i
> really want.I input a bitmap whose bit depth is 8 but i got 32.
> following is the code:
> 
> > (send (read-bitmap "Fig0305(a)(DFT_no_log).tif")
> get-depth)
> 32
> 
> Is there anything I am wrong? Thank you!
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users

From mflatt at cs.utah.edu  Mon Sep 29 11:28:14 2014
From: mflatt at cs.utah.edu (Matthew Flatt)
Date: Mon, 29 Sep 2014 09:28:14 -0600
Subject: [racket] rsound docs
In-Reply-To: <CAGdT1gq=E_-O9JcnaMnOEQkEK51vWxdPsFHHNSudzRoJCwHm2g@mail.gmail.com>
References: <CAGdT1gq=E_-O9JcnaMnOEQkEK51vWxdPsFHHNSudzRoJCwHm2g@mail.gmail.com>
Message-ID: <20140929152815.C87E6650194@mail-svr1.cs.utah.edu>

When you install the package, you also install documentation for
RSound. You should be able to find it by selecting "Racket
Documentation" from the "Help" menu in DrRacket, by right-clicking on
an identifier in DrRacket that is provided by the package's library, or
by running `raco docs` on the command line.

You can also find documentation linked from the "rsound" entry at
http://pkgs.racket-lang.org/, which goes to

  http://pkg-build.racket-lang.org/doc/rsound/index.html


At Mon, 29 Sep 2014 19:09:17 +0530, C K Kashyap wrote:
> Hi,
> 
> I saw this nice racketcon video (
> https://www.youtube.com/watch?v=DkIVzHNjNEA&list=PLXr4KViVC0qI9t3lizitiFJ1cFIeN2
> Gdh&index=9)
> on rsound. I've installed the rsound package and have been able to run the
> example from the video.
> 
> Is there a place that has mode documentation/tutorial on rsound?
> 
> Regards,
> Kashyap
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users

From robby at eecs.northwestern.edu  Mon Sep 29 11:40:08 2014
From: robby at eecs.northwestern.edu (Robby Findler)
Date: Mon, 29 Sep 2014 10:40:08 -0500
Subject: [racket] typed racket, filters, and polymorphism
In-Reply-To: <CAL3TdONRKwUJ0pV3WhXDQvRgMT6mhSp6D6QGr4=W4y7g7DsK4w@mail.gmail.com>
References: <1FB59E1F-0D4F-4276-BAB7-C6328866CE70@knauth.org>
 <1411920801296.55e3aa5@Nodemailer>
 <B849BCC0-7F5C-402E-B98C-01F97E1762D8@knauth.org>
 <2B2B37BB-377F-4D28-8E69-EFEFCF8181A2@ccs.neu.edu>
 <B52D6D46-C9C1-46D6-9F1E-46800DB7B23C@knauth.org>
 <C3486009-3F00-4717-BD42-E18FF074FD4B@ccs.neu.edu>
 <21545.7176.252047.613404@Ordinateur-de-Catherine-Konrad.local>
 <CAL3TdONRKwUJ0pV3WhXDQvRgMT6mhSp6D6QGr4=W4y7g7DsK4w@mail.gmail.com>
Message-ID: <CAL3TdOPrHreseE2RxJW6x5jm5kSnKVQkjvAt-A8RuiKaarCZZw@mail.gmail.com>

Whoops, sorry: this is the same thing Konrad was talking about! Duh.
Well, hopefully the link is useful.

Robby


On Mon, Sep 29, 2014 at 9:32 AM, Robby Findler
<robby at eecs.northwestern.edu> wrote:
> Just in case, here is one piece of related work from the PL world and
> is probably a reasonable starting point for finding others' attempts
> at this problem:
>
> http://research.microsoft.com/en-us/um/people/akenn/units/
>
> Robby
>
>
> On Mon, Sep 29, 2014 at 3:44 AM, Konrad Hinsen
> <konrad.hinsen at fastmail.net> wrote:
>> Matthias Felleisen writes:
>>
>>  > Your message points out where gradual typing badly fails.
>>
>> Not just gradual typing. I am not aware of any good unit
>> implementation in any type system, with the exception of F# and Frink
>> whose type systems were explicitly modified to add unit checking as a
>> special case. Dependent types should permit a good solution, but I
>> haven't seen it done yet.
>>
>> The reason you need dependent types is that the product of two
>> measures is another measure with a new unit, so you must be able to
>> construct new types (representing units) from values. For example, a
>> length divided by a time yields a speed. Once you can do that, scaling
>> a unit (e.g. from meters to kilometers) becomes a special case.
>>
>>  > In this specific case, you have two aspects of dimensionality:
>>  > dimension per se (I am sure there is a word for it) and the chosen
>>  > unit to represent it.
>>
>> The two terms used for this distinction are in fact "dimension" and
>> "unit".
>>
>>  > If someone writes (area-of-rectangle 1 [mm] 1 [km]), there is
>>  > nothing wrong with it -- except that the type checker should insert
>>  > a coercion from mm to m and from km to m (multiplying/dividing by
>>  > 1,000).
>>
>> That's a topic of hot debate in the scientific computing community.
>> Some argue that the type checker should consider your example as an
>> error, and not do any implicit conversion. The motivation is that the
>> expression you used is more likely to be a mistake than the meaningful
>> use of expressive language, since best practices recommend to use a
>> minimal set of units inside any piece of code.
>>
>> Anyway, that's a minor detail. If your type system can handle
>> automatic conversion, then it can also handle its absence.
>>
>>
>> Alexander D. Knauth writes:
>>
>>  > What I had in mind was for the structs to be available at run-time,
>>  > but that ideally the optimizer could take them out for the
>>  > intermediate operations and put them back for the final result.
>>
>> You are actually adding a very Racket-specific requirement to the
>> already difficult units-as-types problem: the interplay between a
>> typed and an untyped dialect of the same language. I agree this would
>> be nice to have.
>>
>> Konrad.
>> ____________________
>>   Racket Users list:
>>   http://lists.racket-lang.org/users

From konrad.hinsen at fastmail.net  Mon Sep 29 12:10:29 2014
From: konrad.hinsen at fastmail.net (Konrad Hinsen)
Date: Mon, 29 Sep 2014 18:10:29 +0200
Subject: [racket] Cyclic data structures in Typed Racket
Message-ID: <21545.33909.430329.240145@Konrad-Hinsens-MacBook-Pro-2.local>

Hi everyone,

I am trying to port code creating a cyclic data structure to Typed
Racket, but notice that this is not as straightforward as I expected.

Here is the untyped code I start from:

   #lang racket

   (struct foo (x) #:mutable)
   (struct bar (x))

   (define-values (f b)
     (shared ([f (foo b)]
              [b (bar f)])
             (values f b)))

   (eq? (bar-x b) f)
   (eq? (foo-x f) b)

Switching to Typed Racket and annotating the struct definitions, I get
an error message saying that make-placeholder and placeholder-set!
from racket/base are lacking type annotations. I also get the recommendation
to use require/typed for importing them. OK, let's try:

   #lang typed/racket

   (require/typed racket/base
     [#:opaque Placeholder placeholder?]
     [make-placeholder (-> Any Placeholder)]
     [placeholder-set! (-> Placeholder Any Void)])

   (struct foo ([x : bar]) #:mutable)
   (struct bar ([x : foo]))

   (define-values (f b)
     (shared ([f (foo b)]
              [b (bar f)])
             (values f b)))

   (eq? (bar-x b) f)
   (eq? (foo-x f) b)

Same error message... which is not really surprising. I suppose typed/racket
already requires racket/base, so my additional require/typed has no effect.
But how can I fix this? I don't see any way to exclude items from typed/racket.

Konrad.

From samth at cs.indiana.edu  Mon Sep 29 12:17:29 2014
From: samth at cs.indiana.edu (Sam Tobin-Hochstadt)
Date: Mon, 29 Sep 2014 12:17:29 -0400
Subject: [racket] Cyclic data structures in Typed Racket
In-Reply-To: <21545.33909.430329.240145@Konrad-Hinsens-MacBook-Pro-2.local>
References: <21545.33909.430329.240145@Konrad-Hinsens-MacBook-Pro-2.local>
Message-ID: <CAK=HD+Y2x3d1e=caisU9xayyFqRvwm+EHT+Rg66yDCpEAhNGtA@mail.gmail.com>

I recommend doing the mutation yourself, and not using `shared`.
That's the most obvious solution.

The reason that your `require/typed` doesn't work is that it creates
new versions of the placeholder functions, but those aren't the ones
that `shared` uses internally, so Typed Racket doesn't use the types
you've given.

Note also that the types you gave aren't safe -- the way shared uses
placeholders expects particular types for particular ones (this would
result in a contract error at runtime if `shared` used your bindings).

Sam

On Mon, Sep 29, 2014 at 12:10 PM, Konrad Hinsen
<konrad.hinsen at fastmail.net> wrote:
> Hi everyone,
>
> I am trying to port code creating a cyclic data structure to Typed
> Racket, but notice that this is not as straightforward as I expected.
>
> Here is the untyped code I start from:
>
>    #lang racket
>
>    (struct foo (x) #:mutable)
>    (struct bar (x))
>
>    (define-values (f b)
>      (shared ([f (foo b)]
>               [b (bar f)])
>              (values f b)))
>
>    (eq? (bar-x b) f)
>    (eq? (foo-x f) b)
>
> Switching to Typed Racket and annotating the struct definitions, I get
> an error message saying that make-placeholder and placeholder-set!
> from racket/base are lacking type annotations. I also get the recommendation
> to use require/typed for importing them. OK, let's try:
>
>    #lang typed/racket
>
>    (require/typed racket/base
>      [#:opaque Placeholder placeholder?]
>      [make-placeholder (-> Any Placeholder)]
>      [placeholder-set! (-> Placeholder Any Void)])
>
>    (struct foo ([x : bar]) #:mutable)
>    (struct bar ([x : foo]))
>
>    (define-values (f b)
>      (shared ([f (foo b)]
>               [b (bar f)])
>              (values f b)))
>
>    (eq? (bar-x b) f)
>    (eq? (foo-x f) b)
>
> Same error message... which is not really surprising. I suppose typed/racket
> already requires racket/base, so my additional require/typed has no effect.
> But how can I fix this? I don't see any way to exclude items from typed/racket.
>
> Konrad.
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users

From konrad.hinsen at fastmail.net  Mon Sep 29 12:33:38 2014
From: konrad.hinsen at fastmail.net (Konrad Hinsen)
Date: Mon, 29 Sep 2014 18:33:38 +0200
Subject: [racket] Cyclic data structures in Typed Racket
In-Reply-To: <CAK=HD+Y2x3d1e=caisU9xayyFqRvwm+EHT+Rg66yDCpEAhNGtA@mail.gmail.com>
References: <21545.33909.430329.240145@Konrad-Hinsens-MacBook-Pro-2.local>
 <CAK=HD+Y2x3d1e=caisU9xayyFqRvwm+EHT+Rg66yDCpEAhNGtA@mail.gmail.com>
Message-ID: <21545.35298.582248.45625@Konrad-Hinsens-MacBook-Pro-2.local>

Sam Tobin-Hochstadt writes:

 > I recommend doing the mutation yourself, and not using `shared`.
 > That's the most obvious solution.

Not for me, unfortunately, but probably I am missing something
obvious.

Here's explicit mutation in untyped Racket:

   #lang racket

   (struct foo (x) #:mutable)
   (struct bar (x))

   (define f (foo (void)))
   (define b (bar f))
   (set-foo-x! f b)

   (eq? (bar-x b) f)
   (eq? (foo-x f) b)

That works fine. Moving to Typed Racket, this is rejected by the type
checker because (void) is of type Void.  Since I need a bar to make a
foo and a foo to make a bar, I don't see how I can ever initialize my
foo structure.

 > The reason that your `require/typed` doesn't work is that it creates
 > new versions of the placeholder functions, but those aren't the ones
 > that `shared` uses internally, so Typed Racket doesn't use the types
 > you've given.

That makes sense, thanks for the explanation!

Konrad.

From samth at cs.indiana.edu  Mon Sep 29 12:52:25 2014
From: samth at cs.indiana.edu (Sam Tobin-Hochstadt)
Date: Mon, 29 Sep 2014 12:52:25 -0400
Subject: [racket] Cyclic data structures in Typed Racket
In-Reply-To: <21545.35298.582248.45625@Konrad-Hinsens-MacBook-Pro-2.local>
References: <21545.33909.430329.240145@Konrad-Hinsens-MacBook-Pro-2.local>
 <CAK=HD+Y2x3d1e=caisU9xayyFqRvwm+EHT+Rg66yDCpEAhNGtA@mail.gmail.com>
 <21545.35298.582248.45625@Konrad-Hinsens-MacBook-Pro-2.local>
Message-ID: <CAK=HD+Ywfnfv-mjQPpo9GmOO8u+aV7f6ue-4motks_GkVnNRaQ@mail.gmail.com>

Ah, if you want to create truly cyclic structure with no base case
like this, then you'll have to do more work. Either you'll need to
expand the type of the `x` field in `foo` to allow an initial value,
or you'll need to create a dummy `foo` with that extended type, and
then copy the cyclic data once you've created it to a new version of
`foo`, using hash tables to track cycles. That's basically how
`shared` works, although I haven't tried implementing it in TR and it
might not be possible.

Sam

On Mon, Sep 29, 2014 at 12:33 PM, Konrad Hinsen
<konrad.hinsen at fastmail.net> wrote:
> Sam Tobin-Hochstadt writes:
>
>  > I recommend doing the mutation yourself, and not using `shared`.
>  > That's the most obvious solution.
>
> Not for me, unfortunately, but probably I am missing something
> obvious.
>
> Here's explicit mutation in untyped Racket:
>
>    #lang racket
>
>    (struct foo (x) #:mutable)
>    (struct bar (x))
>
>    (define f (foo (void)))
>    (define b (bar f))
>    (set-foo-x! f b)
>
>    (eq? (bar-x b) f)
>    (eq? (foo-x f) b)
>
> That works fine. Moving to Typed Racket, this is rejected by the type
> checker because (void) is of type Void.  Since I need a bar to make a
> foo and a foo to make a bar, I don't see how I can ever initialize my
> foo structure.
>
>  > The reason that your `require/typed` doesn't work is that it creates
>  > new versions of the placeholder functions, but those aren't the ones
>  > that `shared` uses internally, so Typed Racket doesn't use the types
>  > you've given.
>
> That makes sense, thanks for the explanation!
>
> Konrad.

From blg59 at cornell.edu  Mon Sep 29 14:00:12 2014
From: blg59 at cornell.edu (Benjamin Greenman)
Date: Mon, 29 Sep 2014 14:00:12 -0400
Subject: [racket] Cyclic data structures in Typed Racket
In-Reply-To: <CAK=HD+Ywfnfv-mjQPpo9GmOO8u+aV7f6ue-4motks_GkVnNRaQ@mail.gmail.com>
References: <21545.33909.430329.240145@Konrad-Hinsens-MacBook-Pro-2.local>
 <CAK=HD+Y2x3d1e=caisU9xayyFqRvwm+EHT+Rg66yDCpEAhNGtA@mail.gmail.com>
 <21545.35298.582248.45625@Konrad-Hinsens-MacBook-Pro-2.local>
 <CAK=HD+Ywfnfv-mjQPpo9GmOO8u+aV7f6ue-4motks_GkVnNRaQ@mail.gmail.com>
Message-ID: <CAAtAoRpHmBjH4FW1BwSOoDh41=m=BZt2q4wz9xdq+Wfj8RsRFQ@mail.gmail.com>

To be sure: Is there any TR analog to this OCaml code, initializing a
cyclic, 3-element list? I've tried using letrec in TR, but the typechecker
yells at me even if I annotate the a,b,c,d.

# type 'a mylist = Nil | Cons of 'a * 'a mylist;;
type 'a mylist = Nil | Cons of 'a * 'a mylist
# let mylist =
    let rec a = Cons(1, b)
    and     b = Cons(2, c)
    and     c = Cons(3, d)
    and     d = Cons(4, a) in
    a;;
val mylist : int mylist = Cons (1, Cons (2, Cons (3, Cons (4, <cycle>))))

On Mon, Sep 29, 2014 at 12:52 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu>
wrote:

> Ah, if you want to create truly cyclic structure with no base case
> like this, then you'll have to do more work. Either you'll need to
> expand the type of the `x` field in `foo` to allow an initial value,
> or you'll need to create a dummy `foo` with that extended type, and
> then copy the cyclic data once you've created it to a new version of
> `foo`, using hash tables to track cycles. That's basically how
> `shared` works, although I haven't tried implementing it in TR and it
> might not be possible.
>
> Sam
>
> On Mon, Sep 29, 2014 at 12:33 PM, Konrad Hinsen
> <konrad.hinsen at fastmail.net> wrote:
> > Sam Tobin-Hochstadt writes:
> >
> >  > I recommend doing the mutation yourself, and not using `shared`.
> >  > That's the most obvious solution.
> >
> > Not for me, unfortunately, but probably I am missing something
> > obvious.
> >
> > Here's explicit mutation in untyped Racket:
> >
> >    #lang racket
> >
> >    (struct foo (x) #:mutable)
> >    (struct bar (x))
> >
> >    (define f (foo (void)))
> >    (define b (bar f))
> >    (set-foo-x! f b)
> >
> >    (eq? (bar-x b) f)
> >    (eq? (foo-x f) b)
> >
> > That works fine. Moving to Typed Racket, this is rejected by the type
> > checker because (void) is of type Void.  Since I need a bar to make a
> > foo and a foo to make a bar, I don't see how I can ever initialize my
> > foo structure.
> >
> >  > The reason that your `require/typed` doesn't work is that it creates
> >  > new versions of the placeholder functions, but those aren't the ones
> >  > that `shared` uses internally, so Typed Racket doesn't use the types
> >  > you've given.
> >
> > That makes sense, thanks for the explanation!
> >
> > Konrad.
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140929/152cdf29/attachment-0001.html>

From norman at astro.gla.ac.uk  Mon Sep 29 14:06:32 2014
From: norman at astro.gla.ac.uk (Norman Gray)
Date: Mon, 29 Sep 2014 19:06:32 +0100
Subject: [racket] aws/glacier: credential scope
In-Reply-To: <CAGspUn1Q5MvFq4i6q3KuoxBcszOpTh-RdjL=QWy+JbwHUUrXSA@mail.gmail.com>
References: <0CA916BA-DD1E-47E9-BFA1-7092EC3F17F5@astro.gla.ac.uk>
 <CAGspUn0tqgSVUaDuk1C3Wb6hm6s6nQ2ps-j42pkJJSY2UESL0A@mail.gmail.com>
 <CAGspUn2au5xydsH+o9E2Y3JTA1iQbPRX58zRcitywAPWh5Djng@mail.gmail.com>
 <CAGspUn2t_ZOaEAgM2PDvSJDrTZbw2aX0VO66QdqmXi6REjJcMQ@mail.gmail.com>
 <CAGspUn1tMRkaZjK30rvxg+aUS7wXG5L_XT=_toU5o0YBg0pV+A@mail.gmail.com>
 <73617243-E0D5-43C6-872B-EA0121AD5142@astro.gla.ac.uk>
 <CAGspUn2cc44ouZpTmvu8e6nW_EmtjWNccco6WoH5p2QuU=4zTQ@mail.gmail.com>
 <CAGspUn1Q5MvFq4i6q3KuoxBcszOpTh-RdjL=QWy+JbwHUUrXSA@mail.gmail.com>
Message-ID: <45C827C4-EF46-4B7C-8328-5902E05E4ECE@astro.gla.ac.uk>


Greg, hello.

On 2014 Sep 29, at 14:13, Greg Hendershott <greghendershott at gmail.com> wrote:

> It looks like the package server did eventually refresh a few hours ago:

Hero: I've updated, and it works perfectly!

All the best,

Norman


-- 
Norman Gray  :  http://nxg.me.uk
SUPA School of Physics and Astronomy, University of Glasgow, UK


From schuster at ccs.neu.edu  Mon Sep 29 14:06:35 2014
From: schuster at ccs.neu.edu (Jonathan Schuster)
Date: Mon, 29 Sep 2014 14:06:35 -0400
Subject: [racket] Better discoverability for errortrace
Message-ID: <CAKUgKwzz17a6aUBdg-pQY5Vhskc5wF0OWSscrSrExT0wtB-p0g@mail.gmail.com>

We were having a discussion today in the Northeastern lab about how the
Racket stack traces sometimes don't give enough information, and the
solution to the problem (errrortrace) is not very discoverable if you're
not using DrRacket.

Here's an example of the kind of case I mean:

$ racket -t jpg.rkt
Assertion failed on #f
  context...:
   do-it
   /Users/acobb/programs/jpg-parse/jpg.rkt:402:2: for-loop
   decode-jpeg
   /Users/acobb/programs/jpg-parse/jpg.rkt: [running body]

$ racket -l errortrace -t jpg.rkt
Assertion failed on #f
  errortrace...:
   /Users/acobb/programs/jpg-parse/jpg.rkt:126:16: (assert #f)
   /Users/acobb/programs/jpg-parse/jpg.rkt:363:36: (huff-dc-decode dc-tree
bits)
   /Users/acobb/programs/jpg-parse/jpg.rkt:407:37: (reader bits)
   <eliding for length>
  context...:
   huff-dc-decode
   do-it
   /Users/acobb/programs/jpg-parse/jpg.rkt:406:5: for-loop
   /Users/acobb/programs/jpg-parse/jpg.rkt:402:2: for-loop
   decode-jpeg
   /Users/acobb/programs/jpg-parse/jpg.rkt: [running body]


To solve the discoverability issue, what would people think of putting a
message like "If this information is not sufficient, try errrortrace with
racket -l errortrace -t <prog>" after the default trace? Also, can anyone
think of a way to programmatically detect when a trace is not informative
enough and only display the message in those cases? I haven't looked into
the feasibility of this myself; it's just an idea that came up during our
discussion.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140929/675e9ee4/attachment.html>

From robby at eecs.northwestern.edu  Mon Sep 29 14:12:51 2014
From: robby at eecs.northwestern.edu (Robby Findler)
Date: Mon, 29 Sep 2014 13:12:51 -0500
Subject: [racket] Better discoverability for errortrace
In-Reply-To: <CAKUgKwzz17a6aUBdg-pQY5Vhskc5wF0OWSscrSrExT0wtB-p0g@mail.gmail.com>
References: <CAKUgKwzz17a6aUBdg-pQY5Vhskc5wF0OWSscrSrExT0wtB-p0g@mail.gmail.com>
Message-ID: <CAL3TdON=s4Zeyw3v10rQmUESVmavBuoinC2fTOg_yB4M9JJKXQ@mail.gmail.com>

I don't think that that particular message makes sense as part of the
default error messages. Specifically, if you disable errortrace in
drracket, then you wouldn't want to see those messages (or, perhaps
more accurately, you'd want to see different error messages). Maybe
the repl that racket starts up can do that, tho?

But perhaps this is something to add to xrepl?

Robby

On Mon, Sep 29, 2014 at 1:06 PM, Jonathan Schuster <schuster at ccs.neu.edu> wrote:
> We were having a discussion today in the Northeastern lab about how the
> Racket stack traces sometimes don't give enough information, and the
> solution to the problem (errrortrace) is not very discoverable if you're not
> using DrRacket.
>
> Here's an example of the kind of case I mean:
>
> $ racket -t jpg.rkt
> Assertion failed on #f
>   context...:
>    do-it
>    /Users/acobb/programs/jpg-parse/jpg.rkt:402:2: for-loop
>    decode-jpeg
>    /Users/acobb/programs/jpg-parse/jpg.rkt: [running body]
>
> $ racket -l errortrace -t jpg.rkt
> Assertion failed on #f
>   errortrace...:
>    /Users/acobb/programs/jpg-parse/jpg.rkt:126:16: (assert #f)
>    /Users/acobb/programs/jpg-parse/jpg.rkt:363:36: (huff-dc-decode dc-tree
> bits)
>    /Users/acobb/programs/jpg-parse/jpg.rkt:407:37: (reader bits)
>    <eliding for length>
>   context...:
>    huff-dc-decode
>    do-it
>    /Users/acobb/programs/jpg-parse/jpg.rkt:406:5: for-loop
>    /Users/acobb/programs/jpg-parse/jpg.rkt:402:2: for-loop
>    decode-jpeg
>    /Users/acobb/programs/jpg-parse/jpg.rkt: [running body]
>
>
> To solve the discoverability issue, what would people think of putting a
> message like "If this information is not sufficient, try errrortrace with
> racket -l errortrace -t <prog>" after the default trace? Also, can anyone
> think of a way to programmatically detect when a trace is not informative
> enough and only display the message in those cases? I haven't looked into
> the feasibility of this myself; it's just an idea that came up during our
> discussion.
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>

From leif at leifandersen.net  Mon Sep 29 14:19:37 2014
From: leif at leifandersen.net (Leif Andersen)
Date: Mon, 29 Sep 2014 14:19:37 -0400
Subject: [racket] Better discoverability for errortrace
In-Reply-To: <CAL3TdON=s4Zeyw3v10rQmUESVmavBuoinC2fTOg_yB4M9JJKXQ@mail.gmail.com>
References: <CAKUgKwzz17a6aUBdg-pQY5Vhskc5wF0OWSscrSrExT0wtB-p0g@mail.gmail.com>
 <CAL3TdON=s4Zeyw3v10rQmUESVmavBuoinC2fTOg_yB4M9JJKXQ@mail.gmail.com>
Message-ID: <CAAVaeED_3qijc_9Fj3Mo8088HAmq-ZjVYDYorPXLxJXFrpgivw@mail.gmail.com>

Another possibility we discussed was to give errortrace it's own flag that
shows up when you do racket -h. It certainly wouldn't really scale for all
of the libraries, but errortrace seems like it's something common enough
that it might deserve it's own flag. Maybe something like:

racket --errortrace <prog>


~Leif Andersen

On Mon, Sep 29, 2014 at 2:12 PM, Robby Findler <robby at eecs.northwestern.edu>
wrote:

> I don't think that that particular message makes sense as part of the
> default error messages. Specifically, if you disable errortrace in
> drracket, then you wouldn't want to see those messages (or, perhaps
> more accurately, you'd want to see different error messages). Maybe
> the repl that racket starts up can do that, tho?
>
> But perhaps this is something to add to xrepl?
>
> Robby
>
> On Mon, Sep 29, 2014 at 1:06 PM, Jonathan Schuster <schuster at ccs.neu.edu>
> wrote:
> > We were having a discussion today in the Northeastern lab about how the
> > Racket stack traces sometimes don't give enough information, and the
> > solution to the problem (errrortrace) is not very discoverable if you're
> not
> > using DrRacket.
> >
> > Here's an example of the kind of case I mean:
> >
> > $ racket -t jpg.rkt
> > Assertion failed on #f
> >   context...:
> >    do-it
> >    /Users/acobb/programs/jpg-parse/jpg.rkt:402:2: for-loop
> >    decode-jpeg
> >    /Users/acobb/programs/jpg-parse/jpg.rkt: [running body]
> >
> > $ racket -l errortrace -t jpg.rkt
> > Assertion failed on #f
> >   errortrace...:
> >    /Users/acobb/programs/jpg-parse/jpg.rkt:126:16: (assert #f)
> >    /Users/acobb/programs/jpg-parse/jpg.rkt:363:36: (huff-dc-decode
> dc-tree
> > bits)
> >    /Users/acobb/programs/jpg-parse/jpg.rkt:407:37: (reader bits)
> >    <eliding for length>
> >   context...:
> >    huff-dc-decode
> >    do-it
> >    /Users/acobb/programs/jpg-parse/jpg.rkt:406:5: for-loop
> >    /Users/acobb/programs/jpg-parse/jpg.rkt:402:2: for-loop
> >    decode-jpeg
> >    /Users/acobb/programs/jpg-parse/jpg.rkt: [running body]
> >
> >
> > To solve the discoverability issue, what would people think of putting a
> > message like "If this information is not sufficient, try errrortrace with
> > racket -l errortrace -t <prog>" after the default trace? Also, can anyone
> > think of a way to programmatically detect when a trace is not informative
> > enough and only display the message in those cases? I haven't looked into
> > the feasibility of this myself; it's just an idea that came up during our
> > discussion.
> >
> > ____________________
> >   Racket Users list:
> >   http://lists.racket-lang.org/users
> >
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140929/3ce2c361/attachment.html>

From schuster at ccs.neu.edu  Mon Sep 29 14:42:19 2014
From: schuster at ccs.neu.edu (Jonathan Schuster)
Date: Mon, 29 Sep 2014 14:42:19 -0400
Subject: [racket] Better discoverability for errortrace
In-Reply-To: <CAL3TdON=s4Zeyw3v10rQmUESVmavBuoinC2fTOg_yB4M9JJKXQ@mail.gmail.com>
References: <CAKUgKwzz17a6aUBdg-pQY5Vhskc5wF0OWSscrSrExT0wtB-p0g@mail.gmail.com>
 <CAL3TdON=s4Zeyw3v10rQmUESVmavBuoinC2fTOg_yB4M9JJKXQ@mail.gmail.com>
Message-ID: <CAKUgKwxf9Vh3k3uYg03Wj7Y7yeOO26PwO6oxh4tA5wxTACZ3_g@mail.gmail.com>

Hmm, I agree that there are cases when the message is extraneous, but I
think adding it only to something like xrepl doesn't solve the core
discoverability issue for those who use plain vanilla Racket on the command
line. What about a parameter that determines if the message should be
printed or not, and DrRacket could enable or disable it as needed? It could
be enabled by default, so that places like the REPL would still show the
message.

On Mon, Sep 29, 2014 at 2:12 PM, Robby Findler <robby at eecs.northwestern.edu>
wrote:

> I don't think that that particular message makes sense as part of the
> default error messages. Specifically, if you disable errortrace in
> drracket, then you wouldn't want to see those messages (or, perhaps
> more accurately, you'd want to see different error messages). Maybe
> the repl that racket starts up can do that, tho?
>
> But perhaps this is something to add to xrepl?
>
> Robby
>
> On Mon, Sep 29, 2014 at 1:06 PM, Jonathan Schuster <schuster at ccs.neu.edu>
> wrote:
> > We were having a discussion today in the Northeastern lab about how the
> > Racket stack traces sometimes don't give enough information, and the
> > solution to the problem (errrortrace) is not very discoverable if you're
> not
> > using DrRacket.
> >
> > Here's an example of the kind of case I mean:
> >
> > $ racket -t jpg.rkt
> > Assertion failed on #f
> >   context...:
> >    do-it
> >    /Users/acobb/programs/jpg-parse/jpg.rkt:402:2: for-loop
> >    decode-jpeg
> >    /Users/acobb/programs/jpg-parse/jpg.rkt: [running body]
> >
> > $ racket -l errortrace -t jpg.rkt
> > Assertion failed on #f
> >   errortrace...:
> >    /Users/acobb/programs/jpg-parse/jpg.rkt:126:16: (assert #f)
> >    /Users/acobb/programs/jpg-parse/jpg.rkt:363:36: (huff-dc-decode
> dc-tree
> > bits)
> >    /Users/acobb/programs/jpg-parse/jpg.rkt:407:37: (reader bits)
> >    <eliding for length>
> >   context...:
> >    huff-dc-decode
> >    do-it
> >    /Users/acobb/programs/jpg-parse/jpg.rkt:406:5: for-loop
> >    /Users/acobb/programs/jpg-parse/jpg.rkt:402:2: for-loop
> >    decode-jpeg
> >    /Users/acobb/programs/jpg-parse/jpg.rkt: [running body]
> >
> >
> > To solve the discoverability issue, what would people think of putting a
> > message like "If this information is not sufficient, try errrortrace with
> > racket -l errortrace -t <prog>" after the default trace? Also, can anyone
> > think of a way to programmatically detect when a trace is not informative
> > enough and only display the message in those cases? I haven't looked into
> > the feasibility of this myself; it's just an idea that came up during our
> > discussion.
> >
> > ____________________
> >   Racket Users list:
> >   http://lists.racket-lang.org/users
> >
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140929/2cb0df69/attachment-0001.html>

From matthias at ccs.neu.edu  Mon Sep 29 15:30:52 2014
From: matthias at ccs.neu.edu (Matthias Felleisen)
Date: Mon, 29 Sep 2014 15:30:52 -0400
Subject: [racket] Cyclic data structures in Typed Racket
In-Reply-To: <CAK=HD+Ywfnfv-mjQPpo9GmOO8u+aV7f6ue-4motks_GkVnNRaQ@mail.gmail.com>
References: <21545.33909.430329.240145@Konrad-Hinsens-MacBook-Pro-2.local>
 <CAK=HD+Y2x3d1e=caisU9xayyFqRvwm+EHT+Rg66yDCpEAhNGtA@mail.gmail.com>
 <21545.35298.582248.45625@Konrad-Hinsens-MacBook-Pro-2.local>
 <CAK=HD+Ywfnfv-mjQPpo9GmOO8u+aV7f6ue-4motks_GkVnNRaQ@mail.gmail.com>
Message-ID: <3597E44F-0EF5-4837-AA2D-E06E728B8B71@ccs.neu.edu>


I guess the real problem is to create suitable default values so 
that the pass the type checker or to figure out a cast that forces
TR to accept some other value as the default value (and making sure
you never use the object). 

Undefined not completely solved. Argh. -- Matthias





On Sep 29, 2014, at 12:52 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> wrote:

> Ah, if you want to create truly cyclic structure with no base case
> like this, then you'll have to do more work. Either you'll need to
> expand the type of the `x` field in `foo` to allow an initial value,
> or you'll need to create a dummy `foo` with that extended type, and
> then copy the cyclic data once you've created it to a new version of
> `foo`, using hash tables to track cycles. That's basically how
> `shared` works, although I haven't tried implementing it in TR and it
> might not be possible.
> 
> Sam
> 
> On Mon, Sep 29, 2014 at 12:33 PM, Konrad Hinsen
> <konrad.hinsen at fastmail.net> wrote:
>> Sam Tobin-Hochstadt writes:
>> 
>>> I recommend doing the mutation yourself, and not using `shared`.
>>> That's the most obvious solution.
>> 
>> Not for me, unfortunately, but probably I am missing something
>> obvious.
>> 
>> Here's explicit mutation in untyped Racket:
>> 
>>   #lang racket
>> 
>>   (struct foo (x) #:mutable)
>>   (struct bar (x))
>> 
>>   (define f (foo (void)))
>>   (define b (bar f))
>>   (set-foo-x! f b)
>> 
>>   (eq? (bar-x b) f)
>>   (eq? (foo-x f) b)
>> 
>> That works fine. Moving to Typed Racket, this is rejected by the type
>> checker because (void) is of type Void.  Since I need a bar to make a
>> foo and a foo to make a bar, I don't see how I can ever initialize my
>> foo structure.
>> 
>>> The reason that your `require/typed` doesn't work is that it creates
>>> new versions of the placeholder functions, but those aren't the ones
>>> that `shared` uses internally, so Typed Racket doesn't use the types
>>> you've given.
>> 
>> That makes sense, thanks for the explanation!
>> 
>> Konrad.
> ____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users



From neil at neilvandyke.org  Mon Sep 29 15:33:58 2014
From: neil at neilvandyke.org (Neil Van Dyke)
Date: Mon, 29 Sep 2014 15:33:58 -0400
Subject: [racket] Better discoverability for errortrace
In-Reply-To: <CAKUgKwxf9Vh3k3uYg03Wj7Y7yeOO26PwO6oxh4tA5wxTACZ3_g@mail.gmail.com>
References: <CAKUgKwzz17a6aUBdg-pQY5Vhskc5wF0OWSscrSrExT0wtB-p0g@mail.gmail.com>
 <CAL3TdON=s4Zeyw3v10rQmUESVmavBuoinC2fTOg_yB4M9JJKXQ@mail.gmail.com>
 <CAKUgKwxf9Vh3k3uYg03Wj7Y7yeOO26PwO6oxh4tA5wxTACZ3_g@mail.gmail.com>
Message-ID: <5429B426.7020404@neilvandyke.org>

Small comment: In addition to discoverability, try to have this go 
direct to *understanding* of how&when to use "errortrace".

(Multiple times over the years, I have been doing crucial performance 
optimizations to server code, and remnants of various versions of 
"errortrace" kept getting added back into the code. Pattern: get it 
optimized, incidentally also remove killer "errortrace" bits, then 
someone merges the changes, and some terrible "#! ... -l errortrace" or  
ancient "require ... errortrace.ss" somehow sneaks back into several 
different places each in multiple branches, sneaks back into hurting 
production capacity.  Even if you provide a switch in the app-specific 
build environment to enable/disable errortrace.  With lots of code and 
lots of branches and merging, it seems that errortrace can be like bed 
bugs -- terrorizing you with small bites everywhere, and you just can't 
get rid of it.  This might seem like a CM and merging problem, not an 
"errortrace" problem, but it seems somewhat peculiar to "errortrace" in 
practice.  I am forming a theory that some programmers are afraid to let 
a merge remove whatever cargo-cult incantations they found in the past 
gave them better backtraces, damn any consequences. :)  Hence my 
suggestion to have your improved discoverability send the programmer 
direct to *understanding*, rather than to cargo-cult "add this magical 
snippet to your project" with an implied "and never lose the faith".)

(Like many programming pragmatics, I know it looks silly to consider 
this seemingly little thing a matter of importance worth mention, but my 
view from the ground is that it has been important.)

Neil V.


From robby at eecs.northwestern.edu  Mon Sep 29 15:59:19 2014
From: robby at eecs.northwestern.edu (Robby Findler)
Date: Mon, 29 Sep 2014 14:59:19 -0500
Subject: [racket] Better discoverability for errortrace
In-Reply-To: <CAKUgKwxf9Vh3k3uYg03Wj7Y7yeOO26PwO6oxh4tA5wxTACZ3_g@mail.gmail.com>
References: <CAKUgKwzz17a6aUBdg-pQY5Vhskc5wF0OWSscrSrExT0wtB-p0g@mail.gmail.com>
 <CAL3TdON=s4Zeyw3v10rQmUESVmavBuoinC2fTOg_yB4M9JJKXQ@mail.gmail.com>
 <CAKUgKwxf9Vh3k3uYg03Wj7Y7yeOO26PwO6oxh4tA5wxTACZ3_g@mail.gmail.com>
Message-ID: <CAL3TdOPCAahyza53fPqtMQH23bumTO=tTOubpq+YX3Pr8ok6_g@mail.gmail.com>

If you want to go that way, I think the right thing is to change the
initial error-display-handler that racket installs when it is started
up in "open a repl" mode.

I will note, however, that there are lots of things that aren't
discoverable in this way. Indeed, I would prefer that if someone were
unhappy with their feedback of the racket repl that they were directed
to a) use xrepl, b) use Greg's emacs mode, or c) use drracket, all
ahead of being directed to use "racket -l errortrace -t file.rkt".
That latter one I view as a convenience for wizards, not something
where discovery is important.

Robby


On Mon, Sep 29, 2014 at 1:42 PM, Jonathan Schuster <schuster at ccs.neu.edu> wrote:
> Hmm, I agree that there are cases when the message is extraneous, but I
> think adding it only to something like xrepl doesn't solve the core
> discoverability issue for those who use plain vanilla Racket on the command
> line. What about a parameter that determines if the message should be
> printed or not, and DrRacket could enable or disable it as needed? It could
> be enabled by default, so that places like the REPL would still show the
> message.
>
> On Mon, Sep 29, 2014 at 2:12 PM, Robby Findler <robby at eecs.northwestern.edu>
> wrote:
>>
>> I don't think that that particular message makes sense as part of the
>> default error messages. Specifically, if you disable errortrace in
>> drracket, then you wouldn't want to see those messages (or, perhaps
>> more accurately, you'd want to see different error messages). Maybe
>> the repl that racket starts up can do that, tho?
>>
>> But perhaps this is something to add to xrepl?
>>
>> Robby
>>
>> On Mon, Sep 29, 2014 at 1:06 PM, Jonathan Schuster <schuster at ccs.neu.edu>
>> wrote:
>> > We were having a discussion today in the Northeastern lab about how the
>> > Racket stack traces sometimes don't give enough information, and the
>> > solution to the problem (errrortrace) is not very discoverable if you're
>> > not
>> > using DrRacket.
>> >
>> > Here's an example of the kind of case I mean:
>> >
>> > $ racket -t jpg.rkt
>> > Assertion failed on #f
>> >   context...:
>> >    do-it
>> >    /Users/acobb/programs/jpg-parse/jpg.rkt:402:2: for-loop
>> >    decode-jpeg
>> >    /Users/acobb/programs/jpg-parse/jpg.rkt: [running body]
>> >
>> > $ racket -l errortrace -t jpg.rkt
>> > Assertion failed on #f
>> >   errortrace...:
>> >    /Users/acobb/programs/jpg-parse/jpg.rkt:126:16: (assert #f)
>> >    /Users/acobb/programs/jpg-parse/jpg.rkt:363:36: (huff-dc-decode
>> > dc-tree
>> > bits)
>> >    /Users/acobb/programs/jpg-parse/jpg.rkt:407:37: (reader bits)
>> >    <eliding for length>
>> >   context...:
>> >    huff-dc-decode
>> >    do-it
>> >    /Users/acobb/programs/jpg-parse/jpg.rkt:406:5: for-loop
>> >    /Users/acobb/programs/jpg-parse/jpg.rkt:402:2: for-loop
>> >    decode-jpeg
>> >    /Users/acobb/programs/jpg-parse/jpg.rkt: [running body]
>> >
>> >
>> > To solve the discoverability issue, what would people think of putting a
>> > message like "If this information is not sufficient, try errrortrace
>> > with
>> > racket -l errortrace -t <prog>" after the default trace? Also, can
>> > anyone
>> > think of a way to programmatically detect when a trace is not
>> > informative
>> > enough and only display the message in those cases? I haven't looked
>> > into
>> > the feasibility of this myself; it's just an idea that came up during
>> > our
>> > discussion.
>> >
>> > ____________________
>> >   Racket Users list:
>> >   http://lists.racket-lang.org/users
>> >
>
>

From blg59 at cornell.edu  Mon Sep 29 16:32:19 2014
From: blg59 at cornell.edu (Benjamin Greenman)
Date: Mon, 29 Sep 2014 16:32:19 -0400
Subject: [racket] Cyclic data structures in Typed Racket
In-Reply-To: <3597E44F-0EF5-4837-AA2D-E06E728B8B71@ccs.neu.edu>
References: <21545.33909.430329.240145@Konrad-Hinsens-MacBook-Pro-2.local>
 <CAK=HD+Y2x3d1e=caisU9xayyFqRvwm+EHT+Rg66yDCpEAhNGtA@mail.gmail.com>
 <21545.35298.582248.45625@Konrad-Hinsens-MacBook-Pro-2.local>
 <CAK=HD+Ywfnfv-mjQPpo9GmOO8u+aV7f6ue-4motks_GkVnNRaQ@mail.gmail.com>
 <3597E44F-0EF5-4837-AA2D-E06E728B8B71@ccs.neu.edu>
Message-ID: <CAAtAoRoBwwg=DDMVKL49SxAJUNLzqRfcB5TSdbpxR0MFeTRODQ@mail.gmail.com>

Could Andrew Myer's work on masked types (to eliminate nulls for object
initialization) be relevant?
http://www.cs.cornell.edu/Projects/jmask/

On Mon, Sep 29, 2014 at 3:30 PM, Matthias Felleisen <matthias at ccs.neu.edu>
wrote:

>
> I guess the real problem is to create suitable default values so
> that the pass the type checker or to figure out a cast that forces
> TR to accept some other value as the default value (and making sure
> you never use the object).
>
> Undefined not completely solved. Argh. -- Matthias
>
>
>
>
>
> On Sep 29, 2014, at 12:52 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu>
> wrote:
>
> > Ah, if you want to create truly cyclic structure with no base case
> > like this, then you'll have to do more work. Either you'll need to
> > expand the type of the `x` field in `foo` to allow an initial value,
> > or you'll need to create a dummy `foo` with that extended type, and
> > then copy the cyclic data once you've created it to a new version of
> > `foo`, using hash tables to track cycles. That's basically how
> > `shared` works, although I haven't tried implementing it in TR and it
> > might not be possible.
> >
> > Sam
> >
> > On Mon, Sep 29, 2014 at 12:33 PM, Konrad Hinsen
> > <konrad.hinsen at fastmail.net> wrote:
> >> Sam Tobin-Hochstadt writes:
> >>
> >>> I recommend doing the mutation yourself, and not using `shared`.
> >>> That's the most obvious solution.
> >>
> >> Not for me, unfortunately, but probably I am missing something
> >> obvious.
> >>
> >> Here's explicit mutation in untyped Racket:
> >>
> >>   #lang racket
> >>
> >>   (struct foo (x) #:mutable)
> >>   (struct bar (x))
> >>
> >>   (define f (foo (void)))
> >>   (define b (bar f))
> >>   (set-foo-x! f b)
> >>
> >>   (eq? (bar-x b) f)
> >>   (eq? (foo-x f) b)
> >>
> >> That works fine. Moving to Typed Racket, this is rejected by the type
> >> checker because (void) is of type Void.  Since I need a bar to make a
> >> foo and a foo to make a bar, I don't see how I can ever initialize my
> >> foo structure.
> >>
> >>> The reason that your `require/typed` doesn't work is that it creates
> >>> new versions of the placeholder functions, but those aren't the ones
> >>> that `shared` uses internally, so Typed Racket doesn't use the types
> >>> you've given.
> >>
> >> That makes sense, thanks for the explanation!
> >>
> >> Konrad.
> > ____________________
> >  Racket Users list:
> >  http://lists.racket-lang.org/users
>
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140929/0ff9b0cf/attachment-0001.html>

From moshedeutsch115 at gmail.com  Mon Sep 29 16:40:15 2014
From: moshedeutsch115 at gmail.com (Moshe Deutsch)
Date: Mon, 29 Sep 2014 16:40:15 -0400
Subject: [racket] users Digest, Vol 109, Issue 79
In-Reply-To: <mailman.496.1412022767.679.users@racket-lang.org>
References: <mailman.496.1412022767.679.users@racket-lang.org>
Message-ID: <CAKpSqLDBQh=92z9jovMmVgDb5fbE1FuD0Bt2o_5g29-9-bM6Rg@mail.gmail.com>

Please take me off the list

Thanks


On Mon, Sep 29, 2014 at 4:32 PM,  <users-request at racket-lang.org> wrote:
> Send users mailing list submissions to
>         users at racket-lang.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>         http://lists.racket-lang.org/users/listinfo
> or, via email, send a message with subject or body 'help' to
>         users-request at racket-lang.org
>
> You can reach the person managing the list at
>         users-owner at racket-lang.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of users digest..."
>
>
> [Racket Users list:
>  http://lists.racket-lang.org/users ]
>
>
> Today's Topics:
>
>    1. Re: Cyclic data structures in Typed Racket (Matthias Felleisen)
>    2. Re: Better discoverability for errortrace (Neil Van Dyke)
>    3. Re: Better discoverability for errortrace (Robby Findler)
>    4. Re: Cyclic data structures in Typed Racket (Benjamin Greenman)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Mon, 29 Sep 2014 15:30:52 -0400
> From: Matthias Felleisen <matthias at ccs.neu.edu>
> To: Sam Tobin-Hochstadt <samth at cs.indiana.edu>
> Cc: users <users at racket-lang.org>
> Subject: Re: [racket] Cyclic data structures in Typed Racket
> Message-ID: <3597E44F-0EF5-4837-AA2D-E06E728B8B71 at ccs.neu.edu>
> Content-Type: text/plain; charset=us-ascii
>
>
> I guess the real problem is to create suitable default values so
> that the pass the type checker or to figure out a cast that forces
> TR to accept some other value as the default value (and making sure
> you never use the object).
>
> Undefined not completely solved. Argh. -- Matthias
>
>
>
>
>
> On Sep 29, 2014, at 12:52 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> wrote:
>
>> Ah, if you want to create truly cyclic structure with no base case
>> like this, then you'll have to do more work. Either you'll need to
>> expand the type of the `x` field in `foo` to allow an initial value,
>> or you'll need to create a dummy `foo` with that extended type, and
>> then copy the cyclic data once you've created it to a new version of
>> `foo`, using hash tables to track cycles. That's basically how
>> `shared` works, although I haven't tried implementing it in TR and it
>> might not be possible.
>>
>> Sam
>>
>> On Mon, Sep 29, 2014 at 12:33 PM, Konrad Hinsen
>> <konrad.hinsen at fastmail.net> wrote:
>>> Sam Tobin-Hochstadt writes:
>>>
>>>> I recommend doing the mutation yourself, and not using `shared`.
>>>> That's the most obvious solution.
>>>
>>> Not for me, unfortunately, but probably I am missing something
>>> obvious.
>>>
>>> Here's explicit mutation in untyped Racket:
>>>
>>>   #lang racket
>>>
>>>   (struct foo (x) #:mutable)
>>>   (struct bar (x))
>>>
>>>   (define f (foo (void)))
>>>   (define b (bar f))
>>>   (set-foo-x! f b)
>>>
>>>   (eq? (bar-x b) f)
>>>   (eq? (foo-x f) b)
>>>
>>> That works fine. Moving to Typed Racket, this is rejected by the type
>>> checker because (void) is of type Void.  Since I need a bar to make a
>>> foo and a foo to make a bar, I don't see how I can ever initialize my
>>> foo structure.
>>>
>>>> The reason that your `require/typed` doesn't work is that it creates
>>>> new versions of the placeholder functions, but those aren't the ones
>>>> that `shared` uses internally, so Typed Racket doesn't use the types
>>>> you've given.
>>>
>>> That makes sense, thanks for the explanation!
>>>
>>> Konrad.
>> ____________________
>>  Racket Users list:
>>  http://lists.racket-lang.org/users
>
>
>
>
> ------------------------------
>
> Message: 2
> Date: Mon, 29 Sep 2014 15:33:58 -0400
> From: Neil Van Dyke <neil at neilvandyke.org>
> To: Jonathan Schuster <schuster at ccs.neu.edu>, Robby Findler
>         <robby at eecs.northwestern.edu>
> Cc: Racket List <users at racket-lang.org>
> Subject: Re: [racket] Better discoverability for errortrace
> Message-ID: <5429B426.7020404 at neilvandyke.org>
> Content-Type: text/plain; charset=UTF-8; format=flowed
>
> Small comment: In addition to discoverability, try to have this go
> direct to *understanding* of how&when to use "errortrace".
>
> (Multiple times over the years, I have been doing crucial performance
> optimizations to server code, and remnants of various versions of
> "errortrace" kept getting added back into the code. Pattern: get it
> optimized, incidentally also remove killer "errortrace" bits, then
> someone merges the changes, and some terrible "#! ... -l errortrace" or
> ancient "require ... errortrace.ss" somehow sneaks back into several
> different places each in multiple branches, sneaks back into hurting
> production capacity.  Even if you provide a switch in the app-specific
> build environment to enable/disable errortrace.  With lots of code and
> lots of branches and merging, it seems that errortrace can be like bed
> bugs -- terrorizing you with small bites everywhere, and you just can't
> get rid of it.  This might seem like a CM and merging problem, not an
> "errortrace" problem, but it seems somewhat peculiar to "errortrace" in
> practice.  I am forming a theory that some programmers are afraid to let
> a merge remove whatever cargo-cult incantations they found in the past
> gave them better backtraces, damn any consequences. :)  Hence my
> suggestion to have your improved discoverability send the programmer
> direct to *understanding*, rather than to cargo-cult "add this magical
> snippet to your project" with an implied "and never lose the faith".)
>
> (Like many programming pragmatics, I know it looks silly to consider
> this seemingly little thing a matter of importance worth mention, but my
> view from the ground is that it has been important.)
>
> Neil V.
>
>
>
> ------------------------------
>
> Message: 3
> Date: Mon, 29 Sep 2014 14:59:19 -0500
> From: Robby Findler <robby at eecs.northwestern.edu>
> To: Jonathan Schuster <schuster at ccs.neu.edu>
> Cc: Racket List <users at racket-lang.org>
> Subject: Re: [racket] Better discoverability for errortrace
> Message-ID:
>         <CAL3TdOPCAahyza53fPqtMQH23bumTO=tTOubpq+YX3Pr8ok6_g at mail.gmail.com>
> Content-Type: text/plain; charset=UTF-8
>
> If you want to go that way, I think the right thing is to change the
> initial error-display-handler that racket installs when it is started
> up in "open a repl" mode.
>
> I will note, however, that there are lots of things that aren't
> discoverable in this way. Indeed, I would prefer that if someone were
> unhappy with their feedback of the racket repl that they were directed
> to a) use xrepl, b) use Greg's emacs mode, or c) use drracket, all
> ahead of being directed to use "racket -l errortrace -t file.rkt".
> That latter one I view as a convenience for wizards, not something
> where discovery is important.
>
> Robby
>
>
> On Mon, Sep 29, 2014 at 1:42 PM, Jonathan Schuster <schuster at ccs.neu.edu> wrote:
>> Hmm, I agree that there are cases when the message is extraneous, but I
>> think adding it only to something like xrepl doesn't solve the core
>> discoverability issue for those who use plain vanilla Racket on the command
>> line. What about a parameter that determines if the message should be
>> printed or not, and DrRacket could enable or disable it as needed? It could
>> be enabled by default, so that places like the REPL would still show the
>> message.
>>
>> On Mon, Sep 29, 2014 at 2:12 PM, Robby Findler <robby at eecs.northwestern.edu>
>> wrote:
>>>
>>> I don't think that that particular message makes sense as part of the
>>> default error messages. Specifically, if you disable errortrace in
>>> drracket, then you wouldn't want to see those messages (or, perhaps
>>> more accurately, you'd want to see different error messages). Maybe
>>> the repl that racket starts up can do that, tho?
>>>
>>> But perhaps this is something to add to xrepl?
>>>
>>> Robby
>>>
>>> On Mon, Sep 29, 2014 at 1:06 PM, Jonathan Schuster <schuster at ccs.neu.edu>
>>> wrote:
>>> > We were having a discussion today in the Northeastern lab about how the
>>> > Racket stack traces sometimes don't give enough information, and the
>>> > solution to the problem (errrortrace) is not very discoverable if you're
>>> > not
>>> > using DrRacket.
>>> >
>>> > Here's an example of the kind of case I mean:
>>> >
>>> > $ racket -t jpg.rkt
>>> > Assertion failed on #f
>>> >   context...:
>>> >    do-it
>>> >    /Users/acobb/programs/jpg-parse/jpg.rkt:402:2: for-loop
>>> >    decode-jpeg
>>> >    /Users/acobb/programs/jpg-parse/jpg.rkt: [running body]
>>> >
>>> > $ racket -l errortrace -t jpg.rkt
>>> > Assertion failed on #f
>>> >   errortrace...:
>>> >    /Users/acobb/programs/jpg-parse/jpg.rkt:126:16: (assert #f)
>>> >    /Users/acobb/programs/jpg-parse/jpg.rkt:363:36: (huff-dc-decode
>>> > dc-tree
>>> > bits)
>>> >    /Users/acobb/programs/jpg-parse/jpg.rkt:407:37: (reader bits)
>>> >    <eliding for length>
>>> >   context...:
>>> >    huff-dc-decode
>>> >    do-it
>>> >    /Users/acobb/programs/jpg-parse/jpg.rkt:406:5: for-loop
>>> >    /Users/acobb/programs/jpg-parse/jpg.rkt:402:2: for-loop
>>> >    decode-jpeg
>>> >    /Users/acobb/programs/jpg-parse/jpg.rkt: [running body]
>>> >
>>> >
>>> > To solve the discoverability issue, what would people think of putting a
>>> > message like "If this information is not sufficient, try errrortrace
>>> > with
>>> > racket -l errortrace -t <prog>" after the default trace? Also, can
>>> > anyone
>>> > think of a way to programmatically detect when a trace is not
>>> > informative
>>> > enough and only display the message in those cases? I haven't looked
>>> > into
>>> > the feasibility of this myself; it's just an idea that came up during
>>> > our
>>> > discussion.
>>> >
>>> > ____________________
>>> >   Racket Users list:
>>> >   http://lists.racket-lang.org/users
>>> >
>>
>>
>
>
> ------------------------------
>
> Message: 4
> Date: Mon, 29 Sep 2014 16:32:19 -0400
> From: Benjamin Greenman <blg59 at cornell.edu>
> To: Matthias Felleisen <matthias at ccs.neu.edu>
> Cc: users <users at racket-lang.org>, Sam Tobin-Hochstadt
>         <samth at cs.indiana.edu>
> Subject: Re: [racket] Cyclic data structures in Typed Racket
> Message-ID:
>         <CAAtAoRoBwwg=DDMVKL49SxAJUNLzqRfcB5TSdbpxR0MFeTRODQ at mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
>
> Could Andrew Myer's work on masked types (to eliminate nulls for object
> initialization) be relevant?
> http://www.cs.cornell.edu/Projects/jmask/
>
> On Mon, Sep 29, 2014 at 3:30 PM, Matthias Felleisen <matthias at ccs.neu.edu>
> wrote:
>
>>
>> I guess the real problem is to create suitable default values so
>> that the pass the type checker or to figure out a cast that forces
>> TR to accept some other value as the default value (and making sure
>> you never use the object).
>>
>> Undefined not completely solved. Argh. -- Matthias
>>
>>
>>
>>
>>
>> On Sep 29, 2014, at 12:52 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu>
>> wrote:
>>
>> > Ah, if you want to create truly cyclic structure with no base case
>> > like this, then you'll have to do more work. Either you'll need to
>> > expand the type of the `x` field in `foo` to allow an initial value,
>> > or you'll need to create a dummy `foo` with that extended type, and
>> > then copy the cyclic data once you've created it to a new version of
>> > `foo`, using hash tables to track cycles. That's basically how
>> > `shared` works, although I haven't tried implementing it in TR and it
>> > might not be possible.
>> >
>> > Sam
>> >
>> > On Mon, Sep 29, 2014 at 12:33 PM, Konrad Hinsen
>> > <konrad.hinsen at fastmail.net> wrote:
>> >> Sam Tobin-Hochstadt writes:
>> >>
>> >>> I recommend doing the mutation yourself, and not using `shared`.
>> >>> That's the most obvious solution.
>> >>
>> >> Not for me, unfortunately, but probably I am missing something
>> >> obvious.
>> >>
>> >> Here's explicit mutation in untyped Racket:
>> >>
>> >>   #lang racket
>> >>
>> >>   (struct foo (x) #:mutable)
>> >>   (struct bar (x))
>> >>
>> >>   (define f (foo (void)))
>> >>   (define b (bar f))
>> >>   (set-foo-x! f b)
>> >>
>> >>   (eq? (bar-x b) f)
>> >>   (eq? (foo-x f) b)
>> >>
>> >> That works fine. Moving to Typed Racket, this is rejected by the type
>> >> checker because (void) is of type Void.  Since I need a bar to make a
>> >> foo and a foo to make a bar, I don't see how I can ever initialize my
>> >> foo structure.
>> >>
>> >>> The reason that your `require/typed` doesn't work is that it creates
>> >>> new versions of the placeholder functions, but those aren't the ones
>> >>> that `shared` uses internally, so Typed Racket doesn't use the types
>> >>> you've given.
>> >>
>> >> That makes sense, thanks for the explanation!
>> >>
>> >> Konrad.
>> > ____________________
>> >  Racket Users list:
>> >  http://lists.racket-lang.org/users
>>
>>
>> ____________________
>>   Racket Users list:
>>   http://lists.racket-lang.org/users
>>
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140929/0ff9b0cf/attachment.html>
>
> End of users Digest, Vol 109, Issue 79
> **************************************

From moshedeutsch115 at gmail.com  Mon Sep 29 16:40:43 2014
From: moshedeutsch115 at gmail.com (Moshe Deutsch)
Date: Mon, 29 Sep 2014 16:40:43 -0400
Subject: [racket] users Digest, Vol 109, Issue 77
In-Reply-To: <mailman.492.1412013639.679.users@racket-lang.org>
References: <mailman.492.1412013639.679.users@racket-lang.org>
Message-ID: <CAKpSqLBP7pAaSVC4K52U15KqoUZew9OyP1+JYd4FsrisV-Dpsg@mail.gmail.com>

Please take me off the list

Thanks

On Mon, Sep 29, 2014 at 2:00 PM,  <users-request at racket-lang.org> wrote:
> Send users mailing list submissions to
>         users at racket-lang.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>         http://lists.racket-lang.org/users/listinfo
> or, via email, send a message with subject or body 'help' to
>         users-request at racket-lang.org
>
> You can reach the person managing the list at
>         users-owner at racket-lang.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of users digest..."
>
>
> [Racket Users list:
>  http://lists.racket-lang.org/users ]
>
>
> Today's Topics:
>
>    1. Cyclic data structures in Typed Racket (Konrad Hinsen)
>    2. Re: Cyclic data structures in Typed Racket (Sam Tobin-Hochstadt)
>    3. Re: Cyclic data structures in Typed Racket (Konrad Hinsen)
>    4. Re: Cyclic data structures in Typed Racket (Sam Tobin-Hochstadt)
>    5. Re: Cyclic data structures in Typed Racket (Benjamin Greenman)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Mon, 29 Sep 2014 18:10:29 +0200
> From: Konrad Hinsen <konrad.hinsen at fastmail.net>
> To: users at racket-lang.org
> Subject: [racket] Cyclic data structures in Typed Racket
> Message-ID:
>         <21545.33909.430329.240145 at Konrad-Hinsens-MacBook-Pro-2.local>
> Content-Type: text/plain; charset=us-ascii
>
> Hi everyone,
>
> I am trying to port code creating a cyclic data structure to Typed
> Racket, but notice that this is not as straightforward as I expected.
>
> Here is the untyped code I start from:
>
>    #lang racket
>
>    (struct foo (x) #:mutable)
>    (struct bar (x))
>
>    (define-values (f b)
>      (shared ([f (foo b)]
>               [b (bar f)])
>              (values f b)))
>
>    (eq? (bar-x b) f)
>    (eq? (foo-x f) b)
>
> Switching to Typed Racket and annotating the struct definitions, I get
> an error message saying that make-placeholder and placeholder-set!
> from racket/base are lacking type annotations. I also get the recommendation
> to use require/typed for importing them. OK, let's try:
>
>    #lang typed/racket
>
>    (require/typed racket/base
>      [#:opaque Placeholder placeholder?]
>      [make-placeholder (-> Any Placeholder)]
>      [placeholder-set! (-> Placeholder Any Void)])
>
>    (struct foo ([x : bar]) #:mutable)
>    (struct bar ([x : foo]))
>
>    (define-values (f b)
>      (shared ([f (foo b)]
>               [b (bar f)])
>              (values f b)))
>
>    (eq? (bar-x b) f)
>    (eq? (foo-x f) b)
>
> Same error message... which is not really surprising. I suppose typed/racket
> already requires racket/base, so my additional require/typed has no effect.
> But how can I fix this? I don't see any way to exclude items from typed/racket.
>
> Konrad.
>
>
> ------------------------------
>
> Message: 2
> Date: Mon, 29 Sep 2014 12:17:29 -0400
> From: Sam Tobin-Hochstadt <samth at cs.indiana.edu>
> To: Konrad Hinsen <konrad.hinsen at fastmail.net>
> Cc: users <users at racket-lang.org>
> Subject: Re: [racket] Cyclic data structures in Typed Racket
> Message-ID:
>         <CAK=HD+Y2x3d1e=caisU9xayyFqRvwm+EHT+Rg66yDCpEAhNGtA at mail.gmail.com>
> Content-Type: text/plain; charset=UTF-8
>
> I recommend doing the mutation yourself, and not using `shared`.
> That's the most obvious solution.
>
> The reason that your `require/typed` doesn't work is that it creates
> new versions of the placeholder functions, but those aren't the ones
> that `shared` uses internally, so Typed Racket doesn't use the types
> you've given.
>
> Note also that the types you gave aren't safe -- the way shared uses
> placeholders expects particular types for particular ones (this would
> result in a contract error at runtime if `shared` used your bindings).
>
> Sam
>
> On Mon, Sep 29, 2014 at 12:10 PM, Konrad Hinsen
> <konrad.hinsen at fastmail.net> wrote:
>> Hi everyone,
>>
>> I am trying to port code creating a cyclic data structure to Typed
>> Racket, but notice that this is not as straightforward as I expected.
>>
>> Here is the untyped code I start from:
>>
>>    #lang racket
>>
>>    (struct foo (x) #:mutable)
>>    (struct bar (x))
>>
>>    (define-values (f b)
>>      (shared ([f (foo b)]
>>               [b (bar f)])
>>              (values f b)))
>>
>>    (eq? (bar-x b) f)
>>    (eq? (foo-x f) b)
>>
>> Switching to Typed Racket and annotating the struct definitions, I get
>> an error message saying that make-placeholder and placeholder-set!
>> from racket/base are lacking type annotations. I also get the recommendation
>> to use require/typed for importing them. OK, let's try:
>>
>>    #lang typed/racket
>>
>>    (require/typed racket/base
>>      [#:opaque Placeholder placeholder?]
>>      [make-placeholder (-> Any Placeholder)]
>>      [placeholder-set! (-> Placeholder Any Void)])
>>
>>    (struct foo ([x : bar]) #:mutable)
>>    (struct bar ([x : foo]))
>>
>>    (define-values (f b)
>>      (shared ([f (foo b)]
>>               [b (bar f)])
>>              (values f b)))
>>
>>    (eq? (bar-x b) f)
>>    (eq? (foo-x f) b)
>>
>> Same error message... which is not really surprising. I suppose typed/racket
>> already requires racket/base, so my additional require/typed has no effect.
>> But how can I fix this? I don't see any way to exclude items from typed/racket.
>>
>> Konrad.
>> ____________________
>>   Racket Users list:
>>   http://lists.racket-lang.org/users
>
>
> ------------------------------
>
> Message: 3
> Date: Mon, 29 Sep 2014 18:33:38 +0200
> From: Konrad Hinsen <konrad.hinsen at fastmail.net>
> To: Sam Tobin-Hochstadt <samth at cs.indiana.edu>
> Cc: users <users at racket-lang.org>
> Subject: Re: [racket] Cyclic data structures in Typed Racket
> Message-ID:
>         <21545.35298.582248.45625 at Konrad-Hinsens-MacBook-Pro-2.local>
> Content-Type: text/plain; charset=us-ascii
>
> Sam Tobin-Hochstadt writes:
>
>  > I recommend doing the mutation yourself, and not using `shared`.
>  > That's the most obvious solution.
>
> Not for me, unfortunately, but probably I am missing something
> obvious.
>
> Here's explicit mutation in untyped Racket:
>
>    #lang racket
>
>    (struct foo (x) #:mutable)
>    (struct bar (x))
>
>    (define f (foo (void)))
>    (define b (bar f))
>    (set-foo-x! f b)
>
>    (eq? (bar-x b) f)
>    (eq? (foo-x f) b)
>
> That works fine. Moving to Typed Racket, this is rejected by the type
> checker because (void) is of type Void.  Since I need a bar to make a
> foo and a foo to make a bar, I don't see how I can ever initialize my
> foo structure.
>
>  > The reason that your `require/typed` doesn't work is that it creates
>  > new versions of the placeholder functions, but those aren't the ones
>  > that `shared` uses internally, so Typed Racket doesn't use the types
>  > you've given.
>
> That makes sense, thanks for the explanation!
>
> Konrad.
>
>
> ------------------------------
>
> Message: 4
> Date: Mon, 29 Sep 2014 12:52:25 -0400
> From: Sam Tobin-Hochstadt <samth at cs.indiana.edu>
> To: Konrad Hinsen <konrad.hinsen at fastmail.net>
> Cc: users <users at racket-lang.org>
> Subject: Re: [racket] Cyclic data structures in Typed Racket
> Message-ID:
>         <CAK=HD+Ywfnfv-mjQPpo9GmOO8u+aV7f6ue-4motks_GkVnNRaQ at mail.gmail.com>
> Content-Type: text/plain; charset=UTF-8
>
> Ah, if you want to create truly cyclic structure with no base case
> like this, then you'll have to do more work. Either you'll need to
> expand the type of the `x` field in `foo` to allow an initial value,
> or you'll need to create a dummy `foo` with that extended type, and
> then copy the cyclic data once you've created it to a new version of
> `foo`, using hash tables to track cycles. That's basically how
> `shared` works, although I haven't tried implementing it in TR and it
> might not be possible.
>
> Sam
>
> On Mon, Sep 29, 2014 at 12:33 PM, Konrad Hinsen
> <konrad.hinsen at fastmail.net> wrote:
>> Sam Tobin-Hochstadt writes:
>>
>>  > I recommend doing the mutation yourself, and not using `shared`.
>>  > That's the most obvious solution.
>>
>> Not for me, unfortunately, but probably I am missing something
>> obvious.
>>
>> Here's explicit mutation in untyped Racket:
>>
>>    #lang racket
>>
>>    (struct foo (x) #:mutable)
>>    (struct bar (x))
>>
>>    (define f (foo (void)))
>>    (define b (bar f))
>>    (set-foo-x! f b)
>>
>>    (eq? (bar-x b) f)
>>    (eq? (foo-x f) b)
>>
>> That works fine. Moving to Typed Racket, this is rejected by the type
>> checker because (void) is of type Void.  Since I need a bar to make a
>> foo and a foo to make a bar, I don't see how I can ever initialize my
>> foo structure.
>>
>>  > The reason that your `require/typed` doesn't work is that it creates
>>  > new versions of the placeholder functions, but those aren't the ones
>>  > that `shared` uses internally, so Typed Racket doesn't use the types
>>  > you've given.
>>
>> That makes sense, thanks for the explanation!
>>
>> Konrad.
>
>
> ------------------------------
>
> Message: 5
> Date: Mon, 29 Sep 2014 14:00:12 -0400
> From: Benjamin Greenman <blg59 at cornell.edu>
> To: Sam Tobin-Hochstadt <samth at cs.indiana.edu>
> Cc: users <users at racket-lang.org>
> Subject: Re: [racket] Cyclic data structures in Typed Racket
> Message-ID:
>         <CAAtAoRpHmBjH4FW1BwSOoDh41=m=BZt2q4wz9xdq+Wfj8RsRFQ at mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
>
> To be sure: Is there any TR analog to this OCaml code, initializing a
> cyclic, 3-element list? I've tried using letrec in TR, but the typechecker
> yells at me even if I annotate the a,b,c,d.
>
> # type 'a mylist = Nil | Cons of 'a * 'a mylist;;
> type 'a mylist = Nil | Cons of 'a * 'a mylist
> # let mylist =
>     let rec a = Cons(1, b)
>     and     b = Cons(2, c)
>     and     c = Cons(3, d)
>     and     d = Cons(4, a) in
>     a;;
> val mylist : int mylist = Cons (1, Cons (2, Cons (3, Cons (4, <cycle>))))
>
> On Mon, Sep 29, 2014 at 12:52 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu>
> wrote:
>
>> Ah, if you want to create truly cyclic structure with no base case
>> like this, then you'll have to do more work. Either you'll need to
>> expand the type of the `x` field in `foo` to allow an initial value,
>> or you'll need to create a dummy `foo` with that extended type, and
>> then copy the cyclic data once you've created it to a new version of
>> `foo`, using hash tables to track cycles. That's basically how
>> `shared` works, although I haven't tried implementing it in TR and it
>> might not be possible.
>>
>> Sam
>>
>> On Mon, Sep 29, 2014 at 12:33 PM, Konrad Hinsen
>> <konrad.hinsen at fastmail.net> wrote:
>> > Sam Tobin-Hochstadt writes:
>> >
>> >  > I recommend doing the mutation yourself, and not using `shared`.
>> >  > That's the most obvious solution.
>> >
>> > Not for me, unfortunately, but probably I am missing something
>> > obvious.
>> >
>> > Here's explicit mutation in untyped Racket:
>> >
>> >    #lang racket
>> >
>> >    (struct foo (x) #:mutable)
>> >    (struct bar (x))
>> >
>> >    (define f (foo (void)))
>> >    (define b (bar f))
>> >    (set-foo-x! f b)
>> >
>> >    (eq? (bar-x b) f)
>> >    (eq? (foo-x f) b)
>> >
>> > That works fine. Moving to Typed Racket, this is rejected by the type
>> > checker because (void) is of type Void.  Since I need a bar to make a
>> > foo and a foo to make a bar, I don't see how I can ever initialize my
>> > foo structure.
>> >
>> >  > The reason that your `require/typed` doesn't work is that it creates
>> >  > new versions of the placeholder functions, but those aren't the ones
>> >  > that `shared` uses internally, so Typed Racket doesn't use the types
>> >  > you've given.
>> >
>> > That makes sense, thanks for the explanation!
>> >
>> > Konrad.
>> ____________________
>>   Racket Users list:
>>   http://lists.racket-lang.org/users
>>
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140929/152cdf29/attachment.html>
>
> End of users Digest, Vol 109, Issue 77
> **************************************

From moshedeutsch115 at gmail.com  Mon Sep 29 16:41:00 2014
From: moshedeutsch115 at gmail.com (Moshe Deutsch)
Date: Mon, 29 Sep 2014 16:41:00 -0400
Subject: [racket] users Digest, Vol 109, Issue 76
In-Reply-To: <mailman.1.1412006401.27068.users@racket-lang.org>
References: <mailman.1.1412006401.27068.users@racket-lang.org>
Message-ID: <CAKpSqLAqqAxDc4F-Ei7J0-y7zkhx7Rkz737Ct9Zymap5+=CsHA@mail.gmail.com>

Please take me off the list

Thanks

On Mon, Sep 29, 2014 at 12:00 PM,  <users-request at racket-lang.org> wrote:
> Send users mailing list submissions to
>         users at racket-lang.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>         http://lists.racket-lang.org/users/listinfo
> or, via email, send a message with subject or body 'help' to
>         users-request at racket-lang.org
>
> You can reach the person managing the list at
>         users-owner at racket-lang.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of users digest..."
>
>
> [Racket Users list:
>  http://lists.racket-lang.org/users ]
>
>
> Today's Topics:
>
>    1.  Is there any way to get the bit depth of a bitmap? (rrandom)
>    2. Re: Is there any way to get the bit depth of a bitmap?
>       (Matthew Flatt)
>    3. Re: rsound docs (Matthew Flatt)
>    4. Re: typed racket, filters, and polymorphism (Robby Findler)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Mon, 29 Sep 2014 23:14:43 +0800
> From: rrandom <emanonhere at gmx.com>
> To: Racket Users <users at racket-lang.org>
> Subject: [racket]  Is there any way to get the bit depth of a bitmap?
> Message-ID: <54297763.3020407 at gmx.com>
> Content-Type: text/plain; charset=GB2312
>
> Hi,I am recently working in the book Digital Image Processing with
> racket. I want to know is there any way to get the bit depth or color
> depth of a input bitmap in Racket? I found the get-depth method in
> racket/draw,bitmap%, but it only return 1 and 32, that's not what i
> really want.I input a bitmap whose bit depth is 8 but i got 32.
> following is the code:
>
>> (send (read-bitmap "Fig0305(a)(DFT_no_log).tif")
> get-depth)
> 32
>
> Is there anything I am wrong? Thank you!
>
>
> ------------------------------
>
> Message: 2
> Date: Mon, 29 Sep 2014 09:24:07 -0600
> From: Matthew Flatt <mflatt at cs.utah.edu>
> To: rrandom <emanonhere at gmx.com>
> Cc: Racket Users <users at racket-lang.org>
> Subject: Re: [racket] Is there any way to get the bit depth of a
>         bitmap?
> Message-ID: <20140929152408.F0E42650192 at mail-svr1.cs.utah.edu>
> Content-Type: text/plain; charset=UTF-8
>
> Racket libraries can read and write images of various formats and
> pixels depths, but a `bitmap%` object always represents an image using
> 1 or 32 bits per pixel in memory.
>
> At Mon, 29 Sep 2014 23:14:43 +0800, rrandom wrote:
>> Hi,I am recently working in the book Digital Image Processing with
>> racket. I want to know is there any way to get the bit depth or color
>> depth of a input bitmap in Racket? I found the get-depth method in
>> racket/draw,bitmap%, but it only return 1 and 32, that's not what i
>> really want.I input a bitmap whose bit depth is 8 but i got 32.
>> following is the code:
>>
>> > (send (read-bitmap "Fig0305(a)(DFT_no_log).tif")
>> get-depth)
>> 32
>>
>> Is there anything I am wrong? Thank you!
>> ____________________
>>   Racket Users list:
>>   http://lists.racket-lang.org/users
>
>
> ------------------------------
>
> Message: 3
> Date: Mon, 29 Sep 2014 09:28:14 -0600
> From: Matthew Flatt <mflatt at cs.utah.edu>
> To: C K Kashyap <ckkashyap at gmail.com>
> Cc: users at racket-lang.org
> Subject: Re: [racket] rsound docs
> Message-ID: <20140929152815.C87E6650194 at mail-svr1.cs.utah.edu>
> Content-Type: text/plain; charset=UTF-8
>
> When you install the package, you also install documentation for
> RSound. You should be able to find it by selecting "Racket
> Documentation" from the "Help" menu in DrRacket, by right-clicking on
> an identifier in DrRacket that is provided by the package's library, or
> by running `raco docs` on the command line.
>
> You can also find documentation linked from the "rsound" entry at
> http://pkgs.racket-lang.org/, which goes to
>
>   http://pkg-build.racket-lang.org/doc/rsound/index.html
>
>
> At Mon, 29 Sep 2014 19:09:17 +0530, C K Kashyap wrote:
>> Hi,
>>
>> I saw this nice racketcon video (
>> https://www.youtube.com/watch?v=DkIVzHNjNEA&list=PLXr4KViVC0qI9t3lizitiFJ1cFIeN2
>> Gdh&index=9)
>> on rsound. I've installed the rsound package and have been able to run the
>> example from the video.
>>
>> Is there a place that has mode documentation/tutorial on rsound?
>>
>> Regards,
>> Kashyap
>> ____________________
>>   Racket Users list:
>>   http://lists.racket-lang.org/users
>
>
> ------------------------------
>
> Message: 4
> Date: Mon, 29 Sep 2014 10:40:08 -0500
> From: Robby Findler <robby at eecs.northwestern.edu>
> To: Konrad Hinsen <konrad.hinsen at fastmail.net>
> Cc: racket users list <users at racket-lang.org>
> Subject: Re: [racket] typed racket, filters, and polymorphism
> Message-ID:
>         <CAL3TdOPrHreseE2RxJW6x5jm5kSnKVQkjvAt-A8RuiKaarCZZw at mail.gmail.com>
> Content-Type: text/plain; charset=UTF-8
>
> Whoops, sorry: this is the same thing Konrad was talking about! Duh.
> Well, hopefully the link is useful.
>
> Robby
>
>
> On Mon, Sep 29, 2014 at 9:32 AM, Robby Findler
> <robby at eecs.northwestern.edu> wrote:
>> Just in case, here is one piece of related work from the PL world and
>> is probably a reasonable starting point for finding others' attempts
>> at this problem:
>>
>> http://research.microsoft.com/en-us/um/people/akenn/units/
>>
>> Robby
>>
>>
>> On Mon, Sep 29, 2014 at 3:44 AM, Konrad Hinsen
>> <konrad.hinsen at fastmail.net> wrote:
>>> Matthias Felleisen writes:
>>>
>>>  > Your message points out where gradual typing badly fails.
>>>
>>> Not just gradual typing. I am not aware of any good unit
>>> implementation in any type system, with the exception of F# and Frink
>>> whose type systems were explicitly modified to add unit checking as a
>>> special case. Dependent types should permit a good solution, but I
>>> haven't seen it done yet.
>>>
>>> The reason you need dependent types is that the product of two
>>> measures is another measure with a new unit, so you must be able to
>>> construct new types (representing units) from values. For example, a
>>> length divided by a time yields a speed. Once you can do that, scaling
>>> a unit (e.g. from meters to kilometers) becomes a special case.
>>>
>>>  > In this specific case, you have two aspects of dimensionality:
>>>  > dimension per se (I am sure there is a word for it) and the chosen
>>>  > unit to represent it.
>>>
>>> The two terms used for this distinction are in fact "dimension" and
>>> "unit".
>>>
>>>  > If someone writes (area-of-rectangle 1 [mm] 1 [km]), there is
>>>  > nothing wrong with it -- except that the type checker should insert
>>>  > a coercion from mm to m and from km to m (multiplying/dividing by
>>>  > 1,000).
>>>
>>> That's a topic of hot debate in the scientific computing community.
>>> Some argue that the type checker should consider your example as an
>>> error, and not do any implicit conversion. The motivation is that the
>>> expression you used is more likely to be a mistake than the meaningful
>>> use of expressive language, since best practices recommend to use a
>>> minimal set of units inside any piece of code.
>>>
>>> Anyway, that's a minor detail. If your type system can handle
>>> automatic conversion, then it can also handle its absence.
>>>
>>>
>>> Alexander D. Knauth writes:
>>>
>>>  > What I had in mind was for the structs to be available at run-time,
>>>  > but that ideally the optimizer could take them out for the
>>>  > intermediate operations and put them back for the final result.
>>>
>>> You are actually adding a very Racket-specific requirement to the
>>> already difficult units-as-types problem: the interplay between a
>>> typed and an untyped dialect of the same language. I agree this would
>>> be nice to have.
>>>
>>> Konrad.
>>> ____________________
>>>   Racket Users list:
>>>   http://lists.racket-lang.org/users
>
>
> End of users Digest, Vol 109, Issue 76
> **************************************

From moshedeutsch115 at gmail.com  Mon Sep 29 16:41:10 2014
From: moshedeutsch115 at gmail.com (Moshe Deutsch)
Date: Mon, 29 Sep 2014 16:41:10 -0400
Subject: [racket] users Digest, Vol 109, Issue 75
In-Reply-To: <mailman.490.1412001171.679.users@racket-lang.org>
References: <mailman.490.1412001171.679.users@racket-lang.org>
Message-ID: <CAKpSqLCuHFKUgyDGGHirVfL383DjvJB3DZY890eop6w2PJKCFw@mail.gmail.com>

Please take me off the list

Thanks

On Mon, Sep 29, 2014 at 10:32 AM,  <users-request at racket-lang.org> wrote:
> Send users mailing list submissions to
>         users at racket-lang.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>         http://lists.racket-lang.org/users/listinfo
> or, via email, send a message with subject or body 'help' to
>         users-request at racket-lang.org
>
> You can reach the person managing the list at
>         users-owner at racket-lang.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of users digest..."
>
>
> [Racket Users list:
>  http://lists.racket-lang.org/users ]
>
>
> Today's Topics:
>
>    1. Re: canonical index of Racket courses? [was: Summer programs
>       learning Racket for a student] (jab at math.brown.edu)
>    2. Re: typed racket, filters, and polymorphism (Konrad Hinsen)
>    3. Re: aws/glacier: credential scope (Greg Hendershott)
>    4. rsound docs (C K Kashyap)
>    5. Re: typed racket, filters, and polymorphism (Robby Findler)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Sun, 28 Sep 2014 23:36:05 -0400
> From: jab at math.brown.edu
> To: Sam Tobin-Hochstadt <samth at cs.indiana.edu>, Matthias Felleisen
>         <matthias at ccs.neu.edu>
> Cc: Racket Users <users at racket-lang.org>
> Subject: Re: [racket] canonical index of Racket courses? [was: Summer
>         programs learning Racket for a student]
> Message-ID:
>         <CANwREeXtCqExVYCchvg4pvqFWs06tb4b=jtmU5zpdwTsAaM4Hg at mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
>
> On Wed, Sep 17, 2014 at 10:51 AM, Sam Tobin-Hochstadt <samth at cs.indiana.edu>
> wrote:
>
>> I've now created a wiki page for this, with some initial content:
>> https://github.com/plt/racket/wiki/Courses-using-Racket
>>
>
> And now it's up to 22 revisions! Thanks for creating, Sam, and to everyone
> who added to it.
>
> On Tue, Sep 16, 2014 at 8:43 PM, Matthias Felleisen
>> <matthias at ccs.neu.edu> wrote:
>> > Do you imagine listing college courses such as Brown's 17, which uses
>> > DrRacket and the teaching languages? Or do you want Coursera courses that
>> > everyone can access?
>>
>
> I think linking to any class that makes educational materials publicly
> available is valuable, but they should be divided into one section for
> classes that don't require being admitted into a larger institution to
> participate in (e.g. Bootstrap World, Coursera, summer programs for teens,
> etc.), and another section for those that do.
>
>
>> > And yes, we should probably create a wiki like thing for such an effort
>> or
>> > perhaps something like packages.racket-lang.org. I'll bring it up as we
>> meet
>> > in St Louis.
>>
>
> Thanks, Matthias, and hope you all had a great time in St Louis. Looks like
> it was a wonderful conference.
>
> Can the wiki page above be directly linked off racket-lang.org, perhaps a
> bit more prominently? I hadn't even noticed
> https://github.com/plt/racket/wiki before; now I see the link to it kind of
> buried under the Contributing column of the Community section. Perhaps it's
> not easy to find for others too.
>
> Thanks again for following up on this.
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140928/ac267b87/attachment-0001.html>
>
> ------------------------------
>
> Message: 2
> Date: Mon, 29 Sep 2014 10:44:56 +0200
> From: Konrad Hinsen <konrad.hinsen at fastmail.net>
> To: racket users list <users at racket-lang.org>
> Subject: Re: [racket] typed racket, filters, and polymorphism
> Message-ID:
>         <21545.7176.252047.613404 at Ordinateur-de-Catherine-Konrad.local>
> Content-Type: text/plain; charset=us-ascii
>
> Matthias Felleisen writes:
>
>  > Your message points out where gradual typing badly fails.
>
> Not just gradual typing. I am not aware of any good unit
> implementation in any type system, with the exception of F# and Frink
> whose type systems were explicitly modified to add unit checking as a
> special case. Dependent types should permit a good solution, but I
> haven't seen it done yet.
>
> The reason you need dependent types is that the product of two
> measures is another measure with a new unit, so you must be able to
> construct new types (representing units) from values. For example, a
> length divided by a time yields a speed. Once you can do that, scaling
> a unit (e.g. from meters to kilometers) becomes a special case.
>
>  > In this specific case, you have two aspects of dimensionality:
>  > dimension per se (I am sure there is a word for it) and the chosen
>  > unit to represent it.
>
> The two terms used for this distinction are in fact "dimension" and
> "unit".
>
>  > If someone writes (area-of-rectangle 1 [mm] 1 [km]), there is
>  > nothing wrong with it -- except that the type checker should insert
>  > a coercion from mm to m and from km to m (multiplying/dividing by
>  > 1,000).
>
> That's a topic of hot debate in the scientific computing community.
> Some argue that the type checker should consider your example as an
> error, and not do any implicit conversion. The motivation is that the
> expression you used is more likely to be a mistake than the meaningful
> use of expressive language, since best practices recommend to use a
> minimal set of units inside any piece of code.
>
> Anyway, that's a minor detail. If your type system can handle
> automatic conversion, then it can also handle its absence.
>
>
> Alexander D. Knauth writes:
>
>  > What I had in mind was for the structs to be available at run-time,
>  > but that ideally the optimizer could take them out for the
>  > intermediate operations and put them back for the final result.
>
> You are actually adding a very Racket-specific requirement to the
> already difficult units-as-types problem: the interplay between a
> typed and an untyped dialect of the same language. I agree this would
> be nice to have.
>
> Konrad.
>
>
> ------------------------------
>
> Message: 3
> Date: Mon, 29 Sep 2014 09:13:11 -0400
> From: Greg Hendershott <greghendershott at gmail.com>
> To: Norman Gray <norman at astro.gla.ac.uk>
> Cc: users at racket-lang.org
> Subject: Re: [racket] aws/glacier: credential scope
> Message-ID:
>         <CAGspUn1Q5MvFq4i6q3KuoxBcszOpTh-RdjL=QWy+JbwHUUrXSA at mail.gmail.com>
> Content-Type: text/plain; charset=UTF-8
>
> It looks like the package server did eventually refresh a few hours ago:
>
>     Last Updated:  9/28/2014, 4:21:56 PM  ;change on GitHub
>     Last Checked:  9/29/2014, 7:51:37 AM  ;pkgs.r-l.org refresh
>
> (I believe those times are US MT, -0600.)
>
> On Sun, Sep 28, 2014 at 6:10 PM, Greg Hendershott
> <greghendershott at gmail.com> wrote:
>> If you're in a hurry you could remove and re-install directly from GitHub:
>>
>>  $ raco pkg remove aws
>>  $ raco pkg install git://github.com/greghendershott/aws
>>
>> On Sun, Sep 28, 2014 at 5:40 PM, Norman Gray <norman at astro.gla.ac.uk> wrote:
>>>
>>> On 2014 Sep 28, at 22:17, Greg Hendershott <greghendershott at gmail.com> wrote:
>>>
>>>>>> I clicked "update" on pkgs.racket-lang.org. But it seems to be slower
>>>>>> than usual to refresh. After it does, you can `raco pkg update aws` to
>>>>>> get the fix.
>>>>>
>>>>> Something seems wrong/stuck on pkgs.racket-lang.org. The top of the page says:
>>>>>
>>>>>  "update upload in progress: there may be inconsistencies below"
>>>>>
>>>>> It's been saying that for nearly an hour.
>>>>
>>>> Sometimes instead it says:
>>>>
>>>>  "update upload being computed: the information below may not
>>>> represent all recent changes and updates"
>>>>
>>>> It has switched between those two messages at various intervals over
>>>> the last hour.
>>>
>>> And in case it's a useful data point, I'm getting
>>>
>>> % raco pkg update aws
>>> Resolving "aws" via http://download.racket-lang.org/releases/6.1/catalog/
>>> Resolving "aws" via http://pkgs.racket-lang.org
>>> No updates available
>>> %
>>>
>>> And Greg, your fix is close enough to what I guessed might be the problem that I'm now kicking myself for not just diving in and trying a fix myself.  Garhhhh...
>>>
>>> Norman
>>>
>>>
>>> --
>>> Norman Gray  :  http://nxg.me.uk
>>> SUPA School of Physics and Astronomy, University of Glasgow, UK
>>>
>
>
> ------------------------------
>
> Message: 4
> Date: Mon, 29 Sep 2014 19:09:17 +0530
> From: C K Kashyap <ckkashyap at gmail.com>
> To: users at racket-lang.org
> Subject: [racket] rsound docs
> Message-ID:
>         <CAGdT1gq=E_-O9JcnaMnOEQkEK51vWxdPsFHHNSudzRoJCwHm2g at mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
>
> Hi,
>
> I saw this nice racketcon video (
> https://www.youtube.com/watch?v=DkIVzHNjNEA&list=PLXr4KViVC0qI9t3lizitiFJ1cFIeN2Gdh&index=9)
> on rsound. I've installed the rsound package and have been able to run the
> example from the video.
>
> Is there a place that has mode documentation/tutorial on rsound?
>
> Regards,
> Kashyap
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140929/7ce26b08/attachment-0001.html>
>
> ------------------------------
>
> Message: 5
> Date: Mon, 29 Sep 2014 09:32:40 -0500
> From: Robby Findler <robby at eecs.northwestern.edu>
> To: Konrad Hinsen <konrad.hinsen at fastmail.net>
> Cc: racket users list <users at racket-lang.org>
> Subject: Re: [racket] typed racket, filters, and polymorphism
> Message-ID:
>         <CAL3TdONRKwUJ0pV3WhXDQvRgMT6mhSp6D6QGr4=W4y7g7DsK4w at mail.gmail.com>
> Content-Type: text/plain; charset=UTF-8
>
> Just in case, here is one piece of related work from the PL world and
> is probably a reasonable starting point for finding others' attempts
> at this problem:
>
> http://research.microsoft.com/en-us/um/people/akenn/units/
>
> Robby
>
>
> On Mon, Sep 29, 2014 at 3:44 AM, Konrad Hinsen
> <konrad.hinsen at fastmail.net> wrote:
>> Matthias Felleisen writes:
>>
>>  > Your message points out where gradual typing badly fails.
>>
>> Not just gradual typing. I am not aware of any good unit
>> implementation in any type system, with the exception of F# and Frink
>> whose type systems were explicitly modified to add unit checking as a
>> special case. Dependent types should permit a good solution, but I
>> haven't seen it done yet.
>>
>> The reason you need dependent types is that the product of two
>> measures is another measure with a new unit, so you must be able to
>> construct new types (representing units) from values. For example, a
>> length divided by a time yields a speed. Once you can do that, scaling
>> a unit (e.g. from meters to kilometers) becomes a special case.
>>
>>  > In this specific case, you have two aspects of dimensionality:
>>  > dimension per se (I am sure there is a word for it) and the chosen
>>  > unit to represent it.
>>
>> The two terms used for this distinction are in fact "dimension" and
>> "unit".
>>
>>  > If someone writes (area-of-rectangle 1 [mm] 1 [km]), there is
>>  > nothing wrong with it -- except that the type checker should insert
>>  > a coercion from mm to m and from km to m (multiplying/dividing by
>>  > 1,000).
>>
>> That's a topic of hot debate in the scientific computing community.
>> Some argue that the type checker should consider your example as an
>> error, and not do any implicit conversion. The motivation is that the
>> expression you used is more likely to be a mistake than the meaningful
>> use of expressive language, since best practices recommend to use a
>> minimal set of units inside any piece of code.
>>
>> Anyway, that's a minor detail. If your type system can handle
>> automatic conversion, then it can also handle its absence.
>>
>>
>> Alexander D. Knauth writes:
>>
>>  > What I had in mind was for the structs to be available at run-time,
>>  > but that ideally the optimizer could take them out for the
>>  > intermediate operations and put them back for the final result.
>>
>> You are actually adding a very Racket-specific requirement to the
>> already difficult units-as-types problem: the interplay between a
>> typed and an untyped dialect of the same language. I agree this would
>> be nice to have.
>>
>> Konrad.
>> ____________________
>>   Racket Users list:
>>   http://lists.racket-lang.org/users
>
>
> End of users Digest, Vol 109, Issue 75
> **************************************

From moshedeutsch115 at gmail.com  Mon Sep 29 16:41:19 2014
From: moshedeutsch115 at gmail.com (Moshe Deutsch)
Date: Mon, 29 Sep 2014 16:41:19 -0400
Subject: [racket] users Digest, Vol 109, Issue 74
In-Reply-To: <mailman.487.1411959631.679.users@racket-lang.org>
References: <mailman.487.1411959631.679.users@racket-lang.org>
Message-ID: <CAKpSqLAx_CUR+OJW4yreyoPq_d4sSGcMKDy9hydKGoGbDyepCw@mail.gmail.com>

Please take me off the list

Thanks

On Sun, Sep 28, 2014 at 11:00 PM,  <users-request at racket-lang.org> wrote:
> Send users mailing list submissions to
>         users at racket-lang.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>         http://lists.racket-lang.org/users/listinfo
> or, via email, send a message with subject or body 'help' to
>         users-request at racket-lang.org
>
> You can reach the person managing the list at
>         users-owner at racket-lang.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of users digest..."
>
>
> [Racket Users list:
>  http://lists.racket-lang.org/users ]
>
>
> Today's Topics:
>
>    1. Re: typed racket, filters, and polymorphism (Alexander D. Knauth)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Sun, 28 Sep 2014 23:00:17 -0400
> From: "Alexander D. Knauth" <alexander at knauth.org>
> To: racket users list <users at racket-lang.org>
> Subject: Re: [racket] typed racket, filters, and polymorphism
> Message-ID: <0BDCBA09-5AA1-410A-A954-D2B6A8B30465 at knauth.org>
> Content-Type: text/plain; charset="utf-8"
>
> I just realized that typed racket seems to ignore the guard anyway, so I could do something like this and convince the type-checker that any value I want has any type that I want:
>
> (struct (a) foo ([a : a])
>   #:guard (lambda (a _) "this value"))
>
> (: x : (foo 1))
> (define x (foo 1))
>
>> (foo-a x)
> - : Integer [more precisely: One]
> "this value?
>
>
> On Sep 28, 2014, at 10:41 PM, Benjamin Greenman <blg59 at cornell.edu> wrote:
>
>> I was able to get something very similar to your example program to compile.
>>
>> (struct (a) foo ([a : a]) #:transparent
>>         #:guard (lambda (a _)
>>                   (define a* a)  ;; a* not "polluted" by occurrence typing
>>                   (unless (exact-integer? a)
>>                     (error 'foo "expected Integer, given ~v" a))
>>                   a*))  ;; we return a value of type a, so the typechecker is happy
>>
>> From here, you can safely cast a foo-a into an Integer. It's not type safe to ann because we (deliberately) don't have the occurrence typing information, but we the programmers can be sure that the casts will never fail.
>>
>> (cast (foo (ann 1 Any)) (foo Integer))
>> > (foo 1)
>>
>>
>>
>> On Sun, Sep 28, 2014 at 10:29 PM, Matthias Felleisen <matthias at ccs.neu.edu> wrote:
>>
>> You might be able to unwrap such things for a local context, but then you need to wrap it again on the way out. That cost might become quite high if it happens in a loop. Then again Racket programs rely on this a lot too (as do many dynamically typed programs) so it might not be a cost that kills performance completely. But do be aware of how others deal with this issue -- Matthias
>>
>>
>>
>>
>>
>>
>>
>> On Sep 28, 2014, at 10:03 PM, Alexander D. Knauth wrote:
>>
>>>
>>>
>>> On Sep 28, 2014, at 8:25 PM, Matthias Felleisen <matthias at ccs.neu.edu> wrote:
>>>
>>>>
>>>> Your message points out where gradual typing badly fails.
>>>>
>>>> Type systems combine two nearly orthogonal ideas: (1) proving theorems about your program and (2) bridging the gap between abstract linguistic constructs and concrete machine implementations. In a sense, the design of a type system should always have the goal to eliminate as much run-time overhead as possible.
>>>>
>>>> In this specific case, you have two aspects of dimensionality: dimension per se (I am sure there is a word for it) and the chosen unit to represent it. So if you know that a number denotes a distance, you can choose mm, m, and km to represent it. If you want to represent an area, you choose mm^2, m^2, km^2. Now if you want to compute the area of a rectangle, you write
>>>>
>>>>  area-of-rectangle :  Real [Distance in m] * Real [Distance in m] -> Real [Area in m^2]
>>>>
>>>> If someone writes (area-of-rectangle 1 [mm] 1 [km]), there is nothing wrong with it -- except that the type checker should insert a coercion from mm to m and from km to m (multiplying/dividing by 1,000). That is, a type system for this application ought to use the type elaboration process idea and translate code as it goes.
>>>>
>>>> This is quite comparable to the idea that a machine-oriented type system needs to insert a coercion from integer to real (for example) as Algol 60 and its degraded successors do.
>>>
>>> What I had in mind was for the structs to be available at run-time, but that ideally the optimizer could take them out for the intermediate operations and put them back for the final result.
>>>
>>> And then if this value goes into untyped code, then it knows at run-time that it has the unit (u* millimeter kilometer), which is unit=? to square-meter.
>>>
>>> Is this sort of like what happens with flonum unboxing?
>>>
>>>
>>>
>>>>
>>>> Our gradual type system lacks some (intensiona) expressive power, which is why you can't get close to this ideal.
>>>>
>>>> -- Matthias
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>> On Sep 28, 2014, at 8:14 PM, Alexander D. Knauth wrote:
>>>>
>>>>> Yes I do want run-time residual of the dimensions, and yes I want a unit struct.
>>>>>
>>>>> Also if there?s run-time residual, then I can use predicates and occurrence typing, and it can also be used in untyped code with contracts instead of types.
>>>>>
>>>>> And also the type-checker would check if it is dimension-correct, but at run-time it would would still convert between different units of the same dimension, multiply and divide units, etc.
>>>>>
>>>>> On Sep 28, 2014, at 6:58 PM, Matthias Felleisen <matthias at ccs.neu.edu> wrote:
>>>>>
>>>>>>
>>>>>> If you want the type checker to ensure your program is unit-correct, I assume you also want no run-time residual of the dimensions but that is in conflict with wanting a structure because it imposes a run-time cost. Are you sure you want a unit? -- Matthias
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>> On Sep 28, 2014, at 1:37 PM, Alexander D. Knauth wrote:
>>>>>>
>>>>>>> No because I want the unit to be a struct that has a dimension field, not a symbol with various dimensions defined as unions of units.
>>>>>>> I want the unit to be based on the dimension, not the other way around, so that new units can be made that have the same dimension.
>>>>>>>
>>>>>>> I have something like the number+unit struct (I called it measure), but I?ll work on more that after I have the unit struct figured out.
>>>>>>>
>>>>>>> On Sep 28, 2014, at 12:13 PM, Spencer florence <spencerflorence at gmail.com> wrote:
>>>>>>>
>>>>>>>> would something like this work?
>>>>>>>>
>>>>>>>> #lang typed/racket
>>>>>>>>
>>>>>>>> (struct (U) number+unit ([amount : Real] [unit : U]))
>>>>>>>>
>>>>>>>> (define-type Weight-Unit (U 'kg 'g 'mg '?g))
>>>>>>>> (define-type Weight (number+unit Weight-Unit))
>>>>>>>> (define-predicate weight? Weight)
>>>>>>>>
>>>>>>>> (: make-weight : Real Weight-Unit -> Weight)
>>>>>>>> (define (make-weight n u)
>>>>>>>>   (number+unit n u))
>>>>>>>>
>>>>>>>> (: +/weight : Weight Weight -> Weight)
>>>>>>>> ;; something something needs unit conversion
>>>>>>>> (define (+/weight w1 w2)
>>>>>>>>   (number+unit (+ (number+unit-amount w1)
>>>>>>>>                   (number+unit-amount w1))
>>>>>>>>                (number+unit-unit w1)))
>>>>>>>>
>>>>>>>> (+/weight (make-weight 1 'kg) (make-weight 1 'kg))
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> On Sun, Sep 28, 2014 at 11:03 AM, Alexander D. Knauth <alexander at knauth.org> wrote:
>>>>>>>>
>>>>>>>> Because the struct is representing a unit (kilograms, meters, seconds, etc.), and a unit has a dimension (mass, length, time, etc.) and I want the type-checker to be able to know what the dimension of a unit is so that the types of functions can specify the dimension that something should have.
>>>>>>>> The real solution to this would probably be bounded polymorphism, but I was wondering if there was some other way to do it with occurrence typing in the guard or something like that.
>>>>>>>>
>>>>>>>> On Sep 28, 2014, at 11:48 AM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> wrote:
>>>>>>>>
>>>>>>>> > Why not do this with the type, instead of making this polymorphic?
>>>>>>>> >
>>>>>>>> > Sam
>>>>>>>> >
>>>>>>>> > On Fri, Sep 26, 2014 at 7:35 PM, Alexander D. Knauth
>>>>>>>> > <alexander at knauth.org> wrote:
>>>>>>>> >> Is it possible to have a struct that does certain things according to the
>>>>>>>> >> guard?
>>>>>>>> >> #lang typed/racket
>>>>>>>> >>
>>>>>>>> >> (struct (a) foo ([a : a]) #:transparent
>>>>>>>> >> #:guard (lambda (a _)
>>>>>>>> >> (unless (exact-integer? a)
>>>>>>>> >> (error 'foo "expected Integer, given ~v" a))
>>>>>>>> >> a))
>>>>>>>> >>
>>>>>>>> >> (ann (foo (ann 1 Any)) (foo Integer))
>>>>>>>> >>
>>>>>>>> >> (: x : (foo Any))
>>>>>>>> >> (define x (foo 1))
>>>>>>>> >>
>>>>>>>> >> (ann (foo-a x) Integer)
>>>>>>>> >>
>>>>>>>> >> ;. Type Checker: Polymorphic function `foo1' could not be applied to
>>>>>>>> >> arguments:
>>>>>>>> >> ;Argument 1:
>>>>>>>> >> ; Expected: a
>>>>>>>> >> ; Given: Any
>>>>>>>> >> ;
>>>>>>>> >> ;Result type: (foo a)
>>>>>>>> >> ;Expected result: (foo Integer)
>>>>>>>> >> ; in: (foo (ann 1 Any))
>>>>>>>> >> ;. Type Checker: Polymorphic function `foo-a' could not be applied to
>>>>>>>> >> arguments:
>>>>>>>> >> ;Argument 1:
>>>>>>>> >> ; Expected: (foo a)
>>>>>>>> >> ; Given: (foo Any)
>>>>>>>> >> ;
>>>>>>>> >> ;Result type: (a : ....)
>>>>>>>> >> ;Expected result: Integer
>>>>>>>> >> ; in: (foo-a x)
>>>>>>>> >>
>>>>>>>> >> On Sep 25, 2014, at 9:42 PM, Alexander D. Knauth <alexander at knauth.org>
>>>>>>>> >> wrote:
>>>>>>>> >>
>>>>>>>> >> What I?m trying to accomplish is something more like this:
>>>>>>>> >> #lang typed/racket
>>>>>>>> >>
>>>>>>>> >> (require "dimensions.rkt")
>>>>>>>> >>
>>>>>>>> >> (struct (d) unit ([name : Any] [scalar : Positive-Real] [dimension : d])
>>>>>>>> >> #:transparent
>>>>>>>> >> #:guard (lambda (name scalar dimension _)
>>>>>>>> >> (unless (dimension? dimension)
>>>>>>>> >> (error 'unit "expected Dimension, given ~v" dimension))
>>>>>>>> >> (values name scalar dimension)))
>>>>>>>> >>
>>>>>>>> >> (define-type (Unitof d) (unit d))
>>>>>>>> >>
>>>>>>>> >> (define-type Unit (Unitof Dimension))
>>>>>>>> >>
>>>>>>>> >> (define Unit? (make-predicate Unit))
>>>>>>>> >>
>>>>>>>> >> (define-type Unitish
>>>>>>>> >> (U (Unitof Any)
>>>>>>>> >> Dimension
>>>>>>>> >> Positive-Real))
>>>>>>>> >>
>>>>>>>> >> (: ->unit : (All (d) (case-> [(Unitof d) -> (Unitof d)]
>>>>>>>> >> [Unitish -> Unit])))
>>>>>>>> >> (define (->unit u)
>>>>>>>> >> (cond [(unit? u)
>>>>>>>> >> (unless (Unit? u) ; this should never happen anyway because of the
>>>>>>>> >> guard
>>>>>>>> >> (error '->unit "expected (Unitof Dimension), given ~v" u))
>>>>>>>> >> u]
>>>>>>>> >> [(dimension? u) (unit u 1 u)]
>>>>>>>> >> [(positive-real? u) (unit u u dimensionless-dimension)]))
>>>>>>>> >>
>>>>>>>> >>
>>>>>>>> >> On Sep 25, 2014, at 6:19 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu>
>>>>>>>> >> wrote:
>>>>>>>> >>
>>>>>>>> >> No, I don't think you can do this. Can you say more about what you're
>>>>>>>> >> trying to accomplish?
>>>>>>>> >>
>>>>>>>> >> Sam
>>>>>>>> >>
>>>>>>>> >> On Thu, Sep 25, 2014 at 6:15 PM, Alexander D. Knauth
>>>>>>>> >> <alexander at knauth.org> wrote:
>>>>>>>> >>
>>>>>>>> >> Do any of you have any advice for getting a function like this to
>>>>>>>> >> type-check?
>>>>>>>> >> #lang typed/racket
>>>>>>>> >>
>>>>>>>> >> (: check-int : (All (a) (case-> [a -> a]
>>>>>>>> >> [Any -> Integer])))
>>>>>>>> >> (define (check-int int)
>>>>>>>> >> (unless (exact-integer? int)
>>>>>>>> >> (error 'check-int "expected Integer, given ~v" int))
>>>>>>>> >> int)
>>>>>>>> >>
>>>>>>>> >> ;. Type Checker: type mismatch
>>>>>>>> >> ; expected: a
>>>>>>>> >> ; given: Integer in: int
>>>>>>>> >>
>>>>>>>> >>
>>>>>>>> >>
>>>>>>>> >> ____________________
>>>>>>>> >> Racket Users list:
>>>>>>>> >> http://lists.racket-lang.org/users
>>>>>>>> >>
>>>>>>>> >>
>>>>>>>> >> ____________________
>>>>>>>> >> Racket Users list:
>>>>>>>> >> http://lists.racket-lang.org/users
>>>>>>>> >>
>>>>>>>> >>
>>>>>>>>
>>>>>>>>
>>>>>>>> ____________________
>>>>>>>> Racket Users list:
>>>>>>>> http://lists.racket-lang.org/users
>>>>>>>>
>>>>>>>>
>>>>>>>
>>>>>>> ____________________
>>>>>>>  Racket Users list:
>>>>>>>  http://lists.racket-lang.org/users
>>>>>>
>>>>>
>>>>
>>>
>>
>>
>
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140928/baf5ed5d/attachment.html>
>
> End of users Digest, Vol 109, Issue 74
> **************************************


From moshedeutsch115 at gmail.com  Mon Sep 29 16:41:29 2014
From: moshedeutsch115 at gmail.com (Moshe Deutsch)
Date: Mon, 29 Sep 2014 16:41:29 -0400
Subject: [racket] users Digest, Vol 109, Issue 73
In-Reply-To: <mailman.485.1411957788.679.users@racket-lang.org>
References: <mailman.485.1411957788.679.users@racket-lang.org>
Message-ID: <CAKpSqLDpc4sO0or7Q88yb7uBOUffQYb8s_PpzaRB0yaDhTD+gA@mail.gmail.com>

Please take me off the list

Thanks

On Sun, Sep 28, 2014 at 10:29 PM,  <users-request at racket-lang.org> wrote:
> Send users mailing list submissions to
>         users at racket-lang.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>         http://lists.racket-lang.org/users/listinfo
> or, via email, send a message with subject or body 'help' to
>         users-request at racket-lang.org
>
> You can reach the person managing the list at
>         users-owner at racket-lang.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of users digest..."
>
>
> [Racket Users list:
>  http://lists.racket-lang.org/users ]
>
>
> Today's Topics:
>
>    1. Re: typed racket, filters, and polymorphism (Alexander D. Knauth)
>    2. Re: typed racket, filters, and polymorphism (Matthias Felleisen)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Sun, 28 Sep 2014 22:03:02 -0400
> From: "Alexander D. Knauth" <alexander at knauth.org>
> To: Matthias Felleisen <matthias at ccs.neu.edu>
> Cc: racket users list <users at racket-lang.org>
> Subject: Re: [racket] typed racket, filters, and polymorphism
> Message-ID: <6237F562-9E68-4D2D-A307-49668BB8E5DF at knauth.org>
> Content-Type: text/plain; charset="utf-8"
>
>
>
> On Sep 28, 2014, at 8:25 PM, Matthias Felleisen <matthias at ccs.neu.edu> wrote:
>
>>
>> Your message points out where gradual typing badly fails.
>>
>> Type systems combine two nearly orthogonal ideas: (1) proving theorems about your program and (2) bridging the gap between abstract linguistic constructs and concrete machine implementations. In a sense, the design of a type system should always have the goal to eliminate as much run-time overhead as possible.
>>
>> In this specific case, you have two aspects of dimensionality: dimension per se (I am sure there is a word for it) and the chosen unit to represent it. So if you know that a number denotes a distance, you can choose mm, m, and km to represent it. If you want to represent an area, you choose mm^2, m^2, km^2. Now if you want to compute the area of a rectangle, you write
>>
>>  area-of-rectangle :  Real [Distance in m] * Real [Distance in m] -> Real [Area in m^2]
>>
>> If someone writes (area-of-rectangle 1 [mm] 1 [km]), there is nothing wrong with it -- except that the type checker should insert a coercion from mm to m and from km to m (multiplying/dividing by 1,000). That is, a type system for this application ought to use the type elaboration process idea and translate code as it goes.
>>
>> This is quite comparable to the idea that a machine-oriented type system needs to insert a coercion from integer to real (for example) as Algol 60 and its degraded successors do.
>
> What I had in mind was for the structs to be available at run-time, but that ideally the optimizer could take them out for the intermediate operations and put them back for the final result.
>
> And then if this value goes into untyped code, then it knows at run-time that it has the unit (u* millimeter kilometer), which is unit=? to square-meter.
>
> Is this sort of like what happens with flonum unboxing?
>
>
>
>>
>> Our gradual type system lacks some (intensiona) expressive power, which is why you can't get close to this ideal.
>>
>> -- Matthias
>>
>>
>>
>>
>>
>>
>>
>>
>> On Sep 28, 2014, at 8:14 PM, Alexander D. Knauth wrote:
>>
>>> Yes I do want run-time residual of the dimensions, and yes I want a unit struct.
>>>
>>> Also if there?s run-time residual, then I can use predicates and occurrence typing, and it can also be used in untyped code with contracts instead of types.
>>>
>>> And also the type-checker would check if it is dimension-correct, but at run-time it would would still convert between different units of the same dimension, multiply and divide units, etc.
>>>
>>> On Sep 28, 2014, at 6:58 PM, Matthias Felleisen <matthias at ccs.neu.edu> wrote:
>>>
>>>>
>>>> If you want the type checker to ensure your program is unit-correct, I assume you also want no run-time residual of the dimensions but that is in conflict with wanting a structure because it imposes a run-time cost. Are you sure you want a unit? -- Matthias
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>> On Sep 28, 2014, at 1:37 PM, Alexander D. Knauth wrote:
>>>>
>>>>> No because I want the unit to be a struct that has a dimension field, not a symbol with various dimensions defined as unions of units.
>>>>> I want the unit to be based on the dimension, not the other way around, so that new units can be made that have the same dimension.
>>>>>
>>>>> I have something like the number+unit struct (I called it measure), but I?ll work on more that after I have the unit struct figured out.
>>>>>
>>>>> On Sep 28, 2014, at 12:13 PM, Spencer florence <spencerflorence at gmail.com> wrote:
>>>>>
>>>>>> would something like this work?
>>>>>>
>>>>>> #lang typed/racket
>>>>>>
>>>>>> (struct (U) number+unit ([amount : Real] [unit : U]))
>>>>>>
>>>>>> (define-type Weight-Unit (U 'kg 'g 'mg '?g))
>>>>>> (define-type Weight (number+unit Weight-Unit))
>>>>>> (define-predicate weight? Weight)
>>>>>>
>>>>>> (: make-weight : Real Weight-Unit -> Weight)
>>>>>> (define (make-weight n u)
>>>>>>   (number+unit n u))
>>>>>>
>>>>>> (: +/weight : Weight Weight -> Weight)
>>>>>> ;; something something needs unit conversion
>>>>>> (define (+/weight w1 w2)
>>>>>>   (number+unit (+ (number+unit-amount w1)
>>>>>>                   (number+unit-amount w1))
>>>>>>                (number+unit-unit w1)))
>>>>>>
>>>>>> (+/weight (make-weight 1 'kg) (make-weight 1 'kg))
>>>>>>
>>>>>>
>>>>>>
>>>>>> On Sun, Sep 28, 2014 at 11:03 AM, Alexander D. Knauth <alexander at knauth.org> wrote:
>>>>>>
>>>>>> Because the struct is representing a unit (kilograms, meters, seconds, etc.), and a unit has a dimension (mass, length, time, etc.) and I want the type-checker to be able to know what the dimension of a unit is so that the types of functions can specify the dimension that something should have.
>>>>>> The real solution to this would probably be bounded polymorphism, but I was wondering if there was some other way to do it with occurrence typing in the guard or something like that.
>>>>>>
>>>>>> On Sep 28, 2014, at 11:48 AM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> wrote:
>>>>>>
>>>>>> > Why not do this with the type, instead of making this polymorphic?
>>>>>> >
>>>>>> > Sam
>>>>>> >
>>>>>> > On Fri, Sep 26, 2014 at 7:35 PM, Alexander D. Knauth
>>>>>> > <alexander at knauth.org> wrote:
>>>>>> >> Is it possible to have a struct that does certain things according to the
>>>>>> >> guard?
>>>>>> >> #lang typed/racket
>>>>>> >>
>>>>>> >> (struct (a) foo ([a : a]) #:transparent
>>>>>> >> #:guard (lambda (a _)
>>>>>> >> (unless (exact-integer? a)
>>>>>> >> (error 'foo "expected Integer, given ~v" a))
>>>>>> >> a))
>>>>>> >>
>>>>>> >> (ann (foo (ann 1 Any)) (foo Integer))
>>>>>> >>
>>>>>> >> (: x : (foo Any))
>>>>>> >> (define x (foo 1))
>>>>>> >>
>>>>>> >> (ann (foo-a x) Integer)
>>>>>> >>
>>>>>> >> ;. Type Checker: Polymorphic function `foo1' could not be applied to
>>>>>> >> arguments:
>>>>>> >> ;Argument 1:
>>>>>> >> ; Expected: a
>>>>>> >> ; Given: Any
>>>>>> >> ;
>>>>>> >> ;Result type: (foo a)
>>>>>> >> ;Expected result: (foo Integer)
>>>>>> >> ; in: (foo (ann 1 Any))
>>>>>> >> ;. Type Checker: Polymorphic function `foo-a' could not be applied to
>>>>>> >> arguments:
>>>>>> >> ;Argument 1:
>>>>>> >> ; Expected: (foo a)
>>>>>> >> ; Given: (foo Any)
>>>>>> >> ;
>>>>>> >> ;Result type: (a : ....)
>>>>>> >> ;Expected result: Integer
>>>>>> >> ; in: (foo-a x)
>>>>>> >>
>>>>>> >> On Sep 25, 2014, at 9:42 PM, Alexander D. Knauth <alexander at knauth.org>
>>>>>> >> wrote:
>>>>>> >>
>>>>>> >> What I?m trying to accomplish is something more like this:
>>>>>> >> #lang typed/racket
>>>>>> >>
>>>>>> >> (require "dimensions.rkt")
>>>>>> >>
>>>>>> >> (struct (d) unit ([name : Any] [scalar : Positive-Real] [dimension : d])
>>>>>> >> #:transparent
>>>>>> >> #:guard (lambda (name scalar dimension _)
>>>>>> >> (unless (dimension? dimension)
>>>>>> >> (error 'unit "expected Dimension, given ~v" dimension))
>>>>>> >> (values name scalar dimension)))
>>>>>> >>
>>>>>> >> (define-type (Unitof d) (unit d))
>>>>>> >>
>>>>>> >> (define-type Unit (Unitof Dimension))
>>>>>> >>
>>>>>> >> (define Unit? (make-predicate Unit))
>>>>>> >>
>>>>>> >> (define-type Unitish
>>>>>> >> (U (Unitof Any)
>>>>>> >> Dimension
>>>>>> >> Positive-Real))
>>>>>> >>
>>>>>> >> (: ->unit : (All (d) (case-> [(Unitof d) -> (Unitof d)]
>>>>>> >> [Unitish -> Unit])))
>>>>>> >> (define (->unit u)
>>>>>> >> (cond [(unit? u)
>>>>>> >> (unless (Unit? u) ; this should never happen anyway because of the
>>>>>> >> guard
>>>>>> >> (error '->unit "expected (Unitof Dimension), given ~v" u))
>>>>>> >> u]
>>>>>> >> [(dimension? u) (unit u 1 u)]
>>>>>> >> [(positive-real? u) (unit u u dimensionless-dimension)]))
>>>>>> >>
>>>>>> >>
>>>>>> >> On Sep 25, 2014, at 6:19 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu>
>>>>>> >> wrote:
>>>>>> >>
>>>>>> >> No, I don't think you can do this. Can you say more about what you're
>>>>>> >> trying to accomplish?
>>>>>> >>
>>>>>> >> Sam
>>>>>> >>
>>>>>> >> On Thu, Sep 25, 2014 at 6:15 PM, Alexander D. Knauth
>>>>>> >> <alexander at knauth.org> wrote:
>>>>>> >>
>>>>>> >> Do any of you have any advice for getting a function like this to
>>>>>> >> type-check?
>>>>>> >> #lang typed/racket
>>>>>> >>
>>>>>> >> (: check-int : (All (a) (case-> [a -> a]
>>>>>> >> [Any -> Integer])))
>>>>>> >> (define (check-int int)
>>>>>> >> (unless (exact-integer? int)
>>>>>> >> (error 'check-int "expected Integer, given ~v" int))
>>>>>> >> int)
>>>>>> >>
>>>>>> >> ;. Type Checker: type mismatch
>>>>>> >> ; expected: a
>>>>>> >> ; given: Integer in: int
>>>>>> >>
>>>>>> >>
>>>>>> >>
>>>>>> >> ____________________
>>>>>> >> Racket Users list:
>>>>>> >> http://lists.racket-lang.org/users
>>>>>> >>
>>>>>> >>
>>>>>> >> ____________________
>>>>>> >> Racket Users list:
>>>>>> >> http://lists.racket-lang.org/users
>>>>>> >>
>>>>>> >>
>>>>>>
>>>>>>
>>>>>> ____________________
>>>>>> Racket Users list:
>>>>>> http://lists.racket-lang.org/users
>>>>>>
>>>>>>
>>>>>
>>>>> ____________________
>>>>>  Racket Users list:
>>>>>  http://lists.racket-lang.org/users
>>>>
>>>
>>
>
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140928/ca23ed92/attachment-0001.html>
>
> ------------------------------
>
> Message: 2
> Date: Sun, 28 Sep 2014 22:29:31 -0400
> From: Matthias Felleisen <matthias at ccs.neu.edu>
> To: "Alexander D. Knauth" <alexander at knauth.org>
> Cc: racket users list <users at racket-lang.org>
> Subject: Re: [racket] typed racket, filters, and polymorphism
> Message-ID: <6ADD4F6D-6B79-4F32-AD11-379CD2FF87AB at ccs.neu.edu>
> Content-Type: text/plain; charset="utf-8"
>
>
> You might be able to unwrap such things for a local context, but then you need to wrap it again on the way out. That cost might become quite high if it happens in a loop. Then again Racket programs rely on this a lot too (as do many dynamically typed programs) so it might not be a cost that kills performance completely. But do be aware of how others deal with this issue -- Matthias
>
>
>
>
>
>
>
> On Sep 28, 2014, at 10:03 PM, Alexander D. Knauth wrote:
>
>>
>>
>> On Sep 28, 2014, at 8:25 PM, Matthias Felleisen <matthias at ccs.neu.edu> wrote:
>>
>>>
>>> Your message points out where gradual typing badly fails.
>>>
>>> Type systems combine two nearly orthogonal ideas: (1) proving theorems about your program and (2) bridging the gap between abstract linguistic constructs and concrete machine implementations. In a sense, the design of a type system should always have the goal to eliminate as much run-time overhead as possible.
>>>
>>> In this specific case, you have two aspects of dimensionality: dimension per se (I am sure there is a word for it) and the chosen unit to represent it. So if you know that a number denotes a distance, you can choose mm, m, and km to represent it. If you want to represent an area, you choose mm^2, m^2, km^2. Now if you want to compute the area of a rectangle, you write
>>>
>>>  area-of-rectangle :  Real [Distance in m] * Real [Distance in m] -> Real [Area in m^2]
>>>
>>> If someone writes (area-of-rectangle 1 [mm] 1 [km]), there is nothing wrong with it -- except that the type checker should insert a coercion from mm to m and from km to m (multiplying/dividing by 1,000). That is, a type system for this application ought to use the type elaboration process idea and translate code as it goes.
>>>
>>> This is quite comparable to the idea that a machine-oriented type system needs to insert a coercion from integer to real (for example) as Algol 60 and its degraded successors do.
>>
>> What I had in mind was for the structs to be available at run-time, but that ideally the optimizer could take them out for the intermediate operations and put them back for the final result.
>>
>> And then if this value goes into untyped code, then it knows at run-time that it has the unit (u* millimeter kilometer), which is unit=? to square-meter.
>>
>> Is this sort of like what happens with flonum unboxing?
>>
>>
>>
>>>
>>> Our gradual type system lacks some (intensiona) expressive power, which is why you can't get close to this ideal.
>>>
>>> -- Matthias
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>>
>>> On Sep 28, 2014, at 8:14 PM, Alexander D. Knauth wrote:
>>>
>>>> Yes I do want run-time residual of the dimensions, and yes I want a unit struct.
>>>>
>>>> Also if there?s run-time residual, then I can use predicates and occurrence typing, and it can also be used in untyped code with contracts instead of types.
>>>>
>>>> And also the type-checker would check if it is dimension-correct, but at run-time it would would still convert between different units of the same dimension, multiply and divide units, etc.
>>>>
>>>> On Sep 28, 2014, at 6:58 PM, Matthias Felleisen <matthias at ccs.neu.edu> wrote:
>>>>
>>>>>
>>>>> If you want the type checker to ensure your program is unit-correct, I assume you also want no run-time residual of the dimensions but that is in conflict with wanting a structure because it imposes a run-time cost. Are you sure you want a unit? -- Matthias
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> On Sep 28, 2014, at 1:37 PM, Alexander D. Knauth wrote:
>>>>>
>>>>>> No because I want the unit to be a struct that has a dimension field, not a symbol with various dimensions defined as unions of units.
>>>>>> I want the unit to be based on the dimension, not the other way around, so that new units can be made that have the same dimension.
>>>>>>
>>>>>> I have something like the number+unit struct (I called it measure), but I?ll work on more that after I have the unit struct figured out.
>>>>>>
>>>>>> On Sep 28, 2014, at 12:13 PM, Spencer florence <spencerflorence at gmail.com> wrote:
>>>>>>
>>>>>>> would something like this work?
>>>>>>>
>>>>>>> #lang typed/racket
>>>>>>>
>>>>>>> (struct (U) number+unit ([amount : Real] [unit : U]))
>>>>>>>
>>>>>>> (define-type Weight-Unit (U 'kg 'g 'mg '?g))
>>>>>>> (define-type Weight (number+unit Weight-Unit))
>>>>>>> (define-predicate weight? Weight)
>>>>>>>
>>>>>>> (: make-weight : Real Weight-Unit -> Weight)
>>>>>>> (define (make-weight n u)
>>>>>>>   (number+unit n u))
>>>>>>>
>>>>>>> (: +/weight : Weight Weight -> Weight)
>>>>>>> ;; something something needs unit conversion
>>>>>>> (define (+/weight w1 w2)
>>>>>>>   (number+unit (+ (number+unit-amount w1)
>>>>>>>                   (number+unit-amount w1))
>>>>>>>                (number+unit-unit w1)))
>>>>>>>
>>>>>>> (+/weight (make-weight 1 'kg) (make-weight 1 'kg))
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> On Sun, Sep 28, 2014 at 11:03 AM, Alexander D. Knauth <alexander at knauth.org> wrote:
>>>>>>>
>>>>>>> Because the struct is representing a unit (kilograms, meters, seconds, etc.), and a unit has a dimension (mass, length, time, etc.) and I want the type-checker to be able to know what the dimension of a unit is so that the types of functions can specify the dimension that something should have.
>>>>>>> The real solution to this would probably be bounded polymorphism, but I was wondering if there was some other way to do it with occurrence typing in the guard or something like that.
>>>>>>>
>>>>>>> On Sep 28, 2014, at 11:48 AM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> wrote:
>>>>>>>
>>>>>>> > Why not do this with the type, instead of making this polymorphic?
>>>>>>> >
>>>>>>> > Sam
>>>>>>> >
>>>>>>> > On Fri, Sep 26, 2014 at 7:35 PM, Alexander D. Knauth
>>>>>>> > <alexander at knauth.org> wrote:
>>>>>>> >> Is it possible to have a struct that does certain things according to the
>>>>>>> >> guard?
>>>>>>> >> #lang typed/racket
>>>>>>> >>
>>>>>>> >> (struct (a) foo ([a : a]) #:transparent
>>>>>>> >> #:guard (lambda (a _)
>>>>>>> >> (unless (exact-integer? a)
>>>>>>> >> (error 'foo "expected Integer, given ~v" a))
>>>>>>> >> a))
>>>>>>> >>
>>>>>>> >> (ann (foo (ann 1 Any)) (foo Integer))
>>>>>>> >>
>>>>>>> >> (: x : (foo Any))
>>>>>>> >> (define x (foo 1))
>>>>>>> >>
>>>>>>> >> (ann (foo-a x) Integer)
>>>>>>> >>
>>>>>>> >> ;. Type Checker: Polymorphic function `foo1' could not be applied to
>>>>>>> >> arguments:
>>>>>>> >> ;Argument 1:
>>>>>>> >> ; Expected: a
>>>>>>> >> ; Given: Any
>>>>>>> >> ;
>>>>>>> >> ;Result type: (foo a)
>>>>>>> >> ;Expected result: (foo Integer)
>>>>>>> >> ; in: (foo (ann 1 Any))
>>>>>>> >> ;. Type Checker: Polymorphic function `foo-a' could not be applied to
>>>>>>> >> arguments:
>>>>>>> >> ;Argument 1:
>>>>>>> >> ; Expected: (foo a)
>>>>>>> >> ; Given: (foo Any)
>>>>>>> >> ;
>>>>>>> >> ;Result type: (a : ....)
>>>>>>> >> ;Expected result: Integer
>>>>>>> >> ; in: (foo-a x)
>>>>>>> >>
>>>>>>> >> On Sep 25, 2014, at 9:42 PM, Alexander D. Knauth <alexander at knauth.org>
>>>>>>> >> wrote:
>>>>>>> >>
>>>>>>> >> What I?m trying to accomplish is something more like this:
>>>>>>> >> #lang typed/racket
>>>>>>> >>
>>>>>>> >> (require "dimensions.rkt")
>>>>>>> >>
>>>>>>> >> (struct (d) unit ([name : Any] [scalar : Positive-Real] [dimension : d])
>>>>>>> >> #:transparent
>>>>>>> >> #:guard (lambda (name scalar dimension _)
>>>>>>> >> (unless (dimension? dimension)
>>>>>>> >> (error 'unit "expected Dimension, given ~v" dimension))
>>>>>>> >> (values name scalar dimension)))
>>>>>>> >>
>>>>>>> >> (define-type (Unitof d) (unit d))
>>>>>>> >>
>>>>>>> >> (define-type Unit (Unitof Dimension))
>>>>>>> >>
>>>>>>> >> (define Unit? (make-predicate Unit))
>>>>>>> >>
>>>>>>> >> (define-type Unitish
>>>>>>> >> (U (Unitof Any)
>>>>>>> >> Dimension
>>>>>>> >> Positive-Real))
>>>>>>> >>
>>>>>>> >> (: ->unit : (All (d) (case-> [(Unitof d) -> (Unitof d)]
>>>>>>> >> [Unitish -> Unit])))
>>>>>>> >> (define (->unit u)
>>>>>>> >> (cond [(unit? u)
>>>>>>> >> (unless (Unit? u) ; this should never happen anyway because of the
>>>>>>> >> guard
>>>>>>> >> (error '->unit "expected (Unitof Dimension), given ~v" u))
>>>>>>> >> u]
>>>>>>> >> [(dimension? u) (unit u 1 u)]
>>>>>>> >> [(positive-real? u) (unit u u dimensionless-dimension)]))
>>>>>>> >>
>>>>>>> >>
>>>>>>> >> On Sep 25, 2014, at 6:19 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu>
>>>>>>> >> wrote:
>>>>>>> >>
>>>>>>> >> No, I don't think you can do this. Can you say more about what you're
>>>>>>> >> trying to accomplish?
>>>>>>> >>
>>>>>>> >> Sam
>>>>>>> >>
>>>>>>> >> On Thu, Sep 25, 2014 at 6:15 PM, Alexander D. Knauth
>>>>>>> >> <alexander at knauth.org> wrote:
>>>>>>> >>
>>>>>>> >> Do any of you have any advice for getting a function like this to
>>>>>>> >> type-check?
>>>>>>> >> #lang typed/racket
>>>>>>> >>
>>>>>>> >> (: check-int : (All (a) (case-> [a -> a]
>>>>>>> >> [Any -> Integer])))
>>>>>>> >> (define (check-int int)
>>>>>>> >> (unless (exact-integer? int)
>>>>>>> >> (error 'check-int "expected Integer, given ~v" int))
>>>>>>> >> int)
>>>>>>> >>
>>>>>>> >> ;. Type Checker: type mismatch
>>>>>>> >> ; expected: a
>>>>>>> >> ; given: Integer in: int
>>>>>>> >>
>>>>>>> >>
>>>>>>> >>
>>>>>>> >> ____________________
>>>>>>> >> Racket Users list:
>>>>>>> >> http://lists.racket-lang.org/users
>>>>>>> >>
>>>>>>> >>
>>>>>>> >> ____________________
>>>>>>> >> Racket Users list:
>>>>>>> >> http://lists.racket-lang.org/users
>>>>>>> >>
>>>>>>> >>
>>>>>>>
>>>>>>>
>>>>>>> ____________________
>>>>>>> Racket Users list:
>>>>>>> http://lists.racket-lang.org/users
>>>>>>>
>>>>>>>
>>>>>>
>>>>>> ____________________
>>>>>>  Racket Users list:
>>>>>>  http://lists.racket-lang.org/users
>>>>>
>>>>
>>>
>>
>
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140928/d95c74e9/attachment.html>
>
> End of users Digest, Vol 109, Issue 73
> **************************************


From moshedeutsch115 at gmail.com  Mon Sep 29 16:41:38 2014
From: moshedeutsch115 at gmail.com (Moshe Deutsch)
Date: Mon, 29 Sep 2014 16:41:38 -0400
Subject: [racket] users Digest, Vol 109, Issue 72
In-Reply-To: <mailman.482.1411950363.679.users@racket-lang.org>
References: <mailman.482.1411950363.679.users@racket-lang.org>
Message-ID: <CAKpSqLCcD5z6xjR1tbKPQUFWrx9dAh0zwjEdHiGyTbKKKWjc-w@mail.gmail.com>

Please take me off the list

Thanks

On Sun, Sep 28, 2014 at 8:26 PM,  <users-request at racket-lang.org> wrote:
> Send users mailing list submissions to
>         users at racket-lang.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>         http://lists.racket-lang.org/users/listinfo
> or, via email, send a message with subject or body 'help' to
>         users-request at racket-lang.org
>
> You can reach the person managing the list at
>         users-owner at racket-lang.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of users digest..."
>
>
> [Racket Users list:
>  http://lists.racket-lang.org/users ]
>
>
> Today's Topics:
>
>    1. Re: typed racket, filters, and polymorphism (Alexander D. Knauth)
>    2. Re: typed racket, filters, and polymorphism (Matthias Felleisen)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Sun, 28 Sep 2014 20:14:43 -0400
> From: "Alexander D. Knauth" <alexander at knauth.org>
> To: Matthias Felleisen <matthias at ccs.neu.edu>
> Cc: Spencer florence <spencerflorence at gmail.com>, racket users list
>         <users at racket-lang.org>, Sam Tobin-Hochstadt <samth at cs.indiana.edu>
> Subject: Re: [racket] typed racket, filters, and polymorphism
> Message-ID: <B52D6D46-C9C1-46D6-9F1E-46800DB7B23C at knauth.org>
> Content-Type: text/plain; charset="utf-8"
>
> Yes I do want run-time residual of the dimensions, and yes I want a unit struct.
>
> Also if there?s run-time residual, then I can use predicates and occurrence typing, and it can also be used in untyped code with contracts instead of types.
>
> And also the type-checker would check if it is dimension-correct, but at run-time it would would still convert between different units of the same dimension, multiply and divide units, etc.
>
> On Sep 28, 2014, at 6:58 PM, Matthias Felleisen <matthias at ccs.neu.edu> wrote:
>
>>
>> If you want the type checker to ensure your program is unit-correct, I assume you also want no run-time residual of the dimensions but that is in conflict with wanting a structure because it imposes a run-time cost. Are you sure you want a unit? -- Matthias
>>
>>
>>
>>
>>
>>
>> On Sep 28, 2014, at 1:37 PM, Alexander D. Knauth wrote:
>>
>>> No because I want the unit to be a struct that has a dimension field, not a symbol with various dimensions defined as unions of units.
>>> I want the unit to be based on the dimension, not the other way around, so that new units can be made that have the same dimension.
>>>
>>> I have something like the number+unit struct (I called it measure), but I?ll work on more that after I have the unit struct figured out.
>>>
>>> On Sep 28, 2014, at 12:13 PM, Spencer florence <spencerflorence at gmail.com> wrote:
>>>
>>>> would something like this work?
>>>>
>>>> #lang typed/racket
>>>>
>>>> (struct (U) number+unit ([amount : Real] [unit : U]))
>>>>
>>>> (define-type Weight-Unit (U 'kg 'g 'mg '?g))
>>>> (define-type Weight (number+unit Weight-Unit))
>>>> (define-predicate weight? Weight)
>>>>
>>>> (: make-weight : Real Weight-Unit -> Weight)
>>>> (define (make-weight n u)
>>>>   (number+unit n u))
>>>>
>>>> (: +/weight : Weight Weight -> Weight)
>>>> ;; something something needs unit conversion
>>>> (define (+/weight w1 w2)
>>>>   (number+unit (+ (number+unit-amount w1)
>>>>                   (number+unit-amount w1))
>>>>                (number+unit-unit w1)))
>>>>
>>>> (+/weight (make-weight 1 'kg) (make-weight 1 'kg))
>>>>
>>>>
>>>>
>>>> On Sun, Sep 28, 2014 at 11:03 AM, Alexander D. Knauth <alexander at knauth.org> wrote:
>>>>
>>>> Because the struct is representing a unit (kilograms, meters, seconds, etc.), and a unit has a dimension (mass, length, time, etc.) and I want the type-checker to be able to know what the dimension of a unit is so that the types of functions can specify the dimension that something should have.
>>>> The real solution to this would probably be bounded polymorphism, but I was wondering if there was some other way to do it with occurrence typing in the guard or something like that.
>>>>
>>>> On Sep 28, 2014, at 11:48 AM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> wrote:
>>>>
>>>> > Why not do this with the type, instead of making this polymorphic?
>>>> >
>>>> > Sam
>>>> >
>>>> > On Fri, Sep 26, 2014 at 7:35 PM, Alexander D. Knauth
>>>> > <alexander at knauth.org> wrote:
>>>> >> Is it possible to have a struct that does certain things according to the
>>>> >> guard?
>>>> >> #lang typed/racket
>>>> >>
>>>> >> (struct (a) foo ([a : a]) #:transparent
>>>> >> #:guard (lambda (a _)
>>>> >> (unless (exact-integer? a)
>>>> >> (error 'foo "expected Integer, given ~v" a))
>>>> >> a))
>>>> >>
>>>> >> (ann (foo (ann 1 Any)) (foo Integer))
>>>> >>
>>>> >> (: x : (foo Any))
>>>> >> (define x (foo 1))
>>>> >>
>>>> >> (ann (foo-a x) Integer)
>>>> >>
>>>> >> ;. Type Checker: Polymorphic function `foo1' could not be applied to
>>>> >> arguments:
>>>> >> ;Argument 1:
>>>> >> ; Expected: a
>>>> >> ; Given: Any
>>>> >> ;
>>>> >> ;Result type: (foo a)
>>>> >> ;Expected result: (foo Integer)
>>>> >> ; in: (foo (ann 1 Any))
>>>> >> ;. Type Checker: Polymorphic function `foo-a' could not be applied to
>>>> >> arguments:
>>>> >> ;Argument 1:
>>>> >> ; Expected: (foo a)
>>>> >> ; Given: (foo Any)
>>>> >> ;
>>>> >> ;Result type: (a : ....)
>>>> >> ;Expected result: Integer
>>>> >> ; in: (foo-a x)
>>>> >>
>>>> >> On Sep 25, 2014, at 9:42 PM, Alexander D. Knauth <alexander at knauth.org>
>>>> >> wrote:
>>>> >>
>>>> >> What I?m trying to accomplish is something more like this:
>>>> >> #lang typed/racket
>>>> >>
>>>> >> (require "dimensions.rkt")
>>>> >>
>>>> >> (struct (d) unit ([name : Any] [scalar : Positive-Real] [dimension : d])
>>>> >> #:transparent
>>>> >> #:guard (lambda (name scalar dimension _)
>>>> >> (unless (dimension? dimension)
>>>> >> (error 'unit "expected Dimension, given ~v" dimension))
>>>> >> (values name scalar dimension)))
>>>> >>
>>>> >> (define-type (Unitof d) (unit d))
>>>> >>
>>>> >> (define-type Unit (Unitof Dimension))
>>>> >>
>>>> >> (define Unit? (make-predicate Unit))
>>>> >>
>>>> >> (define-type Unitish
>>>> >> (U (Unitof Any)
>>>> >> Dimension
>>>> >> Positive-Real))
>>>> >>
>>>> >> (: ->unit : (All (d) (case-> [(Unitof d) -> (Unitof d)]
>>>> >> [Unitish -> Unit])))
>>>> >> (define (->unit u)
>>>> >> (cond [(unit? u)
>>>> >> (unless (Unit? u) ; this should never happen anyway because of the
>>>> >> guard
>>>> >> (error '->unit "expected (Unitof Dimension), given ~v" u))
>>>> >> u]
>>>> >> [(dimension? u) (unit u 1 u)]
>>>> >> [(positive-real? u) (unit u u dimensionless-dimension)]))
>>>> >>
>>>> >>
>>>> >> On Sep 25, 2014, at 6:19 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu>
>>>> >> wrote:
>>>> >>
>>>> >> No, I don't think you can do this. Can you say more about what you're
>>>> >> trying to accomplish?
>>>> >>
>>>> >> Sam
>>>> >>
>>>> >> On Thu, Sep 25, 2014 at 6:15 PM, Alexander D. Knauth
>>>> >> <alexander at knauth.org> wrote:
>>>> >>
>>>> >> Do any of you have any advice for getting a function like this to
>>>> >> type-check?
>>>> >> #lang typed/racket
>>>> >>
>>>> >> (: check-int : (All (a) (case-> [a -> a]
>>>> >> [Any -> Integer])))
>>>> >> (define (check-int int)
>>>> >> (unless (exact-integer? int)
>>>> >> (error 'check-int "expected Integer, given ~v" int))
>>>> >> int)
>>>> >>
>>>> >> ;. Type Checker: type mismatch
>>>> >> ; expected: a
>>>> >> ; given: Integer in: int
>>>> >>
>>>> >>
>>>> >>
>>>> >> ____________________
>>>> >> Racket Users list:
>>>> >> http://lists.racket-lang.org/users
>>>> >>
>>>> >>
>>>> >> ____________________
>>>> >> Racket Users list:
>>>> >> http://lists.racket-lang.org/users
>>>> >>
>>>> >>
>>>>
>>>>
>>>> ____________________
>>>> Racket Users list:
>>>> http://lists.racket-lang.org/users
>>>>
>>>>
>>>
>>> ____________________
>>>  Racket Users list:
>>>  http://lists.racket-lang.org/users
>>
>
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140928/f86df9a7/attachment-0001.html>
>
> ------------------------------
>
> Message: 2
> Date: Sun, 28 Sep 2014 20:25:45 -0400
> From: Matthias Felleisen <matthias at ccs.neu.edu>
> To: "Alexander D. Knauth" <alexander at knauth.org>
> Cc: racket users list <users at racket-lang.org>, Benjamin Greenman
>         <blg59 at cornell.edu>
> Subject: Re: [racket] typed racket, filters, and polymorphism
> Message-ID: <C3486009-3F00-4717-BD42-E18FF074FD4B at ccs.neu.edu>
> Content-Type: text/plain; charset="utf-8"
>
>
> Your message points out where gradual typing badly fails.
>
> Type systems combine two nearly orthogonal ideas: (1) proving theorems about your program and (2) bridging the gap between abstract linguistic constructs and concrete machine implementations. In a sense, the design of a type system should always have the goal to eliminate as much run-time overhead as possible.
>
> In this specific case, you have two aspects of dimensionality: dimension per se (I am sure there is a word for it) and the chosen unit to represent it. So if you know that a number denotes a distance, you can choose mm, m, and km to represent it. If you want to represent an area, you choose mm^2, m^2, km^2. Now if you want to compute the area of a rectangle, you write
>
>  area-of-rectangle :  Real [Distance in m] * Real [Distance in m] -> Real [Area in m^2]
>
> If someone writes (area-of-rectangle 1 [mm] 1 [km]), there is nothing wrong with it -- except that the type checker should insert a coercion from mm to m and from km to m (multiplying/dividing by 1,000). That is, a type system for this application ought to use the type elaboration process idea and translate code as it goes.
>
> This is quite comparable to the idea that a machine-oriented type system needs to insert a coercion from integer to real (for example) as Algol 60 and its degraded successors do.
>
> Our gradual type system lacks some (intensiona) expressive power, which is why you can't get close to this ideal.
>
> -- Matthias
>
>
>
>
>
>
>
>
> On Sep 28, 2014, at 8:14 PM, Alexander D. Knauth wrote:
>
>> Yes I do want run-time residual of the dimensions, and yes I want a unit struct.
>>
>> Also if there?s run-time residual, then I can use predicates and occurrence typing, and it can also be used in untyped code with contracts instead of types.
>>
>> And also the type-checker would check if it is dimension-correct, but at run-time it would would still convert between different units of the same dimension, multiply and divide units, etc.
>>
>> On Sep 28, 2014, at 6:58 PM, Matthias Felleisen <matthias at ccs.neu.edu> wrote:
>>
>>>
>>> If you want the type checker to ensure your program is unit-correct, I assume you also want no run-time residual of the dimensions but that is in conflict with wanting a structure because it imposes a run-time cost. Are you sure you want a unit? -- Matthias
>>>
>>>
>>>
>>>
>>>
>>>
>>> On Sep 28, 2014, at 1:37 PM, Alexander D. Knauth wrote:
>>>
>>>> No because I want the unit to be a struct that has a dimension field, not a symbol with various dimensions defined as unions of units.
>>>> I want the unit to be based on the dimension, not the other way around, so that new units can be made that have the same dimension.
>>>>
>>>> I have something like the number+unit struct (I called it measure), but I?ll work on more that after I have the unit struct figured out.
>>>>
>>>> On Sep 28, 2014, at 12:13 PM, Spencer florence <spencerflorence at gmail.com> wrote:
>>>>
>>>>> would something like this work?
>>>>>
>>>>> #lang typed/racket
>>>>>
>>>>> (struct (U) number+unit ([amount : Real] [unit : U]))
>>>>>
>>>>> (define-type Weight-Unit (U 'kg 'g 'mg '?g))
>>>>> (define-type Weight (number+unit Weight-Unit))
>>>>> (define-predicate weight? Weight)
>>>>>
>>>>> (: make-weight : Real Weight-Unit -> Weight)
>>>>> (define (make-weight n u)
>>>>>   (number+unit n u))
>>>>>
>>>>> (: +/weight : Weight Weight -> Weight)
>>>>> ;; something something needs unit conversion
>>>>> (define (+/weight w1 w2)
>>>>>   (number+unit (+ (number+unit-amount w1)
>>>>>                   (number+unit-amount w1))
>>>>>                (number+unit-unit w1)))
>>>>>
>>>>> (+/weight (make-weight 1 'kg) (make-weight 1 'kg))
>>>>>
>>>>>
>>>>>
>>>>> On Sun, Sep 28, 2014 at 11:03 AM, Alexander D. Knauth <alexander at knauth.org> wrote:
>>>>>
>>>>> Because the struct is representing a unit (kilograms, meters, seconds, etc.), and a unit has a dimension (mass, length, time, etc.) and I want the type-checker to be able to know what the dimension of a unit is so that the types of functions can specify the dimension that something should have.
>>>>> The real solution to this would probably be bounded polymorphism, but I was wondering if there was some other way to do it with occurrence typing in the guard or something like that.
>>>>>
>>>>> On Sep 28, 2014, at 11:48 AM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> wrote:
>>>>>
>>>>> > Why not do this with the type, instead of making this polymorphic?
>>>>> >
>>>>> > Sam
>>>>> >
>>>>> > On Fri, Sep 26, 2014 at 7:35 PM, Alexander D. Knauth
>>>>> > <alexander at knauth.org> wrote:
>>>>> >> Is it possible to have a struct that does certain things according to the
>>>>> >> guard?
>>>>> >> #lang typed/racket
>>>>> >>
>>>>> >> (struct (a) foo ([a : a]) #:transparent
>>>>> >> #:guard (lambda (a _)
>>>>> >> (unless (exact-integer? a)
>>>>> >> (error 'foo "expected Integer, given ~v" a))
>>>>> >> a))
>>>>> >>
>>>>> >> (ann (foo (ann 1 Any)) (foo Integer))
>>>>> >>
>>>>> >> (: x : (foo Any))
>>>>> >> (define x (foo 1))
>>>>> >>
>>>>> >> (ann (foo-a x) Integer)
>>>>> >>
>>>>> >> ;. Type Checker: Polymorphic function `foo1' could not be applied to
>>>>> >> arguments:
>>>>> >> ;Argument 1:
>>>>> >> ; Expected: a
>>>>> >> ; Given: Any
>>>>> >> ;
>>>>> >> ;Result type: (foo a)
>>>>> >> ;Expected result: (foo Integer)
>>>>> >> ; in: (foo (ann 1 Any))
>>>>> >> ;. Type Checker: Polymorphic function `foo-a' could not be applied to
>>>>> >> arguments:
>>>>> >> ;Argument 1:
>>>>> >> ; Expected: (foo a)
>>>>> >> ; Given: (foo Any)
>>>>> >> ;
>>>>> >> ;Result type: (a : ....)
>>>>> >> ;Expected result: Integer
>>>>> >> ; in: (foo-a x)
>>>>> >>
>>>>> >> On Sep 25, 2014, at 9:42 PM, Alexander D. Knauth <alexander at knauth.org>
>>>>> >> wrote:
>>>>> >>
>>>>> >> What I?m trying to accomplish is something more like this:
>>>>> >> #lang typed/racket
>>>>> >>
>>>>> >> (require "dimensions.rkt")
>>>>> >>
>>>>> >> (struct (d) unit ([name : Any] [scalar : Positive-Real] [dimension : d])
>>>>> >> #:transparent
>>>>> >> #:guard (lambda (name scalar dimension _)
>>>>> >> (unless (dimension? dimension)
>>>>> >> (error 'unit "expected Dimension, given ~v" dimension))
>>>>> >> (values name scalar dimension)))
>>>>> >>
>>>>> >> (define-type (Unitof d) (unit d))
>>>>> >>
>>>>> >> (define-type Unit (Unitof Dimension))
>>>>> >>
>>>>> >> (define Unit? (make-predicate Unit))
>>>>> >>
>>>>> >> (define-type Unitish
>>>>> >> (U (Unitof Any)
>>>>> >> Dimension
>>>>> >> Positive-Real))
>>>>> >>
>>>>> >> (: ->unit : (All (d) (case-> [(Unitof d) -> (Unitof d)]
>>>>> >> [Unitish -> Unit])))
>>>>> >> (define (->unit u)
>>>>> >> (cond [(unit? u)
>>>>> >> (unless (Unit? u) ; this should never happen anyway because of the
>>>>> >> guard
>>>>> >> (error '->unit "expected (Unitof Dimension), given ~v" u))
>>>>> >> u]
>>>>> >> [(dimension? u) (unit u 1 u)]
>>>>> >> [(positive-real? u) (unit u u dimensionless-dimension)]))
>>>>> >>
>>>>> >>
>>>>> >> On Sep 25, 2014, at 6:19 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu>
>>>>> >> wrote:
>>>>> >>
>>>>> >> No, I don't think you can do this. Can you say more about what you're
>>>>> >> trying to accomplish?
>>>>> >>
>>>>> >> Sam
>>>>> >>
>>>>> >> On Thu, Sep 25, 2014 at 6:15 PM, Alexander D. Knauth
>>>>> >> <alexander at knauth.org> wrote:
>>>>> >>
>>>>> >> Do any of you have any advice for getting a function like this to
>>>>> >> type-check?
>>>>> >> #lang typed/racket
>>>>> >>
>>>>> >> (: check-int : (All (a) (case-> [a -> a]
>>>>> >> [Any -> Integer])))
>>>>> >> (define (check-int int)
>>>>> >> (unless (exact-integer? int)
>>>>> >> (error 'check-int "expected Integer, given ~v" int))
>>>>> >> int)
>>>>> >>
>>>>> >> ;. Type Checker: type mismatch
>>>>> >> ; expected: a
>>>>> >> ; given: Integer in: int
>>>>> >>
>>>>> >>
>>>>> >>
>>>>> >> ____________________
>>>>> >> Racket Users list:
>>>>> >> http://lists.racket-lang.org/users
>>>>> >>
>>>>> >>
>>>>> >> ____________________
>>>>> >> Racket Users list:
>>>>> >> http://lists.racket-lang.org/users
>>>>> >>
>>>>> >>
>>>>>
>>>>>
>>>>> ____________________
>>>>> Racket Users list:
>>>>> http://lists.racket-lang.org/users
>>>>>
>>>>>
>>>>
>>>> ____________________
>>>>  Racket Users list:
>>>>  http://lists.racket-lang.org/users
>>>
>>
>
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140928/d777b137/attachment.html>
>
> End of users Digest, Vol 109, Issue 72
> **************************************


From moshedeutsch115 at gmail.com  Mon Sep 29 16:41:47 2014
From: moshedeutsch115 at gmail.com (Moshe Deutsch)
Date: Mon, 29 Sep 2014 16:41:47 -0400
Subject: [racket] users Digest, Vol 109, Issue 71
In-Reply-To: <mailman.480.1411945118.679.users@racket-lang.org>
References: <mailman.480.1411945118.679.users@racket-lang.org>
Message-ID: <CAKpSqLCH7nBqQbVGUDzGiqHP4rOqXTOR_VyZdM2yhh5CpFNoPA@mail.gmail.com>

Please take me off the list

Thanks

On Sun, Sep 28, 2014 at 6:58 PM,  <users-request at racket-lang.org> wrote:
> Send users mailing list submissions to
>         users at racket-lang.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>         http://lists.racket-lang.org/users/listinfo
> or, via email, send a message with subject or body 'help' to
>         users-request at racket-lang.org
>
> You can reach the person managing the list at
>         users-owner at racket-lang.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of users digest..."
>
>
> [Racket Users list:
>  http://lists.racket-lang.org/users ]
>
>
> Today's Topics:
>
>    1. Re: aws/glacier: credential scope (Greg Hendershott)
>    2. Re: aws/glacier: credential scope (Greg Hendershott)
>    3. Re: aws/glacier: credential scope (Norman Gray)
>    4. Re: aws/glacier: credential scope (Greg Hendershott)
>    5. Re: typed racket, filters, and polymorphism (Matthias Felleisen)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Sun, 28 Sep 2014 17:15:21 -0400
> From: Greg Hendershott <greghendershott at gmail.com>
> To: Jay McCarthy <jay at racket-lang.org>
> Cc: users at racket-lang.org
> Subject: Re: [racket] aws/glacier: credential scope
> Message-ID:
>         <CAGspUn2t_ZOaEAgM2PDvSJDrTZbw2aX0VO66QdqmXi6REjJcMQ at mail.gmail.com>
> Content-Type: text/plain; charset=UTF-8
>
>> I clicked "update" on pkgs.racket-lang.org. But it seems to be slower
>> than usual to refresh. After it does, you can `raco pkg update aws` to
>> get the fix.
>
> Something seems wrong/stuck on pkgs.racket-lang.org. The top of the page says:
>
>   "update upload in progress: there may be inconsistencies below"
>
> It's been saying that for nearly an hour.
>
> In the past, clicking "update" would cause a successful refresh -- the
> package would get a red star indicating it's updated -- within a
> minute or two.
>
>
> ------------------------------
>
> Message: 2
> Date: Sun, 28 Sep 2014 17:17:32 -0400
> From: Greg Hendershott <greghendershott at gmail.com>
> To: Jay McCarthy <jay at racket-lang.org>
> Cc: users at racket-lang.org
> Subject: Re: [racket] aws/glacier: credential scope
> Message-ID:
>         <CAGspUn1tMRkaZjK30rvxg+aUS7wXG5L_XT=_toU5o0YBg0pV+A at mail.gmail.com>
> Content-Type: text/plain; charset=UTF-8
>
>>> I clicked "update" on pkgs.racket-lang.org. But it seems to be slower
>>> than usual to refresh. After it does, you can `raco pkg update aws` to
>>> get the fix.
>>
>> Something seems wrong/stuck on pkgs.racket-lang.org. The top of the page says:
>>
>>   "update upload in progress: there may be inconsistencies below"
>>
>> It's been saying that for nearly an hour.
>
> Sometimes instead it says:
>
>   "update upload being computed: the information below may not
> represent all recent changes and updates"
>
> It has switched between those two messages at various intervals over
> the last hour.
>
>
> ------------------------------
>
> Message: 3
> Date: Sun, 28 Sep 2014 22:40:19 +0100
> From: Norman Gray <norman at astro.gla.ac.uk>
> To: Greg Hendershott <greghendershott at gmail.com>
> Cc: Jay McCarthy <jay at racket-lang.org>, users at racket-lang.org
> Subject: Re: [racket] aws/glacier: credential scope
> Message-ID: <73617243-E0D5-43C6-872B-EA0121AD5142 at astro.gla.ac.uk>
> Content-Type: text/plain; charset=us-ascii
>
>
> On 2014 Sep 28, at 22:17, Greg Hendershott <greghendershott at gmail.com> wrote:
>
>>>> I clicked "update" on pkgs.racket-lang.org. But it seems to be slower
>>>> than usual to refresh. After it does, you can `raco pkg update aws` to
>>>> get the fix.
>>>
>>> Something seems wrong/stuck on pkgs.racket-lang.org. The top of the page says:
>>>
>>>  "update upload in progress: there may be inconsistencies below"
>>>
>>> It's been saying that for nearly an hour.
>>
>> Sometimes instead it says:
>>
>>  "update upload being computed: the information below may not
>> represent all recent changes and updates"
>>
>> It has switched between those two messages at various intervals over
>> the last hour.
>
> And in case it's a useful data point, I'm getting
>
> % raco pkg update aws
> Resolving "aws" via http://download.racket-lang.org/releases/6.1/catalog/
> Resolving "aws" via http://pkgs.racket-lang.org
> No updates available
> %
>
> And Greg, your fix is close enough to what I guessed might be the problem that I'm now kicking myself for not just diving in and trying a fix myself.  Garhhhh...
>
> Norman
>
>
> --
> Norman Gray  :  http://nxg.me.uk
> SUPA School of Physics and Astronomy, University of Glasgow, UK
>
>
>
>
> ------------------------------
>
> Message: 4
> Date: Sun, 28 Sep 2014 18:10:55 -0400
> From: Greg Hendershott <greghendershott at gmail.com>
> To: Norman Gray <norman at astro.gla.ac.uk>
> Cc: users at racket-lang.org
> Subject: Re: [racket] aws/glacier: credential scope
> Message-ID:
>         <CAGspUn2cc44ouZpTmvu8e6nW_EmtjWNccco6WoH5p2QuU=4zTQ at mail.gmail.com>
> Content-Type: text/plain; charset=UTF-8
>
> If you're in a hurry you could remove and re-install directly from GitHub:
>
>  $ raco pkg remove aws
>  $ raco pkg install git://github.com/greghendershott/aws
>
> On Sun, Sep 28, 2014 at 5:40 PM, Norman Gray <norman at astro.gla.ac.uk> wrote:
>>
>> On 2014 Sep 28, at 22:17, Greg Hendershott <greghendershott at gmail.com> wrote:
>>
>>>>> I clicked "update" on pkgs.racket-lang.org. But it seems to be slower
>>>>> than usual to refresh. After it does, you can `raco pkg update aws` to
>>>>> get the fix.
>>>>
>>>> Something seems wrong/stuck on pkgs.racket-lang.org. The top of the page says:
>>>>
>>>>  "update upload in progress: there may be inconsistencies below"
>>>>
>>>> It's been saying that for nearly an hour.
>>>
>>> Sometimes instead it says:
>>>
>>>  "update upload being computed: the information below may not
>>> represent all recent changes and updates"
>>>
>>> It has switched between those two messages at various intervals over
>>> the last hour.
>>
>> And in case it's a useful data point, I'm getting
>>
>> % raco pkg update aws
>> Resolving "aws" via http://download.racket-lang.org/releases/6.1/catalog/
>> Resolving "aws" via http://pkgs.racket-lang.org
>> No updates available
>> %
>>
>> And Greg, your fix is close enough to what I guessed might be the problem that I'm now kicking myself for not just diving in and trying a fix myself.  Garhhhh...
>>
>> Norman
>>
>>
>> --
>> Norman Gray  :  http://nxg.me.uk
>> SUPA School of Physics and Astronomy, University of Glasgow, UK
>>
>
>
> ------------------------------
>
> Message: 5
> Date: Sun, 28 Sep 2014 18:58:20 -0400
> From: Matthias Felleisen <matthias at ccs.neu.edu>
> To: "Alexander D. Knauth" <alexander at knauth.org>
> Cc: Spencer florence <spencerflorence at gmail.com>, racket users list
>         <users at racket-lang.org>, Sam Tobin-Hochstadt <samth at cs.indiana.edu>
> Subject: Re: [racket] typed racket, filters, and polymorphism
> Message-ID: <2B2B37BB-377F-4D28-8E69-EFEFCF8181A2 at ccs.neu.edu>
> Content-Type: text/plain; charset="utf-8"
>
>
> If you want the type checker to ensure your program is unit-correct, I assume you also want no run-time residual of the dimensions but that is in conflict with wanting a structure because it imposes a run-time cost. Are you sure you want a unit? -- Matthias
>
>
>
>
>
>
> On Sep 28, 2014, at 1:37 PM, Alexander D. Knauth wrote:
>
>> No because I want the unit to be a struct that has a dimension field, not a symbol with various dimensions defined as unions of units.
>> I want the unit to be based on the dimension, not the other way around, so that new units can be made that have the same dimension.
>>
>> I have something like the number+unit struct (I called it measure), but I?ll work on more that after I have the unit struct figured out.
>>
>> On Sep 28, 2014, at 12:13 PM, Spencer florence <spencerflorence at gmail.com> wrote:
>>
>>> would something like this work?
>>>
>>> #lang typed/racket
>>>
>>> (struct (U) number+unit ([amount : Real] [unit : U]))
>>>
>>> (define-type Weight-Unit (U 'kg 'g 'mg '?g))
>>> (define-type Weight (number+unit Weight-Unit))
>>> (define-predicate weight? Weight)
>>>
>>> (: make-weight : Real Weight-Unit -> Weight)
>>> (define (make-weight n u)
>>>   (number+unit n u))
>>>
>>> (: +/weight : Weight Weight -> Weight)
>>> ;; something something needs unit conversion
>>> (define (+/weight w1 w2)
>>>   (number+unit (+ (number+unit-amount w1)
>>>                   (number+unit-amount w1))
>>>                (number+unit-unit w1)))
>>>
>>> (+/weight (make-weight 1 'kg) (make-weight 1 'kg))
>>>
>>>
>>>
>>> On Sun, Sep 28, 2014 at 11:03 AM, Alexander D. Knauth <alexander at knauth.org> wrote:
>>>
>>> Because the struct is representing a unit (kilograms, meters, seconds, etc.), and a unit has a dimension (mass, length, time, etc.) and I want the type-checker to be able to know what the dimension of a unit is so that the types of functions can specify the dimension that something should have.
>>> The real solution to this would probably be bounded polymorphism, but I was wondering if there was some other way to do it with occurrence typing in the guard or something like that.
>>>
>>> On Sep 28, 2014, at 11:48 AM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> wrote:
>>>
>>> > Why not do this with the type, instead of making this polymorphic?
>>> >
>>> > Sam
>>> >
>>> > On Fri, Sep 26, 2014 at 7:35 PM, Alexander D. Knauth
>>> > <alexander at knauth.org> wrote:
>>> >> Is it possible to have a struct that does certain things according to the
>>> >> guard?
>>> >> #lang typed/racket
>>> >>
>>> >> (struct (a) foo ([a : a]) #:transparent
>>> >> #:guard (lambda (a _)
>>> >> (unless (exact-integer? a)
>>> >> (error 'foo "expected Integer, given ~v" a))
>>> >> a))
>>> >>
>>> >> (ann (foo (ann 1 Any)) (foo Integer))
>>> >>
>>> >> (: x : (foo Any))
>>> >> (define x (foo 1))
>>> >>
>>> >> (ann (foo-a x) Integer)
>>> >>
>>> >> ;. Type Checker: Polymorphic function `foo1' could not be applied to
>>> >> arguments:
>>> >> ;Argument 1:
>>> >> ; Expected: a
>>> >> ; Given: Any
>>> >> ;
>>> >> ;Result type: (foo a)
>>> >> ;Expected result: (foo Integer)
>>> >> ; in: (foo (ann 1 Any))
>>> >> ;. Type Checker: Polymorphic function `foo-a' could not be applied to
>>> >> arguments:
>>> >> ;Argument 1:
>>> >> ; Expected: (foo a)
>>> >> ; Given: (foo Any)
>>> >> ;
>>> >> ;Result type: (a : ....)
>>> >> ;Expected result: Integer
>>> >> ; in: (foo-a x)
>>> >>
>>> >> On Sep 25, 2014, at 9:42 PM, Alexander D. Knauth <alexander at knauth.org>
>>> >> wrote:
>>> >>
>>> >> What I?m trying to accomplish is something more like this:
>>> >> #lang typed/racket
>>> >>
>>> >> (require "dimensions.rkt")
>>> >>
>>> >> (struct (d) unit ([name : Any] [scalar : Positive-Real] [dimension : d])
>>> >> #:transparent
>>> >> #:guard (lambda (name scalar dimension _)
>>> >> (unless (dimension? dimension)
>>> >> (error 'unit "expected Dimension, given ~v" dimension))
>>> >> (values name scalar dimension)))
>>> >>
>>> >> (define-type (Unitof d) (unit d))
>>> >>
>>> >> (define-type Unit (Unitof Dimension))
>>> >>
>>> >> (define Unit? (make-predicate Unit))
>>> >>
>>> >> (define-type Unitish
>>> >> (U (Unitof Any)
>>> >> Dimension
>>> >> Positive-Real))
>>> >>
>>> >> (: ->unit : (All (d) (case-> [(Unitof d) -> (Unitof d)]
>>> >> [Unitish -> Unit])))
>>> >> (define (->unit u)
>>> >> (cond [(unit? u)
>>> >> (unless (Unit? u) ; this should never happen anyway because of the
>>> >> guard
>>> >> (error '->unit "expected (Unitof Dimension), given ~v" u))
>>> >> u]
>>> >> [(dimension? u) (unit u 1 u)]
>>> >> [(positive-real? u) (unit u u dimensionless-dimension)]))
>>> >>
>>> >>
>>> >> On Sep 25, 2014, at 6:19 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu>
>>> >> wrote:
>>> >>
>>> >> No, I don't think you can do this. Can you say more about what you're
>>> >> trying to accomplish?
>>> >>
>>> >> Sam
>>> >>
>>> >> On Thu, Sep 25, 2014 at 6:15 PM, Alexander D. Knauth
>>> >> <alexander at knauth.org> wrote:
>>> >>
>>> >> Do any of you have any advice for getting a function like this to
>>> >> type-check?
>>> >> #lang typed/racket
>>> >>
>>> >> (: check-int : (All (a) (case-> [a -> a]
>>> >> [Any -> Integer])))
>>> >> (define (check-int int)
>>> >> (unless (exact-integer? int)
>>> >> (error 'check-int "expected Integer, given ~v" int))
>>> >> int)
>>> >>
>>> >> ;. Type Checker: type mismatch
>>> >> ; expected: a
>>> >> ; given: Integer in: int
>>> >>
>>> >>
>>> >>
>>> >> ____________________
>>> >> Racket Users list:
>>> >> http://lists.racket-lang.org/users
>>> >>
>>> >>
>>> >> ____________________
>>> >> Racket Users list:
>>> >> http://lists.racket-lang.org/users
>>> >>
>>> >>
>>>
>>>
>>> ____________________
>>> Racket Users list:
>>> http://lists.racket-lang.org/users
>>>
>>>
>>
>> ____________________
>>  Racket Users list:
>>  http://lists.racket-lang.org/users
>
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140928/0821b16d/attachment.html>
>
> End of users Digest, Vol 109, Issue 71
> **************************************


From moshedeutsch115 at gmail.com  Mon Sep 29 16:41:57 2014
From: moshedeutsch115 at gmail.com (Moshe Deutsch)
Date: Mon, 29 Sep 2014 16:41:57 -0400
Subject: [racket] users Digest, Vol 109, Issue 70
In-Reply-To: <mailman.478.1411936198.679.users@racket-lang.org>
References: <mailman.478.1411936198.679.users@racket-lang.org>
Message-ID: <CAKpSqLBEwno4bLuEhXsq552WszgRF-i23h0wQMQEWSB6U5aAqQ@mail.gmail.com>

Please take me off the list

Thanks

On Sun, Sep 28, 2014 at 4:29 PM,  <users-request at racket-lang.org> wrote:
> Send users mailing list submissions to
>         users at racket-lang.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>         http://lists.racket-lang.org/users/listinfo
> or, via email, send a message with subject or body 'help' to
>         users-request at racket-lang.org
>
> You can reach the person managing the list at
>         users-owner at racket-lang.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of users digest..."
>
>
> [Racket Users list:
>  http://lists.racket-lang.org/users ]
>
>
> Today's Topics:
>
>    1. Re: Help debugging a ffi crash (Eric Dobson)
>    2. How to document a field? (Roman Klochkov)
>    3. Re: proof assistants, DrRacket and Bootstrap (Bill Richter)
>    4. Re: aws/glacier: credential scope (Greg Hendershott)
>    5. Re: aws/glacier: credential scope (Norman Gray)
>    6. Re: aws/glacier: credential scope (Greg Hendershott)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Sun, 28 Sep 2014 11:15:33 -0700
> From: Eric Dobson <eric.n.dobson at gmail.com>
> To: Matthew Flatt <mflatt at cs.utah.edu>
> Cc: "users at racket-lang.org" <users at racket-lang.org>
> Subject: Re: [racket] Help debugging a ffi crash
> Message-ID:
>         <CAAEHQ5tWAKzi5AizADUH9smVsNXpYTbmxtO3teGm9Zj8ZDUnLg at mail.gmail.com>
> Content-Type: text/plain; charset=UTF-8
>
> That is almost surely it, thanks for the second pair of eyes. I had
> similar issues with a different type that I thought was a pointer but
> was actually a struct, but that one I couldn't even get a single call
> to work so it was much more obvious something was up.
>
> On Sun, Sep 28, 2014 at 11:08 AM, Matthew Flatt <mflatt at cs.utah.edu> wrote:
>> Looking at
>>
>>   http://clang.llvm.org/doxygen/CXString_8h_source.html
>>
>> it seems that CXString as returned by clang_getCursorSpelling() is not
>> a pointer:
>>
>>  typedef struct {
>>    const void *data;
>>    unsigned private_flags;
>>  } CXString;
>>
>> If that's right, I'm a little surprised that `cursor-spelling` works
>> --- but when you get representation wrong, strange things can happen,
>> including something working when it shouldn't.
>>
>> Am I looking at the right library/definitions?
>>
>> At Sun, 28 Sep 2014 10:48:06 -0700, Eric Dobson wrote:
>>> I'm trying to debug an FFI crash that I'm seeing, and because it is
>>> dealing with C code the error just presents as a segfault. I believe I
>>> have tracked down what is causing the problem, but don't understand
>>> how it could be doing so.
>>>
>>> I have two racket functions which take a "cursor" (the foreign
>>> libraries object) and return a string representation of it, which I'm
>>> trying to use for debugging.
>>>
>>> (define raw-clang-get-cstring
>>>   (get-ffi-obj "clang_getCString" lib-clang
>>>     (_fun _pointer -> _string)))
>>>
>>> (define raw-cursor-spelling
>>>  (get-ffi-obj "clang_getCursorSpelling" lib-clang
>>>    (_fun _CXCursor -> _pointer)))
>>>
>>> (define (cursor-spelling c)
>>>   (raw-clang-get-cstring (raw-cursor-spelling c)))
>>>
>>> (define cursor-spelling2
>>>  (get-ffi-obj "clang_getCursorSpelling" lib-clang
>>>    (_fun _CXCursor -> (make-ctype _pointer values (? (v)
>>> (raw-clang-get-cstring v))))))
>>>
>>> If I use cursor-spelling, I have not been able to trigger a crash. But
>>> if I use cursor-spelling2 I can reliably trigger a crash.
>>>
>>> Is there anything obvious on how these functions are different?
>>> Because they look to me like they should be doing the same thing. If
>>> it would be helpful I can try to get my code in a portable enough
>>> shape so that it will work/crash on another machine.
>>>
>>> ____________________
>>>   Racket Users list:
>>>   http://lists.racket-lang.org/users
>
>
>
> ------------------------------
>
> Message: 2
> Date: Sun, 28 Sep 2014 22:39:54 +0400
> From: Roman Klochkov <kalimehtar at mail.ru>
> To: racket users list <users at racket-lang.org>
> Subject: [racket] How to document a field?
> Message-ID: <1411929594.923492574 at f67.i.mail.ru>
> Content-Type: text/plain; charset="utf-8"
>
>  At  http://docs.racket-lang.org/scribble/doc-classes.html I see classes, interfaces, methods, but don't see fields.
>
> How to document them? Via @defthing or is there some syntax with (get-field ..) (set-field! ...) template like @defparam ?
>
>
> --
> Roman Klochkov
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140928/ac478807/attachment-0001.html>
>
> ------------------------------
>
> Message: 3
> Date: Sun, 28 Sep 2014 13:51:26 -0500
> From: Bill Richter <richter at math.northwestern.edu>
> Cc: users at racket-lang.org
> Subject: Re: [racket] proof assistants, DrRacket and Bootstrap
> Message-ID:
>         <201409281851.s8SIpQMx006764 at poisson.math.northwestern.edu>
>
> Matthias & Prabhakar, I think I see what you're saying now.  I need the camlp parser to turn my dialect-expressions into trees.  Then I turn my  tree into standard expressions just as one does in Scheme.  The only differences is that in Scheme we have the quote function that skips the parsing step.  So your ML code
>
> MF>   type ast = VAR of String | LAM of String * ast | APP of ast * ast | CONST of int | ADD of ast * ast
> MF>   type val = BASIC of int | FUNCTION of val -> val | ...
>
> looks a lot like the HOL Light definition of preterms:
>
> type preterm = Varp of string * pretype       (* Variable           - v      *)
>              | Constp of string * pretype     (* Constant           - c      *)
>              | Combp of preterm * preterm     (* Combination        - f x    *)
>              | Absp of preterm * preterm      (* Lambda-abstraction - \x. t  *)
>              | Typing of preterm * pretype;;  (* Type constraint    - t : ty *)
>
> and so we can see one of these preterms in a standard term:
>
> # preterm_of_term `x + y`;;
> val it : preterm =
>   Combp
>    (Combp
>      (Constp ("+",
>        Ptycon ("fun",
>         [Ptycon ("num", []);
>          Ptycon ("fun", [Ptycon ("num", []); Ptycon ("num", [])])])),
>      Varp ("x", Ptycon ("num", []))),
>    Varp ("y", Ptycon ("num", [])))
>
> So the Scheme story works fine except that I haven't understood camlp or how HOL Light turns expressions `[...]` into preterms.  But I feel better about investing the time into learning camlp now that I see the Scheme story remains intact in HOL Light. And you're right  about my ignorance:
>
> MF> 1. HtDP explains the above in a couple of sections.
>
> Right, there's something I seem not to have understood in either Scheme or ML, how when we create and evaluate our trees we retain the values of our variables.  I can figure that out, and HtDP sounds like a good place to start.
>
> PR> A Racketeer would probably tell you to use macros to define your
> PR> dialect, instead of using an eval hack. The corresponding tool for
> PR> OCaml is camlp4, and it sounds as if it would be worth your time
> PR> to learn it thoroughly.
>
> That's an interesting comparison, and since macros, maybe I can learn to like camlp.  In fact, maybe I ought to learn something about racket macros first.
>
> --
> Best,
> Bill
>
>
> ------------------------------
>
> Message: 4
> Date: Sun, 28 Sep 2014 16:01:20 -0400
> From: Greg Hendershott <greghendershott at gmail.com>
> To: Norman Gray <norman at astro.gla.ac.uk>
> Cc: users at racket-lang.org
> Subject: Re: [racket] aws/glacier: credential scope
> Message-ID:
>         <CAGspUn0tqgSVUaDuk1C3Wb6hm6s6nQ2ps-j42pkJJSY2UESL0A at mail.gmail.com>
> Content-Type: text/plain; charset=UTF-8
>
> Hi, Norman.
>
> I logged an issue for this:
>
>   https://github.com/greghendershott/aws/issues/32
>
> I see the problem (or at least the main problem) and will push a fix.
>
> The bug is embarrassing, not just because it's such a silly mistake,
> but it's something a unit test could have caught. (I could say that
> doing unit tests for Glacier is challenging, because the retrieval
> process can take hours. Although that's true, I could have done more
> to test _some_ operations working among various regions.)
>
> On Sat, Sep 27, 2014 at 4:14 PM, Norman Gray <norman at astro.gla.ac.uk> wrote:
>>
>> Greetings.
>>
>> I'm trying to use the aws/glacier package, and running into a problem where I'm being told:
>>
>>     Credential should be scoped to a valid region, not 'eu-west-1'
>>
>> I'm following the instructions at <https://github.com/greghendershott/aws/blob/master/aws/manual.md>
>>
>> My test code is:
>>
>>     % cat glacier.rkt
>>     #lang racket/base
>>
>>     (require aws/glacier
>>              aws/keys)
>>
>>     (define vault "testvault")
>>     (region "eu-west-1")
>>     (read-keys "aws-zbu-credentials") ; local file
>>
>>     (module+ main
>>       (printf "region=~a~%" (region))
>>       (printf "Vaults: ~s~%" (list-vaults))
>>       (printf "...specifically: ~s~%" (describe-vault vault)))
>>
>> Running this produces:
>>
>>     % racket glacier.rkt
>>     region=eu-west-1
>>     aws: HTTP/1.1 403 Forbidden
>>     x-amzn-RequestId: Un3-L2zlaJBPyrIVKJrWuQcqtMMYQAr34gYUOSScg6Qepc4
>>     Content-Type: application/json
>>     Content-Length: 129
>>     Date: Sat, 27 Sep 2014 18:35:50 GMT
>>
>>      {"message":"Credential should be scoped to a valid region, not 'eu-west-1'. ","code":"InvalidSignatureException","type":"Client"}
>>     HTTP 403 "Forbidden". AWS Code="InvalidSignatureException" Message="Credential should be scoped to a valid region, not 'eu-west-1'. "
>>       context...:
>>        check-response
>>        /Users/norman/Library/Racket/6.1/pkgs/aws/aws/glacier.rkt:97:22: temp68
>>        request/redirect/uri
>>        (submod /checkouts/me/code/zbu/glacier.rkt main): [running body]
>>
>> Things I thought of:
>>
>>   * Printing (public-key)/(private-key) indicates that the credentials are being read correctly.
>>   * When I change the argument of (region) to "us-west-1", that's the region that appears in the error message.
>>   * My "testvault" vault is in eu-west-1 (and this is indeed one of the valid regions for glacier, reported in <http://docs.aws.amazon.com/general/latest/gr/rande.html> and which does have a host at http://glacier.eu-west-1.amazonaws.com
>>   * As far as I can see, credentials are _not_ scoped, but are all at us-east-1.
>>   * <http://docs.aws.amazon.com/general/latest/gr/signature-v4-troubleshooting.html> says that "IAM [...] accepts only us-east-1 as its region specification", so I'm taking it that (region) is for setting the _vault_'s region.
>>   * I'm not a great AWS expert, so I could have something in my setup broken; but if so, I've no clue what.
>>
>> If, however, I change the (region) argument to "us-east-1", I get a different error message "User: arn:aws:iam::786725553169:user/zbu is not authorized to perform: glacier:ListVaults on resource: arn:aws:glacier:us-east-1:786725553169:vaults/"  That makes sense, since there's no such vault, but it's interesting that it gets _further_ when the (region) matches the region for the IAM service.
>>
>> I don't see any other (region) equivalents for the other services supported by the package.  Is that because all of the other services supported by the package are supported by all the AWS regions, or am I missing a configuration?
>>
>> Thanks for any pointers.
>>
>> All the best,
>>
>> Norman
>>
>>
>> --
>> Norman Gray  :  http://nxg.me.uk
>> SUPA School of Physics and Astronomy, University of Glasgow, UK
>>
>>
>> ____________________
>>   Racket Users list:
>>   http://lists.racket-lang.org/users
>
>
>
> ------------------------------
>
> Message: 5
> Date: Sun, 28 Sep 2014 21:29:03 +0100
> From: Norman Gray <norman at astro.gla.ac.uk>
> To: Greg Hendershott <greghendershott at gmail.com>
> Cc: users at racket-lang.org
> Subject: Re: [racket] aws/glacier: credential scope
> Message-ID: <FE320180-07AE-43E8-AE51-3AA566329E67 at astro.gla.ac.uk>
> Content-Type: text/plain; charset=us-ascii
>
>
> Greg, hello.
>
> On 2014 Sep 28, at 21:01, Greg Hendershott <greghendershott at gmail.com> wrote:
>
>> I logged an issue for this:
>>
>>  https://github.com/greghendershott/aws/issues/32
>
> Ah: I wondered if the problem might be related to that.
>
>> The bug is embarrassing, not just because it's such a silly mistake,
>> but it's something a unit test could have caught. (I could say that
>> doing unit tests for Glacier is challenging, because the retrieval
>> process can take hours.
>
> I can imagine how protracted that would be ... *shudder*.
>
> Thanks for looking at this.
>
> Also, Frank:
>
>> What happens if you read-keys before setting the region?
>
> Thanks for this suggestion: I did try that just a little earlier, but it didn't make any difference.
>
> All the best,
>
> Norman
>
>
> --
> Norman Gray  :  http://nxg.me.uk
> SUPA School of Physics and Astronomy, University of Glasgow, UK
>
>
>
>
> ------------------------------
>
> Message: 6
> Date: Sun, 28 Sep 2014 16:29:12 -0400
> From: Greg Hendershott <greghendershott at gmail.com>
> To: Norman Gray <norman at astro.gla.ac.uk>
> Cc: users at racket-lang.org
> Subject: Re: [racket] aws/glacier: credential scope
> Message-ID:
>         <CAGspUn2au5xydsH+o9E2Y3JTA1iQbPRX58zRcitywAPWh5Djng at mail.gmail.com>
> Content-Type: text/plain; charset=UTF-8
>
> I pushed a fix, including running the tests across a few different AWS
> regions. They pass.
>
> I clicked "update" on pkgs.racket-lang.org. But it seems to be slower
> than usual to refresh. After it does, you can `raco pkg update aws` to
> get the fix.
>
> On Sun, Sep 28, 2014 at 4:01 PM, Greg Hendershott
> <greghendershott at gmail.com> wrote:
>> Hi, Norman.
>>
>> I logged an issue for this:
>>
>>   https://github.com/greghendershott/aws/issues/32
>>
>> I see the problem (or at least the main problem) and will push a fix.
>>
>> The bug is embarrassing, not just because it's such a silly mistake,
>> but it's something a unit test could have caught. (I could say that
>> doing unit tests for Glacier is challenging, because the retrieval
>> process can take hours. Although that's true, I could have done more
>> to test _some_ operations working among various regions.)
>>
>> On Sat, Sep 27, 2014 at 4:14 PM, Norman Gray <norman at astro.gla.ac.uk> wrote:
>>>
>>> Greetings.
>>>
>>> I'm trying to use the aws/glacier package, and running into a problem where I'm being told:
>>>
>>>     Credential should be scoped to a valid region, not 'eu-west-1'
>>>
>>> I'm following the instructions at <https://github.com/greghendershott/aws/blob/master/aws/manual.md>
>>>
>>> My test code is:
>>>
>>>     % cat glacier.rkt
>>>     #lang racket/base
>>>
>>>     (require aws/glacier
>>>              aws/keys)
>>>
>>>     (define vault "testvault")
>>>     (region "eu-west-1")
>>>     (read-keys "aws-zbu-credentials") ; local file
>>>
>>>     (module+ main
>>>       (printf "region=~a~%" (region))
>>>       (printf "Vaults: ~s~%" (list-vaults))
>>>       (printf "...specifically: ~s~%" (describe-vault vault)))
>>>
>>> Running this produces:
>>>
>>>     % racket glacier.rkt
>>>     region=eu-west-1
>>>     aws: HTTP/1.1 403 Forbidden
>>>     x-amzn-RequestId: Un3-L2zlaJBPyrIVKJrWuQcqtMMYQAr34gYUOSScg6Qepc4
>>>     Content-Type: application/json
>>>     Content-Length: 129
>>>     Date: Sat, 27 Sep 2014 18:35:50 GMT
>>>
>>>      {"message":"Credential should be scoped to a valid region, not 'eu-west-1'. ","code":"InvalidSignatureException","type":"Client"}
>>>     HTTP 403 "Forbidden". AWS Code="InvalidSignatureException" Message="Credential should be scoped to a valid region, not 'eu-west-1'. "
>>>       context...:
>>>        check-response
>>>        /Users/norman/Library/Racket/6.1/pkgs/aws/aws/glacier.rkt:97:22: temp68
>>>        request/redirect/uri
>>>        (submod /checkouts/me/code/zbu/glacier.rkt main): [running body]
>>>
>>> Things I thought of:
>>>
>>>   * Printing (public-key)/(private-key) indicates that the credentials are being read correctly.
>>>   * When I change the argument of (region) to "us-west-1", that's the region that appears in the error message.
>>>   * My "testvault" vault is in eu-west-1 (and this is indeed one of the valid regions for glacier, reported in <http://docs.aws.amazon.com/general/latest/gr/rande.html> and which does have a host at http://glacier.eu-west-1.amazonaws.com
>>>   * As far as I can see, credentials are _not_ scoped, but are all at us-east-1.
>>>   * <http://docs.aws.amazon.com/general/latest/gr/signature-v4-troubleshooting.html> says that "IAM [...] accepts only us-east-1 as its region specification", so I'm taking it that (region) is for setting the _vault_'s region.
>>>   * I'm not a great AWS expert, so I could have something in my setup broken; but if so, I've no clue what.
>>>
>>> If, however, I change the (region) argument to "us-east-1", I get a different error message "User: arn:aws:iam::786725553169:user/zbu is not authorized to perform: glacier:ListVaults on resource: arn:aws:glacier:us-east-1:786725553169:vaults/"  That makes sense, since there's no such vault, but it's interesting that it gets _further_ when the (region) matches the region for the IAM service.
>>>
>>> I don't see any other (region) equivalents for the other services supported by the package.  Is that because all of the other services supported by the package are supported by all the AWS regions, or am I missing a configuration?
>>>
>>> Thanks for any pointers.
>>>
>>> All the best,
>>>
>>> Norman
>>>
>>>
>>> --
>>> Norman Gray  :  http://nxg.me.uk
>>> SUPA School of Physics and Astronomy, University of Glasgow, UK
>>>
>>>
>>> ____________________
>>>   Racket Users list:
>>>   http://lists.racket-lang.org/users
>
>
>
> End of users Digest, Vol 109, Issue 70
> **************************************


From moshedeutsch115 at gmail.com  Mon Sep 29 16:42:09 2014
From: moshedeutsch115 at gmail.com (Moshe Deutsch)
Date: Mon, 29 Sep 2014 16:42:09 -0400
Subject: [racket] users Digest, Vol 109, Issue 69
In-Reply-To: <mailman.476.1411927713.679.users@racket-lang.org>
References: <mailman.476.1411927713.679.users@racket-lang.org>
Message-ID: <CAKpSqLCBYk-_1HSjmN-p34B_HCa0wWtaEyBOD8usVOd_26T59Q@mail.gmail.com>

Please take me off the list

Thanks

On Sun, Sep 28, 2014 at 2:08 PM,  <users-request at racket-lang.org> wrote:
> Send users mailing list submissions to
>         users at racket-lang.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>         http://lists.racket-lang.org/users/listinfo
> or, via email, send a message with subject or body 'help' to
>         users-request at racket-lang.org
>
> You can reach the person managing the list at
>         users-owner at racket-lang.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of users digest..."
>
>
> [Racket Users list:
>  http://lists.racket-lang.org/users ]
>
>
> Today's Topics:
>
>    1. Re: typed racket, filters, and polymorphism (Alexander D. Knauth)
>    2. Help debugging a ffi crash (Eric Dobson)
>    3. Re: [GENERAL] Off Topic: Anybody reading this via
>       news.gmane.org? (Adrian Klaver)
>    4. Re: Help debugging a ffi crash (Matthew Flatt)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Sun, 28 Sep 2014 13:37:30 -0400
> From: "Alexander D. Knauth" <alexander at knauth.org>
> To: Spencer florence <spencerflorence at gmail.com>
> Cc: racket users list <users at racket-lang.org>, Sam Tobin-Hochstadt
>         <samth at cs.indiana.edu>
> Subject: Re: [racket] typed racket, filters, and polymorphism
> Message-ID: <B849BCC0-7F5C-402E-B98C-01F97E1762D8 at knauth.org>
> Content-Type: text/plain; charset="utf-8"
>
> No because I want the unit to be a struct that has a dimension field, not a symbol with various dimensions defined as unions of units.
> I want the unit to be based on the dimension, not the other way around, so that new units can be made that have the same dimension.
>
> I have something like the number+unit struct (I called it measure), but I?ll work on more that after I have the unit struct figured out.
>
> On Sep 28, 2014, at 12:13 PM, Spencer florence <spencerflorence at gmail.com> wrote:
>
>> would something like this work?
>>
>> #lang typed/racket
>>
>> (struct (U) number+unit ([amount : Real] [unit : U]))
>>
>> (define-type Weight-Unit (U 'kg 'g 'mg '?g))
>> (define-type Weight (number+unit Weight-Unit))
>> (define-predicate weight? Weight)
>>
>> (: make-weight : Real Weight-Unit -> Weight)
>> (define (make-weight n u)
>>   (number+unit n u))
>>
>> (: +/weight : Weight Weight -> Weight)
>> ;; something something needs unit conversion
>> (define (+/weight w1 w2)
>>   (number+unit (+ (number+unit-amount w1)
>>                   (number+unit-amount w1))
>>                (number+unit-unit w1)))
>>
>> (+/weight (make-weight 1 'kg) (make-weight 1 'kg))
>>
>>
>>
>> On Sun, Sep 28, 2014 at 11:03 AM, Alexander D. Knauth <alexander at knauth.org> wrote:
>>
>> Because the struct is representing a unit (kilograms, meters, seconds, etc.), and a unit has a dimension (mass, length, time, etc.) and I want the type-checker to be able to know what the dimension of a unit is so that the types of functions can specify the dimension that something should have.
>> The real solution to this would probably be bounded polymorphism, but I was wondering if there was some other way to do it with occurrence typing in the guard or something like that.
>>
>> On Sep 28, 2014, at 11:48 AM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> wrote:
>>
>> > Why not do this with the type, instead of making this polymorphic?
>> >
>> > Sam
>> >
>> > On Fri, Sep 26, 2014 at 7:35 PM, Alexander D. Knauth
>> > <alexander at knauth.org> wrote:
>> >> Is it possible to have a struct that does certain things according to the
>> >> guard?
>> >> #lang typed/racket
>> >>
>> >> (struct (a) foo ([a : a]) #:transparent
>> >> #:guard (lambda (a _)
>> >> (unless (exact-integer? a)
>> >> (error 'foo "expected Integer, given ~v" a))
>> >> a))
>> >>
>> >> (ann (foo (ann 1 Any)) (foo Integer))
>> >>
>> >> (: x : (foo Any))
>> >> (define x (foo 1))
>> >>
>> >> (ann (foo-a x) Integer)
>> >>
>> >> ;. Type Checker: Polymorphic function `foo1' could not be applied to
>> >> arguments:
>> >> ;Argument 1:
>> >> ; Expected: a
>> >> ; Given: Any
>> >> ;
>> >> ;Result type: (foo a)
>> >> ;Expected result: (foo Integer)
>> >> ; in: (foo (ann 1 Any))
>> >> ;. Type Checker: Polymorphic function `foo-a' could not be applied to
>> >> arguments:
>> >> ;Argument 1:
>> >> ; Expected: (foo a)
>> >> ; Given: (foo Any)
>> >> ;
>> >> ;Result type: (a : ....)
>> >> ;Expected result: Integer
>> >> ; in: (foo-a x)
>> >>
>> >> On Sep 25, 2014, at 9:42 PM, Alexander D. Knauth <alexander at knauth.org>
>> >> wrote:
>> >>
>> >> What I?m trying to accomplish is something more like this:
>> >> #lang typed/racket
>> >>
>> >> (require "dimensions.rkt")
>> >>
>> >> (struct (d) unit ([name : Any] [scalar : Positive-Real] [dimension : d])
>> >> #:transparent
>> >> #:guard (lambda (name scalar dimension _)
>> >> (unless (dimension? dimension)
>> >> (error 'unit "expected Dimension, given ~v" dimension))
>> >> (values name scalar dimension)))
>> >>
>> >> (define-type (Unitof d) (unit d))
>> >>
>> >> (define-type Unit (Unitof Dimension))
>> >>
>> >> (define Unit? (make-predicate Unit))
>> >>
>> >> (define-type Unitish
>> >> (U (Unitof Any)
>> >> Dimension
>> >> Positive-Real))
>> >>
>> >> (: ->unit : (All (d) (case-> [(Unitof d) -> (Unitof d)]
>> >> [Unitish -> Unit])))
>> >> (define (->unit u)
>> >> (cond [(unit? u)
>> >> (unless (Unit? u) ; this should never happen anyway because of the
>> >> guard
>> >> (error '->unit "expected (Unitof Dimension), given ~v" u))
>> >> u]
>> >> [(dimension? u) (unit u 1 u)]
>> >> [(positive-real? u) (unit u u dimensionless-dimension)]))
>> >>
>> >>
>> >> On Sep 25, 2014, at 6:19 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu>
>> >> wrote:
>> >>
>> >> No, I don't think you can do this. Can you say more about what you're
>> >> trying to accomplish?
>> >>
>> >> Sam
>> >>
>> >> On Thu, Sep 25, 2014 at 6:15 PM, Alexander D. Knauth
>> >> <alexander at knauth.org> wrote:
>> >>
>> >> Do any of you have any advice for getting a function like this to
>> >> type-check?
>> >> #lang typed/racket
>> >>
>> >> (: check-int : (All (a) (case-> [a -> a]
>> >> [Any -> Integer])))
>> >> (define (check-int int)
>> >> (unless (exact-integer? int)
>> >> (error 'check-int "expected Integer, given ~v" int))
>> >> int)
>> >>
>> >> ;. Type Checker: type mismatch
>> >> ; expected: a
>> >> ; given: Integer in: int
>> >>
>> >>
>> >>
>> >> ____________________
>> >> Racket Users list:
>> >> http://lists.racket-lang.org/users
>> >>
>> >>
>> >> ____________________
>> >> Racket Users list:
>> >> http://lists.racket-lang.org/users
>> >>
>> >>
>>
>>
>> ____________________
>> Racket Users list:
>> http://lists.racket-lang.org/users
>>
>>
>
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140928/56742b6b/attachment-0001.html>
>
> ------------------------------
>
> Message: 2
> Date: Sun, 28 Sep 2014 10:48:06 -0700
> From: Eric Dobson <eric.n.dobson at gmail.com>
> To: "users at racket-lang.org" <users at racket-lang.org>
> Subject: [racket] Help debugging a ffi crash
> Message-ID:
>         <CAAEHQ5uRU5V4i1rvNf6sdfcbQsD087ntCwftjRVhKroNErmUXg at mail.gmail.com>
> Content-Type: text/plain; charset=UTF-8
>
> I'm trying to debug an FFI crash that I'm seeing, and because it is
> dealing with C code the error just presents as a segfault. I believe I
> have tracked down what is causing the problem, but don't understand
> how it could be doing so.
>
> I have two racket functions which take a "cursor" (the foreign
> libraries object) and return a string representation of it, which I'm
> trying to use for debugging.
>
> (define raw-clang-get-cstring
>   (get-ffi-obj "clang_getCString" lib-clang
>     (_fun _pointer -> _string)))
>
> (define raw-cursor-spelling
>  (get-ffi-obj "clang_getCursorSpelling" lib-clang
>    (_fun _CXCursor -> _pointer)))
>
> (define (cursor-spelling c)
>   (raw-clang-get-cstring (raw-cursor-spelling c)))
>
> (define cursor-spelling2
>  (get-ffi-obj "clang_getCursorSpelling" lib-clang
>    (_fun _CXCursor -> (make-ctype _pointer values (? (v)
> (raw-clang-get-cstring v))))))
>
> If I use cursor-spelling, I have not been able to trigger a crash. But
> if I use cursor-spelling2 I can reliably trigger a crash.
>
> Is there anything obvious on how these functions are different?
> Because they look to me like they should be doing the same thing. If
> it would be helpful I can try to get my code in a portable enough
> shape so that it will work/crash on another machine.
>
>
>
> ------------------------------
>
> Message: 3
> Date: Thu, 25 Sep 2014 17:25:15 -0700
> From: Adrian Klaver <adrian.klaver at aklaver.com>
> To: George Neuner <gneuner2 at comcast.net>, neil at neilvandyke.org
> Cc: "users at racket-lang.org >> users" <users at racket-lang.org>
> Subject: Re: [racket] [GENERAL] Off Topic: Anybody reading this via
>         news.gmane.org?
> Message-ID: <5424B26B.9080504 at aklaver.com>
> Content-Type: text/plain; charset=windows-1252; format=flowed
>
> On 09/25/2014 05:13 PM, George Neuner wrote:
>>
>> On 9/25/2014 5:26 PM, Neil Van Dyke wrote:
>>> You can check your IP addr against the list at
>>> "http://gmane.org/denied.php".
>>>
>>> Neil V.
>>
>> On 9/25/2014 7:08 PM, Adrian Klaver wrote:
>>>
>>> Take a look here:
>>>
>>> http://gmane.org/denied.php
>>>
>>> My guess is you are the fourth one from the bottom.
>>>
>>>
>>> Might want to take a look at this thread to see what your options are
>>> and what the turn around time is on your request:
>>>
>>> http://thread.gmane.org/gmane.discuss/16309
>>
>> Thanks to both of you - I missed seeing the entry about the denied list
>> in the FAQ.
>>
>> Adrian you were right - it seems that I have been blocked for some
>> reason.  Based on the description of infractions I really don't think I
>> did anything to warrant it ... but there it is.  I'll have to follow my
>> lists by email until I can unblocked.
>
> I think the relevant part at gmane.org/denied.php is:
>
> "Since there is no way to identify users, news readers are denied on a
> domain/IP basis. So if somebody on a machine close to you downloaded the
> news spool yesterday, you're likely to be denied today"
>
>>
>> Thanks again,
>> George
>
>
> --
> Adrian Klaver
> adrian.klaver at aklaver.com
>
>
> ------------------------------
>
> Message: 4
> Date: Sun, 28 Sep 2014 12:08:23 -0600
> From: Matthew Flatt <mflatt at cs.utah.edu>
> To: Eric Dobson <eric.n.dobson at gmail.com>
> Cc: "users at racket-lang.org" <users at racket-lang.org>
> Subject: Re: [racket] Help debugging a ffi crash
> Message-ID: <20140928180825.80D74650196 at mail-svr1.cs.utah.edu>
> Content-Type: text/plain; charset=UTF-8
>
> Looking at
>
>   http://clang.llvm.org/doxygen/CXString_8h_source.html
>
> it seems that CXString as returned by clang_getCursorSpelling() is not
> a pointer:
>
>  typedef struct {
>    const void *data;
>    unsigned private_flags;
>  } CXString;
>
> If that's right, I'm a little surprised that `cursor-spelling` works
> --- but when you get representation wrong, strange things can happen,
> including something working when it shouldn't.
>
> Am I looking at the right library/definitions?
>
> At Sun, 28 Sep 2014 10:48:06 -0700, Eric Dobson wrote:
>> I'm trying to debug an FFI crash that I'm seeing, and because it is
>> dealing with C code the error just presents as a segfault. I believe I
>> have tracked down what is causing the problem, but don't understand
>> how it could be doing so.
>>
>> I have two racket functions which take a "cursor" (the foreign
>> libraries object) and return a string representation of it, which I'm
>> trying to use for debugging.
>>
>> (define raw-clang-get-cstring
>>   (get-ffi-obj "clang_getCString" lib-clang
>>     (_fun _pointer -> _string)))
>>
>> (define raw-cursor-spelling
>>  (get-ffi-obj "clang_getCursorSpelling" lib-clang
>>    (_fun _CXCursor -> _pointer)))
>>
>> (define (cursor-spelling c)
>>   (raw-clang-get-cstring (raw-cursor-spelling c)))
>>
>> (define cursor-spelling2
>>  (get-ffi-obj "clang_getCursorSpelling" lib-clang
>>    (_fun _CXCursor -> (make-ctype _pointer values (? (v)
>> (raw-clang-get-cstring v))))))
>>
>> If I use cursor-spelling, I have not been able to trigger a crash. But
>> if I use cursor-spelling2 I can reliably trigger a crash.
>>
>> Is there anything obvious on how these functions are different?
>> Because they look to me like they should be doing the same thing. If
>> it would be helpful I can try to get my code in a portable enough
>> shape so that it will work/crash on another machine.
>>
>> ____________________
>>   Racket Users list:
>>   http://lists.racket-lang.org/users
>
>
>
> End of users Digest, Vol 109, Issue 69
> **************************************


From moshedeutsch115 at gmail.com  Mon Sep 29 16:42:21 2014
From: moshedeutsch115 at gmail.com (Moshe Deutsch)
Date: Mon, 29 Sep 2014 16:42:21 -0400
Subject: [racket] users Digest, Vol 109, Issue 65
In-Reply-To: <mailman.466.1411866688.679.users@racket-lang.org>
References: <mailman.466.1411866688.679.users@racket-lang.org>
Message-ID: <CAKpSqLBrRnupUkDvdo9LFxBgD4tgLq2Wjp-Bn-2HbFuqcVF=pw@mail.gmail.com>

Please take me off the list

Thanks

On Sat, Sep 27, 2014 at 9:11 PM,  <users-request at racket-lang.org> wrote:
> Send users mailing list submissions to
>         users at racket-lang.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>         http://lists.racket-lang.org/users/listinfo
> or, via email, send a message with subject or body 'help' to
>         users-request at racket-lang.org
>
> You can reach the person managing the list at
>         users-owner at racket-lang.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of users digest..."
>
>
> [Racket Users list:
>  http://lists.racket-lang.org/users ]
>
>
> Today's Topics:
>
>    1. Re: proof assistants, DrRacket and Bootstrap (Bill Richter)
>    2. Re: proof assistants, DrRacket and Bootstrap (Prabhakar Ragde)
>    3. Re: proof assistants, DrRacket and Bootstrap (Bill Richter)
>    4. Re: problem with pict->bitmap and mrlib/write-animated-gif
>       (Martin DeMello)
>    5. Re: proof assistants, DrRacket and Bootstrap (Prabhakar Ragde)
>    6. Re: typed racket and generic interfaces, is there a
>       workaround using properties? (Anthony Carrico)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Sat, 27 Sep 2014 16:50:11 -0500
> From: Bill Richter <richter at math.northwestern.edu>
> To: Prabhakar Ragde <plragde at uwaterloo.ca>
> Cc: users at racket-lang.org
> Subject: Re: [racket] proof assistants, DrRacket and Bootstrap
> Message-ID:
>         <201409272150.s8RLoBQW030663 at poisson.math.northwestern.edu>
>
> Thanks, Prabhakar, this is exactly what I wanted, and you know something about it (I was barely able to install Coq):
>
>    I am teaching a grad course using Coq right now, and using Proof
>    General within Emacs for it. That is not bad, but I would love to
>    have a DrRacket interface.
>
>    > OCaml doesn't have a quote feature, so the question arises how to
>    > write an OCaml interpreter inside OCaml.
>
>    In my senior undergraduate programming language course, I have them
>    write various interpreters and typecheckers in various languages,
>    including Racket, OCaml, and Haskell.
>    [...]
>
> I know this is way off-topic, but it seems to me that Schemers are the only people who want an elegant solution here:
>
>    But parsing an expression written in faux OCaml/Haskell is quite
>    complicated. [...] Parsing is much easier in Racket/Scheme because
>    of the similarity of code/data and the presence of 'read'.
>
> I think it's a parsing question I'm stuck on.  I think I need to use camlp to avoid the following hack/kludge which I learned from the HOL Light experts, found in my file
> hol_light/RichterHilbertAxiomGeometry/readable.ml
> which is part of the HOL Light distribution:
>
> (* From update_database.ml: Execute any OCaml expression given as a string.  *)
>
> let exec = ignore o Toploop.execute_phrase false Format.std_formatter
>   o !Toploop.parse_toplevel_phrase o Lexing.from_string;;
>
> (* Following miz3.ml, exec_thm returns the theorem representing a string, so *)
> (* exec_thm "FORALL_PAIR_THM";; returns                                      *)
> (* val it : thm = |- !P. (!p. P p) <=> (!p1 p2. P (p1,p2))                   *)
>
> let thm_ref = ref TRUTH;;
>
> let exec_thm s =
>   if Str.string_match (Str.regexp "[^;]*;") s 0 then raise Noparse
>   else
>     try exec ("thm_ref := (("^ s ^"): thm);;");
>       !thm_ref
>     with _ -> raise Noparse;;
>
> When I posted my question on the OCaml list, the experts said don't ever use Obj.magic, but I think they didn't think much of me using Toploop either.
>
> --
> Best,
> Bill
>
>
> ------------------------------
>
> Message: 2
> Date: Sat, 27 Sep 2014 18:00:24 -0400
> From: Prabhakar Ragde <plragde at uwaterloo.ca>
> To: users at racket-lang.org
> Subject: Re: [racket] proof assistants, DrRacket and Bootstrap
> Message-ID: <54273378.1070605 at uwaterloo.ca>
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>
> Bill Richter wrote:
>
>> >Carl Eastlund developed Dracula for Rackety use of the ACL2 theorem
>> >prover.
>>
>> Thanks, Ian!  May I suggest that you try to handle Coq, Isabelle or
>> HOL Light?  I believe these are the foremost proof assistants.  The
>> fields medalist Vladimir Voevodsky uses Coq, in which the 4-color
>> theorem and the Feit-Thompson theorem was formalized by Georges
>> Gonthier.  Tom Hales just finished formalizing his proof of Kepler
>> conjecture in HOL Light and Isabelle (which is HOL as well).  I'm no
>> expert, but I think that ACL2 (also Prover9) is an FOL prover, and
>> the bulk of activity in formal proofs uses type theories.  HOL is
>> simple, it's Church's simple types (a version of the Lambda
>> Calculus), and Coq uses a much more complicated type theory.
>
> I'm sorry, I forgot about Carl's Dracula work when composing my earlier
> reply. ACL2 is more limited, but its advantage (besides being probably
> far and away the theorem prover that has had the most industrial impact)
> is that its language (Applicative Common Lisp, hence the acronym) is
> much closer to Racket, allowing for stronger integration both with
> DrRacket and into a Racket-based curriculum. A DrRacket interface to Coq
> or Isabelle would, in contrast, be more superficial, in the style of
> Proof General. --PR
>
>
> ------------------------------
>
> Message: 3
> Date: Sat, 27 Sep 2014 17:26:44 -0500
> From: Bill Richter <richter at math.northwestern.edu>
> To: Prabhakar Ragde <plragde at uwaterloo.ca>
> Cc: users at racket-lang.org
> Subject: Re: [racket] proof assistants, DrRacket and Bootstrap
> Message-ID:
>         <201409272226.s8RMQiEM030897 at poisson.math.northwestern.edu>
>
>    ACL2 [is] probably far and away the theorem prover that has had the
>    most industrial impact
>
> Interesting, Prabhakar!  I had not heard that before.
>
>    A DrRacket interface to Coq or Isabelle would, in contrast, be more
>    superficial, in the style of Proof General.
>
> I think so, but interfaces are really important, and DrRacket has a beautiful interface.
>
> --
> Best,
> Bill
>
>
> ------------------------------
>
> Message: 4
> Date: Sat, 27 Sep 2014 15:46:25 -0700
> From: Martin DeMello <martindemello at gmail.com>
> To: Racket Users List <users at racket-lang.org>
> Subject: Re: [racket] problem with pict->bitmap and
>         mrlib/write-animated-gif
> Message-ID:
>         <CAFrFfuG9GXu+N6Tp6Ccg1-n1QX6KFybMRpyj9p_Drk2+SRcrKw at mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
>
> Solved, with help from Jens Soegaard; I had to modify mrlib/write-gifs to
> take a disposal argument, and use it in all calls to gif-add-control.
>
> martin
>
> On Sat, Sep 27, 2014 at 12:27 AM, Martin DeMello <martindemello at gmail.com>
> wrote:
>
>> Can't figure it out, but something in the interaction between pict->bitmap
>> and write-animated-gif is causing the frames to display one on top of the
>> other when viewing the gif.
>>
>> #lang racket
>>
>> (require pict
>>          mrlib/gif)
>>
>> (define (draw-frame i)
>>   (pict->bitmap (circle (* 50 i))))
>>
>> (write-animated-gif
>>  (map draw-frame (sequence->list (in-range 1 10)))
>>  10
>>  "test1.gif"
>>  #:loop? true
>>  #:one-at-a-time? true)
>>
>> martin
>>
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140927/a42ef6eb/attachment-0001.html>
>
> ------------------------------
>
> Message: 5
> Date: Sat, 27 Sep 2014 18:54:56 -0400
> From: Prabhakar Ragde <plragde at uwaterloo.ca>
> To: Bill Richter <richter at math.northwestern.edu>
> Cc: users at racket-lang.org
> Subject: Re: [racket] proof assistants, DrRacket and Bootstrap
> Message-ID: <54274040.5070902 at uwaterloo.ca>
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>
> On 2014-09-27, 5:50 PM, Bill Richter wrote:
>> Thanks, Prabhakar, this is exactly what I wanted, and you know something about it (I was barely able to install Coq):
>
> I struggled with installing it also, especially since I'm using a Mac.
> Racket has spoiled us in that respect.
>
>>     But parsing an expression written in faux OCaml/Haskell is quite
>>     complicated. [...] Parsing is much easier in Racket/Scheme because
>>     of the similarity of code/data and the presence of 'read'.
>>
>> I think it's a parsing question I'm stuck on.  I think I need to use camlp to avoid the following hack/kludge which I learned from the HOL Light experts, found in my file
>> hol_light/RichterHilbertAxiomGeometry/readable.ml
>> which is part of the HOL Light distribution:
>>
>> (* From update_database.ml: Execute any OCaml expression given as a string.  *)
>
> I don't have enough practice in OCaml to grasp what your code does (and,
> yes, it is way off-topic), but I do know that about once every month or
> two, someone posts to this list about using 'eval' in Racket, and they
> are warned not to use it unless they really know what they are doing and
> have a suitable application (which is rare). That warning probably is
> doubled in a statically-typed language, in which 'eval' should not even
> be possible without subverting the type system (I gather this is what
> you are doing). Translating the gist of suggestions to such posters, I
> would say: do you really need to execute *any* expression? If not, use
> OCamllex and/or OCamlyacc (maybe Menhir, or a smaller and cleaner parser
> combinator library) to write a parser for your language subset, and then
> write an interpreter for the resulting ASTs. --PR
>
>
> ------------------------------
>
> Message: 6
> Date: Sat, 27 Sep 2014 21:11:08 -0400
> From: Anthony Carrico <acarrico at memebeam.org>
> To: "Alexander D. Knauth" <alexander at knauth.org>
> Cc: racket users list <users at racket-lang.org>
> Subject: Re: [racket] typed racket and generic interfaces, is there a
>         workaround using properties?
> Message-ID: <5427602C.5010101 at memebeam.org>
> Content-Type: text/plain; charset="windows-1252"
>
> On 09/27/2014 09:49 AM, Alexander D. Knauth wrote:
>> Like this, I think:  (but for some reason it seems to print it 3 times?)
>> #lang typed/racket
>>
>> (struct foo () #:transparent
>>   #:property prop:custom-write
>>   (lambda (this out mode)
>>     (displayln "whatever")))
>>
>> (print (foo))
>
> I didn't realize #:property was legal in typed struct. I guess there
> should be a documentation update here:
>   http://docs.racket-lang.org/ts-reference/special-forms.html
>
> BTW change your program to use the given port: (displayln "whatever"
> out) to avoid the double printing problem, although I'm not exactly sure
> why it would have that effect.
>
> Thanks!
>
> --
> Anthony Carrico
>
>
> -------------- next part --------------
> A non-text attachment was scrubbed...
> Name: signature.asc
> Type: application/pgp-signature
> Size: 181 bytes
> Desc: OpenPGP digital signature
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140927/63d8e553/attachment.sig>
>
> End of users Digest, Vol 109, Issue 65
> **************************************

From moshedeutsch115 at gmail.com  Mon Sep 29 16:42:32 2014
From: moshedeutsch115 at gmail.com (Moshe Deutsch)
Date: Mon, 29 Sep 2014 16:42:32 -0400
Subject: [racket] users Digest, Vol 109, Issue 64
In-Reply-To: <mailman.462.1411852138.679.users@racket-lang.org>
References: <mailman.462.1411852138.679.users@racket-lang.org>
Message-ID: <CAKpSqLCAUce96yYDu4pkUa344BwvmjnRNqvJqPiLS7jDvoYf2g@mail.gmail.com>

Please take me off the list

Thanks

On Sat, Sep 27, 2014 at 5:08 PM,  <users-request at racket-lang.org> wrote:
> Send users mailing list submissions to
>         users at racket-lang.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>         http://lists.racket-lang.org/users/listinfo
> or, via email, send a message with subject or body 'help' to
>         users-request at racket-lang.org
>
> You can reach the person managing the list at
>         users-owner at racket-lang.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of users digest..."
>
>
> [Racket Users list:
>  http://lists.racket-lang.org/users ]
>
>
> Today's Topics:
>
>    1. Re: typed racket and generic interfaces,  is there a
>       workaround using properties? (Alexander D. Knauth)
>    2. Re: proof assistants, DrRacket and Bootstrap (J. Ian Johnson)
>    3. aws/glacier: credential scope (Norman Gray)
>    4. Re: proof assistants, DrRacket and Bootstrap (Bill Richter)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Sat, 27 Sep 2014 13:55:12 -0400
> From: "Alexander D. Knauth" <alexander at knauth.org>
> To: Spencer florence <spencerflorence at gmail.com>
> Cc: racket users list <users at racket-lang.org>
> Subject: Re: [racket] typed racket and generic interfaces,      is there a
>         workaround using properties?
> Message-ID: <38E7C628-8FA0-4F07-ABAD-E271E2A66426 at knauth.org>
> Content-Type: text/plain; charset="windows-1252"
>
> Actually using prop:dict works (I hadn?t found prop:dict yet when I first asked this)
> #lang typed/racket
>
> (define-type Dict-Ref ([Dict Any] [Any] . ->* . Any))
> (define-type Dict-Set! (Dict Any Any . -> . Void))
> (define-type Dict-Set (Dict Any Any . -> . Dict))
> (define-type Dict-Remove! (Dict Any . -> . Void))
> (define-type Dict-Remove (Dict Any . -> . Dict))
> (define-type Dict-Iterate-First (Dict . -> . Any))
> (define-type Dict-Iterate-Next (Dict Any . -> . Any))
> (define-type Dict-Iterate-Key (Dict Any . -> . Any))
> (define-type Dict-Iterate-Value (Dict Any . -> . Any))
> (define-type Dict-Count (Dict . -> . Natural))
>
> (require/typed racket/dict
>                [#:opaque Dict dict?]
>                [dict-ref Dict-Ref]
>                [dict-set! Dict-Set!]
>                [dict-set Dict-Set]
>                [dict-remove! Dict-Remove!]
>                [dict-remove Dict-Remove]
>                [dict-iterate-first Dict-Iterate-First]
>                [dict-iterate-next Dict-Iterate-Next]
>                [dict-iterate-key Dict-Iterate-Key]
>                [dict-iterate-value Dict-Iterate-Value]
>                [dict-count Dict-Count]
>                [prop:dict Struct-Type-Property])
>
> (define (make-dict-prop #:dict-ref [dict-ref : Dict-Ref]
>                         #:dict-set! [dict-set! : (U Dict-Set! #f) #f]
>                         #:dict-set [dict-set : (U Dict-Set #f) #f]
>                         #:dict-remove! [dict-remove! : (U Dict-Remove! #f) #f]
>                         #:dict-remove [dict-remove : (U Dict-Remove #f) #f]
>                         #:dict-count [dict-count : Dict-Count]
>                         #:dict-iterate-first [dict-iterate-first : Dict-Iterate-First]
>                         #:dict-iterate-next [dict-iterate-next : Dict-Iterate-Next]
>                         #:dict-iterate-key [dict-iterate-key : Dict-Iterate-Key]
>                         #:dict-iterate-value [dict-iterate-value : Dict-Iterate-Value])
>   (vector-immutable dict-ref
>                     dict-set!
>                     dict-set
>                     dict-remove!
>                     dict-remove
>                     dict-count
>                     dict-iterate-first
>                     dict-iterate-next
>                     dict-iterate-key
>                     dict-iterate-value))
>
> (struct foo () #:transparent
>   #:property prop:dict
>   (make-dict-prop #:dict-ref (lambda (this key [failure #f]) "whatever")
>                   #:dict-count (lambda (this) 0)
>                   #:dict-iterate-first (lambda (this) #f)
>                   #:dict-iterate-next (lambda (this pos) 0)
>                   #:dict-iterate-key (lambda (this pos) pos)
>                   #:dict-iterate-value (lambda (this pos) "whatever")))
>
> (dict-ref (assert (foo) dict?) "idontkare")
>
>
> On Sep 27, 2014, at 1:07 AM, Spencer florence <spencerflorence at gmail.com> wrote:
>
>> I don?t think you can. You would need define the struct in an untyped module then require it via require/typed. This is why dict?s don?t work in typed/racket either.
>>
>>
>>
>> On Sat, May 24, 2014 at 9:39 PM, Alexander D. Knauth <alexander at knauth.org> wrote:
>>
>> Do generic interfaces work using structure type properties, and if they do, is there a way to use generic interfaces through properties so that I can do it in typed racket?
>>
>> Specifically I?m trying to use the gen:custom-write and gen:dict generic interfaces. I can use prop:custom-write for the first one, but I don?t know what to do for gen:dict.
>>
>> Otherwise I?ll just put it in an untyped submodule.
>>
>>
>> ____________________
>> Racket Users list:
>> http://lists.racket-lang.org/users
>>
>>
>
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140927/17fc9f08/attachment-0001.html>
>
> ------------------------------
>
> Message: 2
> Date: Sat, 27 Sep 2014 16:09:09 -0400 (EDT)
> From: "J. Ian Johnson" <ianj at ccs.neu.edu>
> To: Bill Richter <richter at math.northwestern.edu>
> Cc: users at racket-lang.org
> Subject: Re: [racket] proof assistants, DrRacket and Bootstrap
> Message-ID: <28884853.169131411848549503.JavaMail.root at zimbra>
> Content-Type: text/plain; charset=utf-8
>
> To answer the first part of your email, yes. Carl Eastlund developed Dracula for Rackety use of the ACL2 theorem prover.
> -Ian
> ----- Original Message -----
> From: Bill Richter <richter at math.northwestern.edu>
> To: users at racket-lang.org
> Sent: Sat, 27 Sep 2014 03:38:48 -0400 (EDT)
> Subject: [racket] proof assistants, DrRacket and Bootstrap
>
> I have a few questions that might be off-topic.  Are you interested in formal proofs?  Have you considered adapting DrRacket to give an integrated editor for a proof assistant?  The proof assistants Coq and Isabelle use jedit and ProofGeneral, which I think aren't nearly as nice as DrRacket.  I actually use HOL Light, which nobody uses an integrated editor for.  Here are the slides for a talk I gave at the Institut Henri Poincar? in the workshop ``Formalization of mathematics in proof assistants''
> http://www.math.northwestern.edu/~richter/RichterIHPslide.pdf
>
> HOL Light and Coq are written in OCaml, a dialect of ML, which is therefore similar to Scheme, but it has one difference that I wonder if anyone here's knows how to deal with.  Scheme is well-suited for writing a Scheme interpreter, because of the quote feature.  OCaml doesn't have a quote feature, so the question arises how to write an OCaml interpreter inside OCaml.  That's not quite what I want to do, but if anyone could explain how to do it, I'd be grateful.
>
> I have a 7th grade student I'm trying to teach Racket, and I'm a bit confused about Bootstrap and Program By Design.  Is there any reason for my student not to just read HtDP?
>
> --
> Best,
> Bill
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>
>
>
>
> ------------------------------
>
> Message: 3
> Date: Sat, 27 Sep 2014 21:14:55 +0100
> From: Norman Gray <norman at astro.gla.ac.uk>
> To: racket users list <users at racket-lang.org>
> Subject: [racket] aws/glacier: credential scope
> Message-ID: <0CA916BA-DD1E-47E9-BFA1-7092EC3F17F5 at astro.gla.ac.uk>
> Content-Type: text/plain; charset=us-ascii
>
>
> Greetings.
>
> I'm trying to use the aws/glacier package, and running into a problem where I'm being told:
>
>     Credential should be scoped to a valid region, not 'eu-west-1'
>
> I'm following the instructions at <https://github.com/greghendershott/aws/blob/master/aws/manual.md>
>
> My test code is:
>
>     % cat glacier.rkt
>     #lang racket/base
>
>     (require aws/glacier
>              aws/keys)
>
>     (define vault "testvault")
>     (region "eu-west-1")
>     (read-keys "aws-zbu-credentials") ; local file
>
>     (module+ main
>       (printf "region=~a~%" (region))
>       (printf "Vaults: ~s~%" (list-vaults))
>       (printf "...specifically: ~s~%" (describe-vault vault)))
>
> Running this produces:
>
>     % racket glacier.rkt
>     region=eu-west-1
>     aws: HTTP/1.1 403 Forbidden
>     x-amzn-RequestId: Un3-L2zlaJBPyrIVKJrWuQcqtMMYQAr34gYUOSScg6Qepc4
>     Content-Type: application/json
>     Content-Length: 129
>     Date: Sat, 27 Sep 2014 18:35:50 GMT
>
>      {"message":"Credential should be scoped to a valid region, not 'eu-west-1'. ","code":"InvalidSignatureException","type":"Client"}
>     HTTP 403 "Forbidden". AWS Code="InvalidSignatureException" Message="Credential should be scoped to a valid region, not 'eu-west-1'. "
>       context...:
>        check-response
>        /Users/norman/Library/Racket/6.1/pkgs/aws/aws/glacier.rkt:97:22: temp68
>        request/redirect/uri
>        (submod /checkouts/me/code/zbu/glacier.rkt main): [running body]
>
> Things I thought of:
>
>   * Printing (public-key)/(private-key) indicates that the credentials are being read correctly.
>   * When I change the argument of (region) to "us-west-1", that's the region that appears in the error message.
>   * My "testvault" vault is in eu-west-1 (and this is indeed one of the valid regions for glacier, reported in <http://docs.aws.amazon.com/general/latest/gr/rande.html> and which does have a host at http://glacier.eu-west-1.amazonaws.com
>   * As far as I can see, credentials are _not_ scoped, but are all at us-east-1.
>   * <http://docs.aws.amazon.com/general/latest/gr/signature-v4-troubleshooting.html> says that "IAM [...] accepts only us-east-1 as its region specification", so I'm taking it that (region) is for setting the _vault_'s region.
>   * I'm not a great AWS expert, so I could have something in my setup broken; but if so, I've no clue what.
>
> If, however, I change the (region) argument to "us-east-1", I get a different error message "User: arn:aws:iam::786725553169:user/zbu is not authorized to perform: glacier:ListVaults on resource: arn:aws:glacier:us-east-1:786725553169:vaults/"  That makes sense, since there's no such vault, but it's interesting that it gets _further_ when the (region) matches the region for the IAM service.
>
> I don't see any other (region) equivalents for the other services supported by the package.  Is that because all of the other services supported by the package are supported by all the AWS regions, or am I missing a configuration?
>
> Thanks for any pointers.
>
> All the best,
>
> Norman
>
>
> --
> Norman Gray  :  http://nxg.me.uk
> SUPA School of Physics and Astronomy, University of Glasgow, UK
>
>
>
>
> ------------------------------
>
> Message: 4
> Date: Sat, 27 Sep 2014 16:08:49 -0500
> From: Bill Richter <richter at math.northwestern.edu>
> To: "J. Ian Johnson" <ianj at ccs.neu.edu>
> Cc: users at racket-lang.org
> Subject: Re: [racket] proof assistants, DrRacket and Bootstrap
> Message-ID:
>         <201409272108.s8RL8nqh030467 at poisson.math.northwestern.edu>
>
>    Carl Eastlund developed Dracula for Rackety use of the ACL2 theorem prover.
>
> Thanks, Ian!  May I suggest that you try to handle Coq, Isabelle or HOL Light?  I believe these are the foremost proof assistants.  The fields medalist Vladimir Voevodsky uses Coq, in which the 4-color theorem and the Feit-Thompson theorem was formalized by Georges Gonthier.  Tom Hales just finished formalizing his proof of Kepler conjecture in HOL Light and Isabelle (which is HOL as well).  I'm no expert, but I think that ACL2 (also Prover9) is an FOL prover, and the bulk of activity in formal proofs uses type theories.  HOL is simple, it's Church's simple types (a version of the Lambda Calculus), and Coq uses a much more complicated type theory.
>
> --
> Best,
> Bill
>
>
> End of users Digest, Vol 109, Issue 64
> **************************************


From moshedeutsch115 at gmail.com  Mon Sep 29 16:42:42 2014
From: moshedeutsch115 at gmail.com (Moshe Deutsch)
Date: Mon, 29 Sep 2014 16:42:42 -0400
Subject: [racket] users Digest, Vol 109, Issue 66
In-Reply-To: <mailman.472.1411918184.679.users@racket-lang.org>
References: <mailman.472.1411918184.679.users@racket-lang.org>
Message-ID: <CAKpSqLB3giUdG2o+oKUR3+CmXoCrEQYYrHQ5Th+NqEPvjHaKfA@mail.gmail.com>

Please take me off the list

Thanks

On Sun, Sep 28, 2014 at 11:29 AM,  <users-request at racket-lang.org> wrote:
> Send users mailing list submissions to
>         users at racket-lang.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>         http://lists.racket-lang.org/users/listinfo
> or, via email, send a message with subject or body 'help' to
>         users-request at racket-lang.org
>
> You can reach the person managing the list at
>         users-owner at racket-lang.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of users digest..."
>
>
> [Racket Users list:
>  http://lists.racket-lang.org/users ]
>
>
> Today's Topics:
>
>    1. Re: typed racket and generic interfaces, is there a
>       workaround using properties? (Sam Tobin-Hochstadt)
>    2. Re: proof assistants, DrRacket and Bootstrap (Bill Richter)
>    3. Re: proof assistants, DrRacket and Bootstrap (Matthias Felleisen)
>    4. Re: proof assistants, DrRacket and Bootstrap (Bill Richter)
>    5. (fourth RacketCon) videos (Asumu Takikawa)
>    6. Re: proof assistants, DrRacket and Bootstrap (Prabhakar Ragde)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Sat, 27 Sep 2014 21:58:14 -0400
> From: Sam Tobin-Hochstadt <samth at cs.indiana.edu>
> To: Anthony Carrico <acarrico at memebeam.org>
> Cc: racket users list <users at racket-lang.org>
> Subject: Re: [racket] typed racket and generic interfaces, is there a
>         workaround using properties?
> Message-ID:
>         <CAK=HD+aVN_KNXZtf9_9WSFftehQN5m_xb7qGFEgHdpb1EuVXCQ at mail.gmail.com>
> Content-Type: text/plain; charset=UTF-8
>
> On Sat, Sep 27, 2014 at 9:11 PM, Anthony Carrico <acarrico at memebeam.org> wrote:
>> On 09/27/2014 09:49 AM, Alexander D. Knauth wrote:
>>> Like this, I think:  (but for some reason it seems to print it 3 times?)
>>> #lang typed/racket
>>>
>>> (struct foo () #:transparent
>>>   #:property prop:custom-write
>>>   (lambda (this out mode)
>>>     (displayln "whatever")))
>>>
>>> (print (foo))
>>
>> I didn't realize #:property was legal in typed struct. I guess there
>> should be a documentation update here:
>>   http://docs.racket-lang.org/ts-reference/special-forms.html
>
> No, instead we need to change typed racket to reject those programs.
> However, that would break a bunch of code with no easy fix that relies
> on getting away with this, so we'll probably have to wait till we have
> a real story here.
>
>> BTW change your program to use the given port: (displayln "whatever"
>> out) to avoid the double printing problem, although I'm not exactly sure
>> why it would have that effect.
>
> It prints multiple times (to hidden output ports) to detect cycles
> among other things.
>
> Sam
>
>> Thanks!
>>
>> --
>> Anthony Carrico
>>
>>
>>
>> ____________________
>>   Racket Users list:
>>   http://lists.racket-lang.org/users
>>
>
>
> ------------------------------
>
> Message: 2
> Date: Sat, 27 Sep 2014 23:57:12 -0500
> From: Bill Richter <richter at math.northwestern.edu>
> To: Prabhakar Ragde <plragde at uwaterloo.ca>
> Cc: users at racket-lang.org
> Subject: Re: [racket] proof assistants, DrRacket and Bootstrap
> Message-ID:
>         <201409280457.s8S4vCrp000773 at poisson.math.northwestern.edu>
>
>    Translating the gist of suggestions to such posters, I would say:
>    do you really need to execute *any* expression? If not, use
>    OCamllex and/or OCamlyacc (maybe Menhir, or a smaller and cleaner
>    parser combinator library) to write a parser for your language
>    subset, and then write an interpreter for the resulting ASTs.
>
> Thanks, Prabhakar!  I think something like what you're saying is true.  HOL Light is based on quite a lot of the OCaml parser camlp, and I've thought for a long time that I had to learn it myself.
>
>    I don't have enough practice in OCaml to grasp what your code does
>
> I'm making a dialect ofHOL Light with different syntax by interpreting one of my programs as a string and then breaking the string up into the component pieces, but then I need this Toploop/exec hack to evaluate my variables.  That sounds like the Scheme eval, and in Scheme you can solve that problem nicely with the quote feature.
>
> --
> Best,
> Bill
>
>
> ------------------------------
>
> Message: 3
> Date: Sat, 27 Sep 2014 16:00:35 -0600
> From: Matthias Felleisen <matthias at ccs.neu.edu>
> To: Bill Richter <richter at math.northwestern.edu>
> Cc: users at racket-lang.org
> Subject: Re: [racket] proof assistants, DrRacket and Bootstrap
> Message-ID: <B7FBF0FD-EA70-44BE-BB63-577E2E1B6CFF at ccs.neu.edu>
> Content-Type: text/plain; charset=us-ascii
>
>
> On Sep 27, 2014, at 1:38 AM, Bill Richter wrote:
>
>> Scheme is well-suited for writing a Scheme interpreter, because of the quote feature.  OCaml doesn't have a quote feature, so the question arises how to write an OCaml interpreter inside OCaml.  That's not quite what I want to do, but if anyone could explain how to do it, I'd be grateful.
>
> Decouple parsing from interpreting and your ML interpreter for ML won't look too different from your Lisp/Racket/Scheme interpreter for Lisp/Racket/Scheme. Indeed, almost everyone who does the latter actually does some form of decoupling usually.
>
> In general,
>
>  (1) pick a representation for your chosen programming language that is tree-oriented, call it AST
>  (2) pick a representation for the values that your programs may return, call it VAL
>  (3) design a function
>
>         interpreter : AST -> VAL
>
>  (4) for a typed language, also design a function
>
>         type_check : AST -> Boolean
>
> In ML, you might try
>
>   type ast = VAR of String | LAM of String * ast | APP of ast * ast | CONST of int | ADD of ast * ast
>   type val = BASIC of int | FUNCTION of val -> val | ...
>
>
>> I have a 7th grade student I'm trying to teach Racket, and I'm a bit confused about Bootstrap and Program By Design.  Is there any reason for my student not to just read HtDP?
>
>
> 1. HtDP explains the above in a couple of sections.
>
> 2. HtDP is for college freshmen (roughly) and for high school students who have the patience to read. The occasional 7th and 8th grader work their way thru HtDP with adults who understand HtDP or learn to understand HtDP that way. The basic 7th grader will benefit from the material at bootstrap-world.org, which is what we use with many middle school students around the country.
>
> -- Matthias
>
>
>
>
>
>
>
>
> ------------------------------
>
> Message: 4
> Date: Sun, 28 Sep 2014 00:39:20 -0500
> From: Bill Richter <richter at math.northwestern.edu>
> To: Matthias Felleisen <matthias at ccs.neu.edu>
> Cc: users at racket-lang.org
> Subject: Re: [racket] proof assistants, DrRacket and Bootstrap
> Message-ID:
>         <201409280539.s8S5dKcx001372 at poisson.math.northwestern.edu>
> Content-Type: text/plain; charset=utf-8
>
>      type ast = VAR of String | LAM of String * ast | APP of ast * ast | CONST of int | ADD of ast * ast
>      type val = BASIC of int | FUNCTION of val -> val | ...
>
> Thanks, Matthias!  I know this is way off topic, but I think it's the sort of thing only Schemers understand, and I think HOL Light would be a lot better off my problem got solved.  There's a lot I didn't understand with your ML explanation.  Is this written down somewhere?  How to write a ML interpreter in ML using your type ideas?  Are you saying I don't need camlp parsing?  I would find that hard to believe.  Formal proofs in HOL Light are really just OCaml programs where a lot of code has already been evaluated.  Here's a simple Hilbert geometry program in my dialect:
>
> let ExistsPointOffLine = theorem `;
>   ?l. Line l  ?  ?Q. Q ? l
>
>   proof
>     intro_TAC ?l, H1;
>     consider A B C such that
>     ?(A = B) ? ?(A = C) ? ?(B = C) ? ?Collinear A B C     [Distinct] by fol I3;
>     assume (A ? l) ? (B ? l) ? (C ? l)     [all_on] by fol ?;
>     Collinear A B C     [] by fol H1 - Collinear_DEF;
>     fol - Distinct;
>   qed;
> `;;
>
> I think I need the `;[...]` construct, which at present just turns my proof into a string without any backslash or newline problems, which is then passed to my function theorem.  The usual HOL Light style is like this, using instead a function prove:
>
> let MULT_0 = prove
>  (`!m. m * 0 = 0`,
>   INDUCT_TAC THEN ASM_REWRITE_TAC[MULT; ADD_CLAUSES]);;
>
> Notice the related `[...]` construct, which I think sends the ordered pair off to heavy camlp parsing.
>
> Thanks for the 7th and 8th advice.  I think my student is smart and hard-working, and I (finally!) learned how to program reading your book HtDP, so it sounds like you're saying that I should be able to teach it to my student.  I went to bootstrap-world.org and, uh, maybe I didn't look hard enough, but I didn't find anything that looked juicy.
>
> --
> Best,
> Bill
>
>
> ------------------------------
>
> Message: 5
> Date: Sun, 28 Sep 2014 11:17:15 -0400
> From: Asumu Takikawa <asumu at ccs.neu.edu>
> To: Racket Users <users at racket-lang.org>
> Subject: [racket] (fourth RacketCon) videos
> Message-ID: <20140928151714.GH30398 at localhost>
> Content-Type: text/plain; charset=us-ascii
>
> Hi all,
>
> The hi-res talk recordings for (fourth RacketCon) are now up on Youtube.
> You can watch all of them via this playlist:
>
>   https://www.youtube.com/playlist?list=PLXr4KViVC0qI9t3lizitiFJ1cFIeN2Gdh
>
> Let me know if you find any glitches.
>
> Cheers,
> Asumu
>
>
> ------------------------------
>
> Message: 6
> Date: Sun, 28 Sep 2014 11:29:32 -0400
> From: Prabhakar Ragde <plragde at uwaterloo.ca>
> To: Bill Richter <richter at math.northwestern.edu>
> Cc: users at racket-lang.org
> Subject: Re: [racket] proof assistants, DrRacket and Bootstrap
> Message-ID: <5428295C.5050006 at uwaterloo.ca>
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>
> On 2014-09-28, 12:57 AM, Bill Richter wrote:
>
>> I'm making a dialect of HOL Light with different syntax by
>> interpreting one of my programs as a string and then breaking the
>> string up into the component pieces, but then I need this
>> Toploop/exec hack to evaluate my variables.  That sounds like the
>> Scheme eval, and in Scheme you can solve that problem nicely with the
>> quote feature.
>
> A Racketeer would probably tell you to use macros to define your
> dialect, instead of using an eval hack. The corresponding tool for OCaml
> is camlp4, and it sounds as if it would be worth your time to learn it
> thoroughly. --PR
>
>
> End of users Digest, Vol 109, Issue 66
> **************************************


From moshedeutsch115 at gmail.com  Mon Sep 29 16:42:52 2014
From: moshedeutsch115 at gmail.com (Moshe Deutsch)
Date: Mon, 29 Sep 2014 16:42:52 -0400
Subject: [racket] users Digest, Vol 109, Issue 61
In-Reply-To: <mailman.456.1411787294.679.users@racket-lang.org>
References: <mailman.456.1411787294.679.users@racket-lang.org>
Message-ID: <CAKpSqLCxRNnjTOapCuqr3FukPYo2jfJVWCT4ah-rA8O-fk9Wfw@mail.gmail.com>

Please take me off the list

Thanks

On Fri, Sep 26, 2014 at 11:08 PM,  <users-request at racket-lang.org> wrote:
> Send users mailing list submissions to
>         users at racket-lang.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>         http://lists.racket-lang.org/users/listinfo
> or, via email, send a message with subject or body 'help' to
>         users-request at racket-lang.org
>
> You can reach the person managing the list at
>         users-owner at racket-lang.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of users digest..."
>
>
> [Racket Users list:
>  http://lists.racket-lang.org/users ]
>
>
> Today's Topics:
>
>    1. Re: typed racket, filters, and polymorphism (Alexander D. Knauth)
>    2. Re: ANN: DOS,     Delimited-continuation-based Operating-system
>       Simulator (Jay McCarthy)
>    3. Re: web server: module servlets (Jay McCarthy)
>    4. Re: typed racket and generic interfaces, is there a
>       workaround using properties? (Anthony Carrico)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Fri, 26 Sep 2014 19:35:24 -0400
> From: "Alexander D. Knauth" <alexander at knauth.org>
> To: racket users list <users at racket-lang.org>
> Cc: Sam Tobin-Hochstadt <samth at cs.indiana.edu>
> Subject: Re: [racket] typed racket, filters, and polymorphism
> Message-ID: <8CF041CE-65AD-488D-B2AE-FAE6C8C1C352 at knauth.org>
> Content-Type: text/plain; charset="windows-1252"
>
> Is it possible to have a struct that does certain things according to the guard?
> #lang typed/racket
>
> (struct (a) foo ([a : a]) #:transparent
>   #:guard (lambda (a _)
>             (unless (exact-integer? a)
>               (error 'foo "expected Integer, given ~v" a))
>             a))
>
> (ann (foo (ann 1 Any)) (foo Integer))
>
> (: x : (foo Any))
> (define x (foo 1))
>
> (ann (foo-a x) Integer)
>
> ;. Type Checker: Polymorphic function `foo1' could not be applied to arguments:
> ;Argument 1:
> ;  Expected: a
> ;  Given:    Any
> ;
> ;Result type:     (foo a)
> ;Expected result: (foo Integer)
> ; in: (foo (ann 1 Any))
> ;. Type Checker: Polymorphic function `foo-a' could not be applied to arguments:
> ;Argument 1:
> ;  Expected: (foo a)
> ;  Given:    (foo Any)
> ;
> ;Result type:     (a : ....)
> ;Expected result: Integer
> ; in: (foo-a x)
>
> On Sep 25, 2014, at 9:42 PM, Alexander D. Knauth <alexander at knauth.org> wrote:
>
>> What I?m trying to accomplish is something more like this:
>> #lang typed/racket
>>
>> (require "dimensions.rkt")
>>
>> (struct (d) unit ([name : Any] [scalar : Positive-Real] [dimension : d]) #:transparent
>>   #:guard (lambda (name scalar dimension _)
>>             (unless (dimension? dimension)
>>               (error 'unit "expected Dimension, given ~v" dimension))
>>             (values name scalar dimension)))
>>
>> (define-type (Unitof d) (unit d))
>>
>> (define-type Unit (Unitof Dimension))
>>
>> (define Unit? (make-predicate Unit))
>>
>> (define-type Unitish
>>   (U (Unitof Any)
>>      Dimension
>>      Positive-Real))
>>
>> (: ->unit : (All (d) (case-> [(Unitof d) -> (Unitof d)]
>>                              [Unitish -> Unit])))
>> (define (->unit u)
>>   (cond [(unit? u)
>>          (unless (Unit? u) ; this should never happen anyway because of the guard
>>            (error '->unit "expected (Unitof Dimension), given ~v" u))
>>          u]
>>         [(dimension? u) (unit u 1 u)]
>>         [(positive-real? u) (unit u u dimensionless-dimension)]))
>>
>>
>> On Sep 25, 2014, at 6:19 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> wrote:
>>
>>> No, I don't think you can do this. Can you say more about what you're
>>> trying to accomplish?
>>>
>>> Sam
>>>
>>> On Thu, Sep 25, 2014 at 6:15 PM, Alexander D. Knauth
>>> <alexander at knauth.org> wrote:
>>>> Do any of you have any advice for getting a function like this to
>>>> type-check?
>>>> #lang typed/racket
>>>>
>>>> (: check-int : (All (a) (case-> [a -> a]
>>>>                                [Any -> Integer])))
>>>> (define (check-int int)
>>>>  (unless (exact-integer? int)
>>>>    (error 'check-int "expected Integer, given ~v" int))
>>>>  int)
>>>>
>>>> ;. Type Checker: type mismatch
>>>> ;  expected: a
>>>> ;  given: Integer in: int
>>>>
>>>>
>>>>
>>>> ____________________
>>>>  Racket Users list:
>>>>  http://lists.racket-lang.org/users
>>>>
>>
>> ____________________
>>  Racket Users list:
>>  http://lists.racket-lang.org/users
>
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140926/bbbbc5ac/attachment-0001.html>
>
> ------------------------------
>
> Message: 2
> Date: Fri, 26 Sep 2014 20:03:57 -0400
> From: Jay McCarthy <jay.mccarthy at gmail.com>
> To: "Michael Bradley, Jr." <michaelsbradleyjr at gmail.com>
> Cc: users <users at racket-lang.org>
> Subject: Re: [racket] ANN: DOS, Delimited-continuation-based
>         Operating-system Simulator
> Message-ID:
>         <CAJYbDanLdZACFv41w+CAtNjvLernOomC0L7goMVHH-kwe07fXw at mail.gmail.com>
> Content-Type: text/plain; charset=UTF-8
>
> Sure, I can make a post about it. However, the core idea is already in
> this post:
>
> http://jeapostrophe.github.io/2012-07-12-cont-sys-post.html
>
> I can make another one, however, to rehash it and talk about the
> benefits of monoid states.
>
> Jay
>
> On Fri, Sep 26, 2014 at 2:15 PM, Michael Bradley, Jr.
> <michaelsbradleyjr at gmail.com> wrote:
>> Jay McCarthy <jay.mccarthy at ...> writes:
>>
>>>
>>> I've just released the game architecture library I talked about in
>>> part 2 of my RacketCon talk.
>>>
>>> The big picture of this library is to make World-style programs more
>>>  compositional by (a) using continuations to hide the internal state
>>> (including control state) of components and (b) using environments as
>>> a standard monoid-based inter-component communication channel. A
>>> monoid is used to ensure that the components can be evaluated in any
>>> order. Despite assumptions some have about continuations and pure
>>> functional programming, it is incredibly efficient and can run at
>>> 60FPS, as demonstrated in get-bonus.
>>>
>>> You can get it with
>>>
>>> raco pkg install dos
>>>
>>> And you can try out the demo with
>>>
>>> racket -l dos/examples/win
>>>
>>> The demo source is a bare 39 lines:
>>>
>>> https://github.com/jeapostrophe/dos/blob/master/dos/examples/win.rkt
>>>
>>> and I provide the non-DOS version for comparison:
>>>
>>> https://github.com/jeapostrophe/dos/blob/master/dos/examples/win-long.rkt
>>>
>>> The core library has a mere 36 lines:
>>>
>>> https://github.com/jeapostrophe/dos/blob/master/dos/main.rkt
>>>
>>
>>
>>
>> Hi Jay,
>>
>> I really enjoyed your talk on get-bonus at RacketCon last weekend, and
>> appreciate your sharing DOS with all of us. Also, your presentation at
>> Strange Loop 2013 was my first introduction to delimited continuations in
>> Racket!
>>
>> I have benefited from a number of your blog posts in the past, e.g. the one
>> on an improved threading macro, and was wondering if you would consider
>> doing a write-up on DOS?
>>
>> Best regards,
>>
>> --
>> Michael Bradley, Jr.
>> @michaelsbradley
>>
>>
>> ____________________
>>   Racket Users list:
>>   http://lists.racket-lang.org/users
>
>
>
> --
> Jay McCarthy
> http://jeapostrophe.github.io
>
>            "Wherefore, be not weary in well-doing,
>       for ye are laying the foundation of a great work.
> And out of small things proceedeth that which is great."
>                           - D&C 64:33
>
>
> ------------------------------
>
> Message: 3
> Date: Fri, 26 Sep 2014 20:08:30 -0400
> From: Jay McCarthy <jay.mccarthy at gmail.com>
> To: George Neuner <gneuner2 at comcast.net>
> Cc: users <users at racket-lang.org>
> Subject: Re: [racket] web server: module servlets
> Message-ID:
>         <CAJYbDa=aYrXcKh6bZK-dBai4CYR8tE+yuXOMHp6P9a-sYs_f4Q at mail.gmail.com>
> Content-Type: text/plain; charset=UTF-8
>
> Hi George,
>
> I suggest that all Racket web-server apps not use the dynamic features
> of serve/servlet either, but instead write the servlet as the single
> "request -> response" function that serve/servlet provides. This would
> ensure that all the libraries are loaded the one time and has the best
> performance out of all the options of the Web server. Furthermore,
> there are many features (such as total URL control) that are only
> available in this model.
>
> In case it is not completely clear, here's a little example (<500
> lines) of a Web application that works this way:
>
> https://github.com/plt/racket/blob/master/pkgs/plt-services/meta/pkg-index/official/dynamic.rkt#L423
>
> Here's another that uses 'serve' directly for non-servlet needs:
>
> https://github.com/jeapostrophe/exp/blob/master/dir-serve.rkt
>
> And here's a considerably longer one:
>
> https://github.com/jeapostrophe/grade-samurai/blob/master/app.rkt#L2020
>
> I highly recommend using serve/servlet like this over any other
> options. In particular, many of the problems you're talking about were
> the reason that I made serve/servlet in the first place.
>
> Jay
>
> On Fri, Sep 26, 2014 at 12:16 AM, George Neuner <gneuner2 at comcast.net> wrote:
>> On 9/25/2014 6:26 PM, Jay McCarthy wrote:
>>
>> web-server/servlet is NOT shared, nor is any net library or common
>> racket library like racket/list. They are all totally unique per
>> servlet.
>>
>> This, by the way, is part of why I don't recommend using dynamic
>> servlets at all and suggest using serve/servlet.
>>
>>
>> But ... IIUC ... a listener started by  serve/servlet  will still
>> dynamically load from <servlet-root>, <server-root>/htdocs/*, etc. if the
>> request doesn't match a hard coded dispatch URL - the regex of the initial
>> servlet or an entry in a dispatcher table.  At least that's the behavior I
>> see with my own application: my initial servlet returns 404 when called -
>> the call to  serve/servlet  just sets up the environment and and all the
>> "real" servlets are demand loaded from disk when first touched.
>>
>> Are you advocating *static* linking and essentially just dispatching to
>> internal functions by URL?  ISTM that that defeats the purpose.
>>
>> George
>
>
>
> --
> Jay McCarthy
> http://jeapostrophe.github.io
>
>            "Wherefore, be not weary in well-doing,
>       for ye are laying the foundation of a great work.
> And out of small things proceedeth that which is great."
>                           - D&C 64:33
>
>
> ------------------------------
>
> Message: 4
> Date: Fri, 26 Sep 2014 23:07:55 -0400
> From: Anthony Carrico <acarrico at memebeam.org>
> To: users at racket-lang.org
> Subject: Re: [racket] typed racket and generic interfaces, is there a
>         workaround using properties?
> Message-ID: <54262A0B.2010901 at memebeam.org>
> Content-Type: text/plain; charset="windows-1252"
>
> Dredging up an old thread: So how exactly do you define custom-write for
> a typed racket struct?
>
> On 05/25/2014 03:30 PM, Alexander D. Knauth wrote:
>> Never mind I just found prop:dict.
>>
>> By the way it would probably be helpful if instead of the value for
>> prop:dict being a vector, it was a hash table with the method names
>> as the keys, so then it would be a lot easier to tell which procedure
>> goes with which method, and also a lot easier to extend it while
>> maintaining backwards compatibility.
>>
>> On May 24, 2014, at 10:38 PM, Alexander D. Knauth
>> <alexander at knauth.org> wrote:
>>
>>> Do generic interfaces work using structure type properties, and if
>>> they do, is there a way to use generic interfaces through
>>> properties so that I can do it in typed racket?
>>>
>>> Specifically I?m trying to use the gen:custom-write and gen:dict
>>> generic interfaces.  I can use prop:custom-write for the first one,
>>> but I don?t know what to do for gen:dict.
>>>
>>> Otherwise I?ll just put it in an untyped submodule.
>
> --
> Anthony Carrico
>
> -------------- next part --------------
> A non-text attachment was scrubbed...
> Name: signature.asc
> Type: application/pgp-signature
> Size: 181 bytes
> Desc: OpenPGP digital signature
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140926/d14018a4/attachment.sig>
>
> End of users Digest, Vol 109, Issue 61
> **************************************

From moshedeutsch115 at gmail.com  Mon Sep 29 16:43:01 2014
From: moshedeutsch115 at gmail.com (Moshe Deutsch)
Date: Mon, 29 Sep 2014 16:43:01 -0400
Subject: [racket] users Digest, Vol 109, Issue 63
In-Reply-To: <mailman.3.1411833602.9920.users@racket-lang.org>
References: <mailman.3.1411833602.9920.users@racket-lang.org>
Message-ID: <CAKpSqLBiQrf3aUng9oi=2Tx=o=-bujAfy=yFe34WO040p+4aTw@mail.gmail.com>

Please take me off the list

Thanks

On Sat, Sep 27, 2014 at 12:00 PM,  <users-request at racket-lang.org> wrote:
> Send users mailing list submissions to
>         users at racket-lang.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>         http://lists.racket-lang.org/users/listinfo
> or, via email, send a message with subject or body 'help' to
>         users-request at racket-lang.org
>
> You can reach the person managing the list at
>         users-owner at racket-lang.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of users digest..."
>
>
> [Racket Users list:
>  http://lists.racket-lang.org/users ]
>
>
> Today's Topics:
>
>    1.  proof assistants, DrRacket and Bootstrap (Prabhakar Ragde)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Sat, 27 Sep 2014 10:33:15 -0400
> From: Prabhakar Ragde <plragde at uwaterloo.ca>
> To: "users at racket-lang.org" <users at racket-lang.org>
> Subject: [racket]  proof assistants, DrRacket and Bootstrap
> Message-ID: <5426CAAB.2010400 at uwaterloo.ca>
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
>
>> I have a few questions that might be off-topic.  Are you interested
>> in formal proofs?  Have you considered adapting DrRacket to give an
>> integrated editor for a proof assistant?  The proof assistants Coq
>> and Isabelle use jedit and ProofGeneral, which I think aren't nearly
>> as nice as DrRacket.  I actually use HOL Light, which nobody uses an
>> integrated editor for.  Here are the slides for a talk I gave at the
>> Institut Henri Poincar? in the workshop ``Formalization of
>> mathematics in proof assistants''
>> http://www.math.northwestern.edu/~richter/RichterIHPslide.pdf
>
> I am teaching a grad course using Coq right now, and using Proof General
> within Emacs for it. That is not bad, but I would love to have a
> DrRacket interface. I know something like this is possible, because at
> Brown they used DrRacket to prepare and run OCaml programs, a few years
> ago. I really don't have the time to work on this, though. (And if I
> did, I would first work on teaching language subsets of Haskell, which I
> also would like to have inside DrRacket.)
>
>> HOL Light and Coq are written in OCaml, a dialect of ML, which is
>> therefore similar to Scheme, but it has one difference that I wonder
>> if anyone here's knows how to deal with.  Scheme is well-suited for
>> writing a Scheme interpreter, because of the quote feature.  OCaml
>> doesn't have a quote feature, so the question arises how to write an
>> OCaml interpreter inside OCaml.  That's not quite what I want to do,
>> but if anyone could explain how to do it, I'd be grateful.
>
> In my senior undergraduate programming language course, I have them
> write various interpreters and typecheckers in various languages,
> including Racket, OCaml, and Haskell. In the latter two, you need to use
> algebraic datatypes to build an abstract syntax tree, and interpret
> that. So the expression (* (+ 1 2) (- 3 4)) might become
>
> (Mult (Plus (Num 1) (Num 2)) (Minus (Num 3) (Num 4)))
>
> Writing an interpreter for such an AST is even easier than in
> Racket/Scheme, due to concise pattern-matching syntax. But parsing an
> expression written in faux OCaml/Haskell is quite complicated. I don't
> ask students to do that, since it really is part of a separate course
> (both languages come with general parsing tools that can be used).
> Parsing is much easier in Racket/Scheme because of the similarity of
> code/data and the presence of 'read'.
>
>> I have a 7th grade student I'm trying to teach Racket, and I'm a bit
>> confused about Bootstrap and Program By Design.  Is there any reason
>> for my student not to just read HtDP?
>
> You should view word from the authors/developers as definitive, but I
> would say that if your student has good algebraic skills, then using
> HtDP/2e is the right path. If not, Bootstrap may be a way to help
> develop those skills. --PR
>
>
> End of users Digest, Vol 109, Issue 63
> **************************************

From moshedeutsch115 at gmail.com  Mon Sep 29 16:43:12 2014
From: moshedeutsch115 at gmail.com (Moshe Deutsch)
Date: Mon, 29 Sep 2014 16:43:12 -0400
Subject: [racket] users Digest, Vol 109, Issue 60
In-Reply-To: <mailman.454.1411755610.679.users@racket-lang.org>
References: <mailman.454.1411755610.679.users@racket-lang.org>
Message-ID: <CAKpSqLBh9R=xKaeUNRKQxOcWzwT6H3-Whn0MAcDGV0frUgVbGw@mail.gmail.com>

Please take me off the list

Thanks

On Fri, Sep 26, 2014 at 2:20 PM,  <users-request at racket-lang.org> wrote:
> Send users mailing list submissions to
>         users at racket-lang.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>         http://lists.racket-lang.org/users/listinfo
> or, via email, send a message with subject or body 'help' to
>         users-request at racket-lang.org
>
> You can reach the person managing the list at
>         users-owner at racket-lang.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of users digest..."
>
>
> [Racket Users list:
>  http://lists.racket-lang.org/users ]
>
>
> Today's Topics:
>
>    1. Tricky case of occurrence typing in Typed Racket (Konrad Hinsen)
>    2. Re: Tricky case of occurrence typing in Typed Racket (Andrew Kent)
>    3. Re: Tricky case of occurrence typing in Typed Racket
>       (Konrad Hinsen)
>    4. Re: Tricky case of occurrence typing in Typed Racket (Andrew Kent)
>    5. Re: ANN: DOS,     Delimited-continuation-based Operating-system
>       Simulator (Michael Bradley, Jr.)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Fri, 26 Sep 2014 18:07:19 +0200
> From: Konrad Hinsen <konrad.hinsen at fastmail.net>
> To: users at racket-lang.org
> Subject: [racket] Tricky case of occurrence typing in Typed Racket
> Message-ID:
>         <21541.36663.168649.25871 at Konrad-Hinsens-MacBook-Pro-2.local>
> Content-Type: text/plain; charset=us-ascii
>
> Hi everyone,
>
> I am trying to convince Typed Racket that after a run-time check on
> some list, its elements are all of a given type. This is just like
> occurrence typing based on a predicate, except that my predicate
> is more complicated.
>
> My first attempt was this:
>
>
>    #lang typed/racket
>
>    (struct: foo ())
>    (struct: bar ())
>    (define-type FooOrBar (U foo bar))
>
>    (: f-bar (-> (Listof bar) Void))
>    (define (f-bar xs)
>      (void))
>
>    (: f-mixed (-> (Listof FooOrBar) Void))
>    (define (f-mixed xs)
>      (void))
>
>    (: f (-> (Listof FooOrBar) Void))
>    (define (f xs)
>      (if (for/and : Boolean ([x xs]) (bar? x))
>          (f-bar xs) ; contains only bars
>          (f-mixed xs) ; may contain foos as well
>          ))
>
> This yields a type error:
>
>    ; /Users/hinsen/projects/racket/foobar.rkt:18:13: Type Checker: type mismatch
>    ;   expected: (Listof bar)
>    ;   given: (Listof FooOrBar)
>    ;   in: xs
>
> After reading section 5 of the Typed Racket Guide, in particular section
> 5.2 entitled "Filters and Predicates", I thought I could simply define my
> own predicate with a type annotation:
>
>    #lang typed/racket
>
>    (struct: foo ())
>    (struct: bar ())
>    (define-type FooOrBar (U foo bar))
>
>    (: f-bar (-> (Listof bar) Void))
>    (define (f-bar xs)
>      (void))
>
>    (: f-mixed (-> (Listof FooOrBar) Void))
>    (define (f-mixed xs)
>      (void))
>
>    (: pure-bars? (-> (Listof FooOrBar) Boolean : (Listof bar)))
>    (define (pure-bars? xs)
>      (for/and : Boolean ([x xs]) (bar? x)))
>
>    (: f (-> (Listof FooOrBar) Void))
>    (define (f xs)
>      (if (pure-bars? xs)
>          (f-bar xs) ; contains only bars
>          (f-mixed xs) ; may contain foos as well
>          ))
>
> But this got me only into more trouble - now I don't even understand the
> error message any more:
>
>    ; /Users/hinsen/projects/racket/foobar.rkt:17:2: Type Checker: type mismatch;
>    ;  mismatch in filter
>    ;   expected: (((Listof bar) @ xs) | (! (Listof bar) @ xs))
>    ;   given: (Top | Top)
>    ;   in: (for/and : Boolean ((x xs)) (bar? x))
>
> Is there a way to do this kind of test?
>
> Konrad.
>
>
> ------------------------------
>
> Message: 2
> Date: Fri, 26 Sep 2014 12:12:29 -0400
> From: Andrew Kent <amk.kent at gmail.com>
> To: Konrad Hinsen <konrad.hinsen at fastmail.net>
> Cc: users at racket-lang.org
> Subject: Re: [racket] Tricky case of occurrence typing in Typed Racket
> Message-ID:
>         <CAMf1y4JeW-jdYeuJP-0x=U=TJB3tahxqvAmuM3TSbeFvGSAYJg at mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
>
> Will andmap work for you?
>
>    (struct: foo ())
>    (struct: bar ())
>    (define-type FooOrBar (U foo bar))
>
>    (: f-bar (-> (Listof bar) Void))
>    (define (f-bar xs)
>      (void))
>
>    (: f-mixed (-> (Listof FooOrBar) Void))
>    (define (f-mixed xs)
>      (void))
>
>    (: f (-> (Listof FooOrBar) Void))
>    (define (f xs)
>      (if (andmap bar? xs)
>          (f-bar xs) ; contains only bars
>          (f-mixed xs) ; may contain foos as well
>          ))
>
> On Fri, Sep 26, 2014 at 12:07 PM, Konrad Hinsen <konrad.hinsen at fastmail.net>
> wrote:
>
>> Hi everyone,
>>
>> I am trying to convince Typed Racket that after a run-time check on
>> some list, its elements are all of a given type. This is just like
>> occurrence typing based on a predicate, except that my predicate
>> is more complicated.
>>
>> My first attempt was this:
>>
>>
>>    #lang typed/racket
>>
>>    (struct: foo ())
>>    (struct: bar ())
>>    (define-type FooOrBar (U foo bar))
>>
>>    (: f-bar (-> (Listof bar) Void))
>>    (define (f-bar xs)
>>      (void))
>>
>>    (: f-mixed (-> (Listof FooOrBar) Void))
>>    (define (f-mixed xs)
>>      (void))
>>
>>    (: f (-> (Listof FooOrBar) Void))
>>    (define (f xs)
>>      (if (for/and : Boolean ([x xs]) (bar? x))
>>          (f-bar xs) ; contains only bars
>>          (f-mixed xs) ; may contain foos as well
>>          ))
>>
>> This yields a type error:
>>
>>    ; /Users/hinsen/projects/racket/foobar.rkt:18:13: Type Checker: type
>> mismatch
>>    ;   expected: (Listof bar)
>>    ;   given: (Listof FooOrBar)
>>    ;   in: xs
>>
>> After reading section 5 of the Typed Racket Guide, in particular section
>> 5.2 entitled "Filters and Predicates", I thought I could simply define my
>> own predicate with a type annotation:
>>
>>    #lang typed/racket
>>
>>    (struct: foo ())
>>    (struct: bar ())
>>    (define-type FooOrBar (U foo bar))
>>
>>    (: f-bar (-> (Listof bar) Void))
>>    (define (f-bar xs)
>>      (void))
>>
>>    (: f-mixed (-> (Listof FooOrBar) Void))
>>    (define (f-mixed xs)
>>      (void))
>>
>>    (: pure-bars? (-> (Listof FooOrBar) Boolean : (Listof bar)))
>>    (define (pure-bars? xs)
>>      (for/and : Boolean ([x xs]) (bar? x)))
>>
>>    (: f (-> (Listof FooOrBar) Void))
>>    (define (f xs)
>>      (if (pure-bars? xs)
>>          (f-bar xs) ; contains only bars
>>          (f-mixed xs) ; may contain foos as well
>>          ))
>>
>> But this got me only into more trouble - now I don't even understand the
>> error message any more:
>>
>>    ; /Users/hinsen/projects/racket/foobar.rkt:17:2: Type Checker: type
>> mismatch;
>>    ;  mismatch in filter
>>    ;   expected: (((Listof bar) @ xs) | (! (Listof bar) @ xs))
>>    ;   given: (Top | Top)
>>    ;   in: (for/and : Boolean ((x xs)) (bar? x))
>>
>> Is there a way to do this kind of test?
>>
>> Konrad.
>> ____________________
>>   Racket Users list:
>>   http://lists.racket-lang.org/users
>>
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140926/72332d84/attachment-0001.html>
>
> ------------------------------
>
> Message: 3
> Date: Fri, 26 Sep 2014 18:28:03 +0200
> From: Konrad Hinsen <konrad.hinsen at fastmail.net>
> To: users at racket-lang.org
> Subject: Re: [racket] Tricky case of occurrence typing in Typed Racket
> Message-ID:
>         <21541.37907.627925.114871 at Konrad-Hinsens-MacBook-Pro-2.local>
> Content-Type: text/plain; charset=us-ascii
>
> Andrew Kent writes:
>
>  > Will andmap work for you?
>
> Interesting... I didn't know about that one.
>
> For my demonstration code, that's indeed a good solution. In my real
> application, the test is more complicated. I need to check all
> elements of a list for conformance to a union type, so I have no
> prefabricated predicate, not even for the elements of my list.
>
> Konrad.
>
>
> ------------------------------
>
> Message: 4
> Date: Fri, 26 Sep 2014 12:48:05 -0400
> From: Andrew Kent <amk.kent at gmail.com>
> To: Konrad Hinsen <konrad.hinsen at fastmail.net>
> Cc: users at racket-lang.org
> Subject: Re: [racket] Tricky case of occurrence typing in Typed Racket
> Message-ID:
>         <CAMf1y4LenSG_Ti2S_sFeaDXY6Zo_P5nO9C8-RC6q_TNScG4G9w at mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
>
> Does using andmap with the custom predicates you alluded to for that union
> you mentioned work?
>
> Something like this perhaps?:
>
> #lang typed/racket
>
> (struct: T1 ())
> (struct: T2 ())
> (define-type T1or2 (U T1 T2))
>
> (: T1or2? (-> Any Boolean : T1or2))
> (define (T1or2? a)
>   (or (T1? a) (T2? a)))
>
>
> (: listof-T1or2 (-> Any Boolean : (Listof T1or2)))
> (define (listof-T1or2 l)
>   (and (list? l) (andmap T1or2? l)))
>
>
>
> On Fri, Sep 26, 2014 at 12:28 PM, Konrad Hinsen <konrad.hinsen at fastmail.net>
> wrote:
>
>> Andrew Kent writes:
>>
>>  > Will andmap work for you?
>>
>> Interesting... I didn't know about that one.
>>
>> For my demonstration code, that's indeed a good solution. In my real
>> application, the test is more complicated. I need to check all
>> elements of a list for conformance to a union type, so I have no
>> prefabricated predicate, not even for the elements of my list.
>>
>> Konrad.
>> ____________________
>>   Racket Users list:
>>   http://lists.racket-lang.org/users
>>
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140926/30da28f4/attachment-0001.html>
>
> ------------------------------
>
> Message: 5
> Date: Fri, 26 Sep 2014 18:15:31 +0000 (UTC)
> From: "Michael Bradley, Jr." <michaelsbradleyjr at gmail.com>
> To: users at racket-lang.org
> Subject: Re: [racket] ANN: DOS, Delimited-continuation-based
>         Operating-system Simulator
> Message-ID: <loom.20140926T201005-291 at post.gmane.org>
> Content-Type: text/plain; charset=us-ascii
>
> Jay McCarthy <jay.mccarthy at ...> writes:
>
>>
>> I've just released the game architecture library I talked about in
>> part 2 of my RacketCon talk.
>>
>> The big picture of this library is to make World-style programs more
>>  compositional by (a) using continuations to hide the internal state
>> (including control state) of components and (b) using environments as
>> a standard monoid-based inter-component communication channel. A
>> monoid is used to ensure that the components can be evaluated in any
>> order. Despite assumptions some have about continuations and pure
>> functional programming, it is incredibly efficient and can run at
>> 60FPS, as demonstrated in get-bonus.
>>
>> You can get it with
>>
>> raco pkg install dos
>>
>> And you can try out the demo with
>>
>> racket -l dos/examples/win
>>
>> The demo source is a bare 39 lines:
>>
>> https://github.com/jeapostrophe/dos/blob/master/dos/examples/win.rkt
>>
>> and I provide the non-DOS version for comparison:
>>
>> https://github.com/jeapostrophe/dos/blob/master/dos/examples/win-long.rkt
>>
>> The core library has a mere 36 lines:
>>
>> https://github.com/jeapostrophe/dos/blob/master/dos/main.rkt
>>
>
>
>
> Hi Jay,
>
> I really enjoyed your talk on get-bonus at RacketCon last weekend, and
> appreciate your sharing DOS with all of us. Also, your presentation at
> Strange Loop 2013 was my first introduction to delimited continuations in
> Racket!
>
> I have benefited from a number of your blog posts in the past, e.g. the one
> on an improved threading macro, and was wondering if you would consider
> doing a write-up on DOS?
>
> Best regards,
>
> --
> Michael Bradley, Jr.
> @michaelsbradley
>
>
>
>
> End of users Digest, Vol 109, Issue 60
> **************************************

From moshedeutsch115 at gmail.com  Mon Sep 29 16:43:22 2014
From: moshedeutsch115 at gmail.com (Moshe Deutsch)
Date: Mon, 29 Sep 2014 16:43:22 -0400
Subject: [racket] users Digest, Vol 109, Issue 62
In-Reply-To: <mailman.460.1411825801.679.users@racket-lang.org>
References: <mailman.460.1411825801.679.users@racket-lang.org>
Message-ID: <CAKpSqLDOqNa4uyg4aro9RQEpWbt-KfN6FzXo80BZ_iLdBoGxVg@mail.gmail.com>

Please take me off the list

Thanks

On Sat, Sep 27, 2014 at 9:50 AM,  <users-request at racket-lang.org> wrote:
> Send users mailing list submissions to
>         users at racket-lang.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>         http://lists.racket-lang.org/users/listinfo
> or, via email, send a message with subject or body 'help' to
>         users-request at racket-lang.org
>
> You can reach the person managing the list at
>         users-owner at racket-lang.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of users digest..."
>
>
> [Racket Users list:
>  http://lists.racket-lang.org/users ]
>
>
> Today's Topics:
>
>    1. "do what i mean" for https://github links (Martin DeMello)
>    2. Re: typed racket and generic interfaces, is there a
>       workaround using properties? (Spencer florence)
>    3. problem with pict->bitmap and mrlib/write-animated-gif
>       (Martin DeMello)
>    4. proof assistants, DrRacket and Bootstrap (Bill Richter)
>    5. Fwd: JFP CFP (Robby Findler)
>    6. Re: typed racket and generic interfaces,  is there a
>       workaround using properties? (Alexander D. Knauth)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Fri, 26 Sep 2014 21:56:52 -0700
> From: Martin DeMello <martindemello at gmail.com>
> To: Racket Users List <users at racket-lang.org>
> Subject: [racket] "do what i mean" for https://github links
> Message-ID:
>         <CAFrFfuH93VtSwahAz7Zo9EBGoTgP2_z=xZU04WdjrmEerd0Smw at mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
>
> It took me a while of reading through the docs to figure out why trying to
> install a package via "https://github.com/user/pkg" failed for want of a
> MANIFEST file. I think the "do what I mean" installer should have
> autoconverted it to git:// , seeing as how github is already special-cased.
>
> (Also my first try of installing via the git clone url
> https://github.com/user/pkg.git should ideally have been supported too,
> since that's what github encourages you to copy/paste as a reference)
>
> martin
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140926/a07de57f/attachment-0001.html>
>
> ------------------------------
>
> Message: 2
> Date: Fri, 26 Sep 2014 22:07:43 -0700 (PDT)
> From: "Spencer florence" <spencerflorence at gmail.com>
> To: "Alexander D. Knauth" <alexander at knauth.org>
> Cc: Racket Users <users at racket-lang.org>
> Subject: Re: [racket] typed racket and generic interfaces, is there a
>         workaround using properties?
> Message-ID: <1411794462639.6720ba81 at Nodemailer>
> Content-Type: text/plain; charset="utf-8"
>
> I don?t think you can. You would need define the struct in an untyped module then require it via require/typed. This is why dict?s don?t work in typed/racket either.
>
> On Sat, May 24, 2014 at 9:39 PM, Alexander D. Knauth <alexander at knauth.org>
> wrote:
>
>> Do generic interfaces work using structure type properties, and if they do, is there a way to use generic interfaces through properties so that I can do it in typed racket?
>> Specifically I?m trying to use the gen:custom-write and gen:dict generic interfaces.  I can use prop:custom-write for the first one, but I don?t know what to do for gen:dict.
>> Otherwise I?ll just put it in an untyped submodule.
>> ____________________
>>   Racket Users list:
>>   http://lists.racket-lang.org/users
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140926/e61d292b/attachment-0001.html>
>
> ------------------------------
>
> Message: 3
> Date: Sat, 27 Sep 2014 00:27:29 -0700
> From: Martin DeMello <martindemello at gmail.com>
> To: Racket Users List <users at racket-lang.org>
> Subject: [racket] problem with pict->bitmap and
>         mrlib/write-animated-gif
> Message-ID:
>         <CAFrFfuHsyYBhWB01PnOf8E5HzwDM_69n0nAKfLyNn2UY1+nkhQ at mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
>
> Can't figure it out, but something in the interaction between pict->bitmap
> and write-animated-gif is causing the frames to display one on top of the
> other when viewing the gif.
>
> #lang racket
>
> (require pict
>          mrlib/gif)
>
> (define (draw-frame i)
>   (pict->bitmap (circle (* 50 i))))
>
> (write-animated-gif
>  (map draw-frame (sequence->list (in-range 1 10)))
>  10
>  "test1.gif"
>  #:loop? true
>  #:one-at-a-time? true)
>
> martin
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140927/9a4a2d74/attachment-0001.html>
>
> ------------------------------
>
> Message: 4
> Date: Sat, 27 Sep 2014 02:38:48 -0500
> From: Bill Richter <richter at math.northwestern.edu>
> To: users at racket-lang.org
> Subject: [racket] proof assistants, DrRacket and Bootstrap
> Message-ID:
>         <201409270738.s8R7cmcd022184 at poisson.math.northwestern.edu>
> Content-Type: text/plain; charset=utf-8
>
> I have a few questions that might be off-topic.  Are you interested in formal proofs?  Have you considered adapting DrRacket to give an integrated editor for a proof assistant?  The proof assistants Coq and Isabelle use jedit and ProofGeneral, which I think aren't nearly as nice as DrRacket.  I actually use HOL Light, which nobody uses an integrated editor for.  Here are the slides for a talk I gave at the Institut Henri Poincar? in the workshop ``Formalization of mathematics in proof assistants''
> http://www.math.northwestern.edu/~richter/RichterIHPslide.pdf
>
> HOL Light and Coq are written in OCaml, a dialect of ML, which is therefore similar to Scheme, but it has one difference that I wonder if anyone here's knows how to deal with.  Scheme is well-suited for writing a Scheme interpreter, because of the quote feature.  OCaml doesn't have a quote feature, so the question arises how to write an OCaml interpreter inside OCaml.  That's not quite what I want to do, but if anyone could explain how to do it, I'd be grateful.
>
> I have a 7th grade student I'm trying to teach Racket, and I'm a bit confused about Bootstrap and Program By Design.  Is there any reason for my student not to just read HtDP?
>
> --
> Best,
> Bill
>
>
> ------------------------------
>
> Message: 5
> Date: Sat, 27 Sep 2014 07:38:54 -0500
> From: Robby Findler <robby at eecs.northwestern.edu>
> To: Racket Users <users at racket-lang.org>
> Subject: [racket] Fwd: JFP CFP
> Message-ID:
>         <CAL3TdOMU-PHh++5qUz2RE9fRxHGd-zjB8=nMNdQA8_FOMSbiag at mail.gmail.com>
> Content-Type: text/plain; charset=UTF-8
>
> If you're working on parallel or concurrent programming, consider submitting!
>
> Robby
>
>
> ------------------------------------------------------------------------------
>                         CALL FOR PAPERS
>
>                        JFP Special Issue
>                                 on
>                   Parallel and Concurrent Programming
>
>                 Submission Deadline: 22 December 2014
>        Expected Publication Date: November/December issue 2015
>
>
> * Scope
>
> Functional languages are uniquely suited to provide programmers with a
> programming model for parallel and concurrent computing. This is reflected in
> the wide range of work that is currently underway, both on parallel and
> concurrent functional languages, but also on bringing functional language
> features to other programming languages. This has resulted in a rapidly
> growing number of practical applications.
>
> The Journal of Functional Programming will devote a special issue to
> functional programming for concurrent and parallel computing. The purpose of
> this special issue is to showcase the state of the art on how functional
> languages and functional concepts currently assist programmers with the task
> of managing the challenges of creating parallel and concurrent systems.
>
> Language designers as well as systems builders and application programmers are
> invited to submit papers describing the current state of the art of language
> support and experiences with functional programming in building real-world
> systems.
>
> We encourage the submission of consolidated, condensed and extended work based
> on prior conference and workshop publications.
>
> * Submission Details
>
> Manuscripts should be submitted to CUP's Manuscript Central formatted
> according to JFP's standards.  Authors should specify in their cover message
> that the paper is intended for the special issue and choose the appropriate
> desired editor from the menu. For submission and formatting details, please
> consult the Journal of Functional Programming [ http://bit.ly/jfpadvice ]
> website.
>
>
> Guest Editors
> ---------------------------------------------------------------------------------
> Gabriele Keller                             Fritz Henglein
> keller @cse.unsw.edu.au                     henglein at diku.dk
> School of Computer Science & Engineering    Department of Computer
> Science (DIKU)
> UNSW Sydney NSW 2052                        University of Copenhagen
> and                                         Universitetsparken 5
> The Software Systems Research Group@ NICTA  DK-2100 Copenhagen
> Australia                                   Denmark
> ---------------------------------------------------------------------------------
>
>
> ------------------------------
>
> Message: 6
> Date: Sat, 27 Sep 2014 09:49:48 -0400
> From: "Alexander D. Knauth" <alexander at knauth.org>
> To: Anthony Carrico <acarrico at memebeam.org>
> Cc: racket users list <users at racket-lang.org>
> Subject: Re: [racket] typed racket and generic interfaces,      is there a
>         workaround using properties?
> Message-ID: <39AEBCAD-6138-4FE6-84A5-6AFF1A1A466C at knauth.org>
> Content-Type: text/plain; charset=windows-1252
>
> Like this, I think:  (but for some reason it seems to print it 3 times?)
> #lang typed/racket
>
> (struct foo () #:transparent
>   #:property prop:custom-write
>   (lambda (this out mode)
>     (displayln "whatever")))
>
> (print (foo))
>
>
> On Sep 26, 2014, at 11:07 PM, Anthony Carrico <acarrico at memebeam.org> wrote:
>
>> Dredging up an old thread: So how exactly do you define custom-write for
>> a typed racket struct?
>>
>> On 05/25/2014 03:30 PM, Alexander D. Knauth wrote:
>>> Never mind I just found prop:dict.
>>>
>>> By the way it would probably be helpful if instead of the value for
>>> prop:dict being a vector, it was a hash table with the method names
>>> as the keys, so then it would be a lot easier to tell which procedure
>>> goes with which method, and also a lot easier to extend it while
>>> maintaining backwards compatibility.
>>>
>>> On May 24, 2014, at 10:38 PM, Alexander D. Knauth
>>> <alexander at knauth.org> wrote:
>>>
>>>> Do generic interfaces work using structure type properties, and if
>>>> they do, is there a way to use generic interfaces through
>>>> properties so that I can do it in typed racket?
>>>>
>>>> Specifically I?m trying to use the gen:custom-write and gen:dict
>>>> generic interfaces.  I can use prop:custom-write for the first one,
>>>> but I don?t know what to do for gen:dict.
>>>>
>>>> Otherwise I?ll just put it in an untyped submodule.
>>
>> --
>> Anthony Carrico
>>
>> ____________________
>>  Racket Users list:
>>  http://lists.racket-lang.org/users
>
>
>
>
> End of users Digest, Vol 109, Issue 62
> **************************************


From moshedeutsch115 at gmail.com  Mon Sep 29 16:43:33 2014
From: moshedeutsch115 at gmail.com (Moshe Deutsch)
Date: Mon, 29 Sep 2014 16:43:33 -0400
Subject: [racket] users Digest, Vol 109, Issue 59
In-Reply-To: <mailman.450.1411705000.679.users@racket-lang.org>
References: <mailman.450.1411705000.679.users@racket-lang.org>
Message-ID: <CAKpSqLAHua7CK-VD702TiiaVb8Ambmpwa=uunVkuYzawTbiUCw@mail.gmail.com>

Please take me off the list

Thanks

On Fri, Sep 26, 2014 at 12:16 AM,  <users-request at racket-lang.org> wrote:
> Send users mailing list submissions to
>         users at racket-lang.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>         http://lists.racket-lang.org/users/listinfo
> or, via email, send a message with subject or body 'help' to
>         users-request at racket-lang.org
>
> You can reach the person managing the list at
>         users-owner at racket-lang.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of users digest..."
>
>
> [Racket Users list:
>  http://lists.racket-lang.org/users ]
>
>
> Today's Topics:
>
>    1. Re: typed racket, filters, and polymorphism (Sam Tobin-Hochstadt)
>    2. Re: web server: module servlets (Jay McCarthy)
>    3. Re: [GENERAL] Off Topic: Anybody reading this via
>       news.gmane.org? (George Neuner)
>    4. Re: typed racket, filters, and polymorphism (Alexander D. Knauth)
>    5. Re: web server: module servlets (George Neuner)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Thu, 25 Sep 2014 18:19:37 -0400
> From: Sam Tobin-Hochstadt <samth at cs.indiana.edu>
> To: "Alexander D. Knauth" <alexander at knauth.org>
> Cc: racket users list <users at racket-lang.org>
> Subject: Re: [racket] typed racket, filters, and polymorphism
> Message-ID:
>         <CAK=HD+ZpPuvYTMZJ1_LXrnPGjcPamZ73Z8aiPizCussR+UxYsw at mail.gmail.com>
> Content-Type: text/plain; charset=UTF-8
>
> No, I don't think you can do this. Can you say more about what you're
> trying to accomplish?
>
> Sam
>
> On Thu, Sep 25, 2014 at 6:15 PM, Alexander D. Knauth
> <alexander at knauth.org> wrote:
>> Do any of you have any advice for getting a function like this to
>> type-check?
>> #lang typed/racket
>>
>> (: check-int : (All (a) (case-> [a -> a]
>>                                 [Any -> Integer])))
>> (define (check-int int)
>>   (unless (exact-integer? int)
>>     (error 'check-int "expected Integer, given ~v" int))
>>   int)
>>
>> ;. Type Checker: type mismatch
>> ;  expected: a
>> ;  given: Integer in: int
>>
>>
>>
>> ____________________
>>   Racket Users list:
>>   http://lists.racket-lang.org/users
>>
>
>
> ------------------------------
>
> Message: 2
> Date: Thu, 25 Sep 2014 18:26:58 -0400
> From: Jay McCarthy <jay.mccarthy at gmail.com>
> To: George Neuner <gneuner2 at comcast.net>
> Cc: users <users at racket-lang.org>
> Subject: Re: [racket] web server: module servlets
> Message-ID:
>         <CAJYbDamsu2F=97vnZJQdx2dRO-Ss46NU0icrf8XwNA_2mMg9gw at mail.gmail.com>
> Content-Type: text/plain; charset=UTF-8
>
> On Thu, Sep 25, 2014 at 4:34 PM, George Neuner <gneuner2 at comcast.net> wrote:
>> Hi Jay,
>>
>> On 9/25/2014 1:04 PM, Jay McCarthy wrote:
>>
>>> If a Racket library is deliberately put into the servlet-namespace, does
>>> that streamline linking?
>>
>> Your assumption about the purpose of this is not correct. Anything in
>> the servlet-namespace will be shared between all servlets, and thus
>> not loaded per-servlet.
>>
>>
>> I think you misunderstood my question, but you may have answered it anyway.
>>
>> What I really was asking was whether each custodian was having to
>> individually load common Racket libraries [ web-server/*, net/*, etc. ] for
>> a new servlet rather than all servlet custodians sharing libraries that are
>> already loaded.  Or, if not actually "loading" the libraries, then having to
>> go to disk to check dependencies.
>>
>> So if I create a shared servlet namespace and put, e.g.,
>> "web-server/servlet" into it, would that in any way speed up starting a new
>> dynamically loaded servlet?
>
> There is a very small set that is automatically shared:
> racket/base
> web-server/private/servlet
> web-server/http
> web-server/servlet/web
> ...
>
> web-server/servlet is NOT shared, nor is any net library or common
> racket library like racket/list. They are all totally unique per
> servlet.
>
> This, by the way, is part of why I don't recommend using dynamic
> servlets at all and suggest using serve/servlet.
>
> Jay
>
>
>
> --
> Jay McCarthy
> http://jeapostrophe.github.io
>
>            "Wherefore, be not weary in well-doing,
>       for ye are laying the foundation of a great work.
> And out of small things proceedeth that which is great."
>                           - D&C 64:33
>
>
> ------------------------------
>
> Message: 3
> Date: Thu, 25 Sep 2014 20:13:54 -0400
> From: George Neuner <gneuner2 at comcast.net>
> To: Adrian Klaver <adrian.klaver at aklaver.com>, neil at neilvandyke.org
> Cc: "users at racket-lang.org >> users" <users at racket-lang.org>
> Subject: Re: [racket] [GENERAL] Off Topic: Anybody reading this via
>         news.gmane.org?
> Message-ID: <5424AFC2.6020302 at comcast.net>
> Content-Type: text/plain; charset="windows-1252"; Format="flowed"
>
>
> On 9/25/2014 5:26 PM, Neil Van Dyke wrote:
>> You can check your IP addr against the list at
>> "http://gmane.org/denied.php".
>>
>> Neil V.
>
> On 9/25/2014 7:08 PM, Adrian Klaver wrote:
>>
>> Take a look here:
>>
>> http://gmane.org/denied.php
>>
>> My guess is you are the fourth one from the bottom.
>>
>>
>> Might want to take a look at this thread to see what your options are
>> and what the turn around time is on your request:
>>
>> http://thread.gmane.org/gmane.discuss/16309
>
> Thanks to both of you - I missed seeing the entry about the denied list
> in the FAQ.
>
> Adrian you were right - it seems that I have been blocked for some
> reason.  Based on the description of infractions I really don't think I
> did anything to warrant it ... but there it is.  I'll have to follow my
> lists by email until I can unblocked.
>
> Thanks again,
> George
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140925/0d2a1f73/attachment-0001.html>
>
> ------------------------------
>
> Message: 4
> Date: Thu, 25 Sep 2014 21:42:00 -0400
> From: "Alexander D. Knauth" <alexander at knauth.org>
> To: Sam Tobin-Hochstadt <samth at cs.indiana.edu>
> Cc: racket users list <users at racket-lang.org>
> Subject: Re: [racket] typed racket, filters, and polymorphism
> Message-ID: <F16C1892-F09C-4A9F-94E2-16D5DE72F57D at knauth.org>
> Content-Type: text/plain; charset="windows-1252"
>
> What I?m trying to accomplish is something more like this:
> #lang typed/racket
>
> (require "dimensions.rkt")
>
> (struct (d) unit ([name : Any] [scalar : Positive-Real] [dimension : d]) #:transparent
>   #:guard (lambda (name scalar dimension _)
>             (unless (dimension? dimension)
>               (error 'unit "expected Dimension, given ~v" dimension))
>             (values name scalar dimension)))
>
> (define-type (Unitof d) (unit d))
>
> (define-type Unit (Unitof Dimension))
>
> (define Unit? (make-predicate Unit))
>
> (define-type Unitish
>   (U (Unitof Any)
>      Dimension
>      Positive-Real))
>
> (: ->unit : (All (d) (case-> [(Unitof d) -> (Unitof d)]
>                              [Unitish -> Unit])))
> (define (->unit u)
>   (cond [(unit? u)
>          (unless (Unit? u) ; this should never happen anyway because of the guard
>            (error '->unit "expected (Unitof Dimension), given ~v" u))
>          u]
>         [(dimension? u) (unit u 1 u)]
>         [(positive-real? u) (unit u u dimensionless-dimension)]))
>
>
> On Sep 25, 2014, at 6:19 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> wrote:
>
>> No, I don't think you can do this. Can you say more about what you're
>> trying to accomplish?
>>
>> Sam
>>
>> On Thu, Sep 25, 2014 at 6:15 PM, Alexander D. Knauth
>> <alexander at knauth.org> wrote:
>>> Do any of you have any advice for getting a function like this to
>>> type-check?
>>> #lang typed/racket
>>>
>>> (: check-int : (All (a) (case-> [a -> a]
>>>                                [Any -> Integer])))
>>> (define (check-int int)
>>>  (unless (exact-integer? int)
>>>    (error 'check-int "expected Integer, given ~v" int))
>>>  int)
>>>
>>> ;. Type Checker: type mismatch
>>> ;  expected: a
>>> ;  given: Integer in: int
>>>
>>>
>>>
>>> ____________________
>>>  Racket Users list:
>>>  http://lists.racket-lang.org/users
>>>
>
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140925/dc09b97f/attachment-0001.html>
>
> ------------------------------
>
> Message: 5
> Date: Fri, 26 Sep 2014 00:16:56 -0400
> From: George Neuner <gneuner2 at comcast.net>
> To: Jay McCarthy <jay.mccarthy at gmail.com>
> Cc: users <users at racket-lang.org>
> Subject: Re: [racket] web server: module servlets
> Message-ID: <5424E8B8.8040501 at comcast.net>
> Content-Type: text/plain; charset="utf-8"; Format="flowed"
>
> On 9/25/2014 6:26 PM, Jay McCarthy wrote:
>> web-server/servlet is NOT shared, nor is any net library or common
>> racket library like racket/list. They are all totally unique per
>> servlet.
>>
>> This, by the way, is part of why I don't recommend using dynamic
>> servlets at all and suggest using serve/servlet.
>
> But ... IIUC ... a listener started by  serve/servlet  will still
> dynamically load from <servlet-root>, <server-root>/htdocs/*, etc. if
> the request doesn't match a hard coded dispatch URL - the regex of the
> initial servlet or an entry in a dispatcher table.  At least that's the
> behavior I see with my own application: my initial servlet returns 404
> when called - the call to  serve/servlet  just sets up the environment
> and and all the "real" servlets are demand loaded from disk when first
> touched.
>
> Are you advocating *static* linking and essentially just dispatching to
> internal functions by URL?  ISTM that that defeats the purpose.
>
> George
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140926/ffaf6bde/attachment.html>
>
> End of users Digest, Vol 109, Issue 59
> **************************************

From moshedeutsch115 at gmail.com  Mon Sep 29 16:43:44 2014
From: moshedeutsch115 at gmail.com (Moshe Deutsch)
Date: Mon, 29 Sep 2014 16:43:44 -0400
Subject: [racket] users Digest, Vol 109, Issue 57
In-Reply-To: <mailman.444.1411664691.679.users@racket-lang.org>
References: <mailman.444.1411664691.679.users@racket-lang.org>
Message-ID: <CAKpSqLDbs_MHTA_YJJmf=7MQZeBXNGcCGoOU1UCFanRHpv7Pyg@mail.gmail.com>

Please take me off the list

Thanks

On Thu, Sep 25, 2014 at 1:04 PM,  <users-request at racket-lang.org> wrote:
> Send users mailing list submissions to
>         users at racket-lang.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>         http://lists.racket-lang.org/users/listinfo
> or, via email, send a message with subject or body 'help' to
>         users-request at racket-lang.org
>
> You can reach the person managing the list at
>         users-owner at racket-lang.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of users digest..."
>
>
> [Racket Users list:
>  http://lists.racket-lang.org/users ]
>
>
> Today's Topics:
>
>    1. htdp: exercise 21.1.3 (Daniel Bastos)
>    2. Re: web server: module servlets (George Neuner)
>    3. ANN: DOS, Delimited-continuation-based Operating-system
>       Simulator (Jay McCarthy)
>    4. Re: web server: module servlets (Jay McCarthy)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Thu, 25 Sep 2014 13:48:05 -0300
> From: Daniel Bastos <dbastos at toledo.com>
> To: Racket Users <users at racket-lang.org>
> Subject: [racket] htdp: exercise 21.1.3
> Message-ID:
>         <CAB08-qr1MMDnJ6uSztfx+qnJMGh3QYztK5iC3_xWiMhKz-onFA at mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
>
> The solution is missing, but I thought of exposing this one here so that it
> gets scrutinized before sending it to Robby Findler.
>
> ;; Exercise 21.1.3. Define natural-f, which is the abstraction of the
> ;; following two functions:
> ;;
> ;; ;; copy : N X  ->  (listof X)
> ;; ;; to create a list that contains
> ;; ;; obj n times
> ;; (define (copy n obj)
> ;;   (cond
> ;;     [(zero? n) empty]
> ;;     [else (cons obj
> ;;                 (copy (sub1 n) obj))]))
> ;;
> ;; ;; n-adder : N number  ->  number
> ;; ;; to add n to x using
> ;; ;; (+ 1 ...) only
> ;; (define (n-adder n x)
> ;;   (cond
> ;;     [(zero? n) x]
> ;;     [else (+ 1
> ;;              (n-adder (sub1 n) x))]))
> ;;
> ;; Don't forget to test natural-f. Also use natural-f to define
> ;; n-multiplier, which consumes n and x and produces n times x with
> ;; additions only. Use the examples to formulate a contract.
> ;;
> ;; Hint: The two function differ more than, say, the functions sum and
> ;; product in exercise 21.1.2. In particular, the base case in one
> ;; instance is a argument of the function, where in the other it is just
> ;; a constant value.
> ;;
> ;; (*) Solution
> ;;
> ;; After following the design recipe for abstraction, we get natural-f as
> ;; shown below. To write the contract, let's first only consider copy,
> ;; then we consider only n-adder and then we consider both.
> ;;
> ;; ;; natural-f-copy: Nat X (X (listof X) -> (listof X)) (listof X) ->
> (listof X)
> ;; ;; natural-f-adder: Nat number (number number -> number) number -> number
> ;;
> ;; The contract for natural-f-adder is more specific than
> ;; natural-f-copy. The difference between these is that in the first, two
> ;; different types are involved --- X and (listof X). Therefore, (listof
> ;; X) works as another variable, so let's call it Y in the final version.
>
> ;; natural-f: Nat X (X Y -> Y) Y -> Y
> (define (natural-f n x f init)
>   (cond
>     [(zero? n) init]
>     [else (f x
>               (natural-f (sub1 n) x f init))]))
>
> ;; (*) Tests
>
> ;; Let's use two different types in copy, say symbol and then number.
>
> (check-expect (copy 3 'a) '(a a a))
> (check-expect (natural-f 3 'a cons empty) (copy 3 'a))
>
> (check-expect (copy 0 'a) empty)
> (check-expect (natural-f 0 'a cons empty) (copy 0 'a))
>
> (check-expect (copy 3 1) (list 1 1 1))
> (check-expect (natural-f 3 1 cons empty) (copy 3 1))
>
> ;; For n-adder, let's make sure to test different init-values.
>
> (check-expect (n-adder 3 3.14) 6.14)
> (check-expect (natural-f 3 1 + 3.14) (n-adder 3 3.14))
>
> (check-expect (n-adder 3 1) 4)
> (check-expect (natural-f 3 1 + 1) (n-adder 3 1))
>
> (check-expect (n-adder 3 2) 5)
> (check-expect (natural-f 3 1 + 2) (n-adder 3 2))
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140925/ea097a92/attachment-0001.html>
>
> ------------------------------
>
> Message: 2
> Date: Thu, 25 Sep 2014 12:51:56 -0400
> From: George Neuner <gneuner2 at comcast.net>
> To: Jay McCarthy <jay.mccarthy at gmail.com>
> Cc: users <users at racket-lang.org>
> Subject: Re: [racket] web server: module servlets
> Message-ID: <5424482C.1030901 at comcast.net>
> Content-Type: text/plain; charset="utf-8"; Format="flowed"
>
> Hi Jay,
>
> On 9/23/2014 10:03 PM, Jay McCarthy wrote:
>> The command-line tool is basically deprecated and only provided for
>> backwards compatibility. There is a huge amount that it can't do at
>> all and it hasn't been the primary way that we recommend using the Web
>> server for a very long time.
>
> I'm not using the plt-web-server app - I created a minimal application
> that set up the environment: directories, ports, etc. and a start
> function that just returns a 404 if called.  My server sits behind
> Apache and only handles servlets, which I would like to be demand load
> modules so they can be added to and updated easily. That's why I am
> interested in being able to unload servlets on command, though not
> necessarily all of them at once (although that also is helpful).
>
>
>> > Not really knowing much about Racket's internals he naively asks:
>> >Coulda server  pre-load the commonly used webserver modules and make
>> >themavailableto new module servlets, or does the custodian
>> >implementation  make doing thatdifficult/impossible?
>>
>> This is the purpose of the make-servlet-namespace argument of
>> configuration-table->web-config@ but there is no option in the
>> configuration file for that argument.
>>
>> http://docs.racket-lang.org/web-server-internal/Web_Servers.html?q=servlet-namespace#%28def._web-config._%28%28lib._web-server%2Fweb-config-unit..rkt%29._configuration-table-~3eweb-config~40%29%29
>
> The namespace facility seems designed more for user written modules than
> for library modules.
>
> My analysis may be off-base as I have very little experience working
> directly with custodians, but it seems that when a dynamic servlet is
> first loaded, its custodian spends a lot of additional time (re)loading
> library modules that already exist in other custodians. [ Though I can't
> tell exactly what's happening, I can see a lot of disk activity when I
> think I'm loading a 5KB servlet. ]  Reasonably I would have expected
> that after loading/linking the first servlet, the libraries common to
> the servlets would be already in memory. But loading additional servlets
> is no quicker [ and causes a similar disk hit ] so clearly I don't
> understand what is happening internally with the custodians.
>
> If a Racket library is deliberately put into the servlet-namespace, does
> that streamline linking?
>
>
>> > My application so far is based on stateful servlets and AJAX ... stateful
>> > mainly because it's easier for me to understand.  Currently there is
>> >little use ofcontinuations, but some planned functionality will use them
>> >  extensively and it certainly wouldhelp if debugging didn't always mean
>> >starting over setting up conditions in theapplication.
>>
>> This comment/question is related to questions 4 and 5 from the FAQ:
>>
>> http://docs.racket-lang.org/web-server/faq.html?q=servlet-namespace#%28part._update-servlets%29
>
> That link leads to a "troubleshooting" page 8-).  I didn't consider the
> issue to be a "problem" per se - I already knew that restarting a
> stateful servlet would lose saved state.  Most of my existing servlets
> are one-shots that don't save any state, but they are written using the
> stateful language.  Only a few use continuations and not extensively (so
> far).  I was just thinking ahead to stuff that will need to use
> continuations more extensively.
>
> Thanks for putting up with my questions.
> George
>
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140925/7c368b80/attachment-0001.html>
>
> ------------------------------
>
> Message: 3
> Date: Thu, 25 Sep 2014 13:02:00 -0400
> From: Jay McCarthy <jay.mccarthy at gmail.com>
> To: users <users at racket-lang.org>
> Subject: [racket] ANN: DOS,     Delimited-continuation-based
>         Operating-system Simulator
> Message-ID:
>         <CAJYbDam0maAG7ZmCBjbha1uqhJyCpX9J+L3=kxqAL4GjdKUvkg at mail.gmail.com>
> Content-Type: text/plain; charset=UTF-8
>
> I've just released the game architecture library I talked about in
> part 2 of my RacketCon talk.
>
> The big picture of this library is to make World-style programs more
>  compositional by (a) using continuations to hide the internal state
> (including control state) of components and (b) using environments as
> a standard monoid-based inter-component communication channel. A
> monoid is used to ensure that the components can be evaluated in any
> order. Despite assumptions some have about continuations and pure
> functional programming, it is incredibly efficient and can run at
> 60FPS, as demonstrated in get-bonus.
>
> You can get it with
>
> raco pkg install dos
>
> And you can try out the demo with
>
> racket -l dos/examples/win
>
> The demo source is a bare 39 lines:
>
> https://github.com/jeapostrophe/dos/blob/master/dos/examples/win.rkt
>
> and I provide the non-DOS version for comparison:
>
> https://github.com/jeapostrophe/dos/blob/master/dos/examples/win-long.rkt
>
> The core library has a mere 36 lines:
>
> https://github.com/jeapostrophe/dos/blob/master/dos/main.rkt
>
> --
> Jay McCarthy
> http://jeapostrophe.github.io
>
>            "Wherefore, be not weary in well-doing,
>       for ye are laying the foundation of a great work.
> And out of small things proceedeth that which is great."
>                           - D&C 64:33
>
>
> ------------------------------
>
> Message: 4
> Date: Thu, 25 Sep 2014 13:04:44 -0400
> From: Jay McCarthy <jay.mccarthy at gmail.com>
> To: George Neuner <gneuner2 at comcast.net>
> Cc: users <users at racket-lang.org>
> Subject: Re: [racket] web server: module servlets
> Message-ID:
>         <CAJYbDa=-DuRMVy7JnCt32Y8waG2+qkHZQw5cVDL4L_d6zth45w at mail.gmail.com>
> Content-Type: text/plain; charset=UTF-8
>
> On Thu, Sep 25, 2014 at 12:51 PM, George Neuner <gneuner2 at comcast.net> wrote:
>> Hi Jay,
>>
>> On 9/23/2014 10:03 PM, Jay McCarthy wrote:
>>
>> The command-line tool is basically deprecated and only provided for
>> backwards compatibility. There is a huge amount that it can't do at
>> all and it hasn't been the primary way that we recommend using the Web
>> server for a very long time.
>>
>>
>> I'm not using the plt-web-server app - I created a minimal application that
>> set up the environment: directories, ports, etc. and a start function that
>> just returns a 404 if called.  My server sits behind Apache and only handles
>> servlets, which I would like to be demand load modules so they can be added
>> to and updated easily.  That's why I am interested in being able to unload
>> servlets on command, though not necessarily all of them at once (although
>> that also is helpful).
>>
>>
>>> Not really knowing much about Racket's internals he naively asks:
>>> Could a server  pre-load the commonly used webserver modules and make
>>> them available to new module servlets, or does the custodian
>>> implementation make doing that difficult/impossible?
>>
>> This is the purpose of the make-servlet-namespace argument of
>> configuration-table->web-config@ but there is no option in the
>> configuration file for that argument.
>>
>> http://docs.racket-lang.org/web-server-internal/Web_Servers.html?q=servlet-namespace#%28def._web-config._%28%28lib._web-server%2Fweb-config-unit..rkt%29._configuration-table-~3eweb-config~40%29%29
>>
>>
>> The namespace facility seems designed more for user written modules than for
>> library modules.
>>
>> My analysis may be off-base as I have very little experience working
>> directly with custodians, but it seems that when a dynamic servlet is first
>> loaded, its custodian spends a lot of additional time (re)loading library
>> modules that already exist in other custodians.  [ Though I can't tell
>> exactly what's happening, I can see a lot of disk activity when I think I'm
>> loading a 5KB servlet. ]  Reasonably I would have expected that after
>> loading/linking the first servlet, the libraries common to the servlets
>> would be already in memory.  But loading additional servlets is no quicker [
>> and causes a similar disk hit ] so clearly I don't understand what is
>> happening internally with the custodians.
>>
>> If a Racket library is deliberately put into the servlet-namespace, does
>> that streamline linking?
>>
>
> Your assumption about the purpose of this is not correct. Anything in
> the servlet-namespace will be shared between all servlets, and thus
> not loaded per-servlet.
>
>>> My application so far is based on stateful servlets and AJAX ... stateful
>>> mainly because it's easier for me to understand.  Currently there is
>>> little use of continuations, but some planned functionality will use them
>>> extensively and it certainly would help if debugging didn't always mean
>>> starting over setting up conditions in the application.
>>
>> This comment/question is related to questions 4 and 5 from the FAQ:
>>
>> http://docs.racket-lang.org/web-server/faq.html?q=servlet-namespace#%28part._update-servlets%29
>>
>>
>> That link leads to a "troubleshooting" page 8-).  I didn't consider the
>> issue to be a "problem" per se - I already knew that restarting a stateful
>> servlet would lose saved state.  Most of my existing servlets are one-shots
>> that don't save any state, but they are written using the stateful language.
>> Only a few use continuations and not extensively (so far).  I was just
>> thinking ahead to stuff that will need to use continuations more
>> extensively.
>>
>> Thanks for putting up with my questions.
>> George
>>
>
>
>
> --
> Jay McCarthy
> http://jeapostrophe.github.io
>
>            "Wherefore, be not weary in well-doing,
>       for ye are laying the foundation of a great work.
> And out of small things proceedeth that which is great."
>                           - D&C 64:33
>
>
> End of users Digest, Vol 109, Issue 57
> **************************************

From moshedeutsch115 at gmail.com  Mon Sep 29 16:43:54 2014
From: moshedeutsch115 at gmail.com (Moshe Deutsch)
Date: Mon, 29 Sep 2014 16:43:54 -0400
Subject: [racket] users Digest, Vol 109, Issue 56
In-Reply-To: <mailman.1.1411660802.25600.users@racket-lang.org>
References: <mailman.1.1411660802.25600.users@racket-lang.org>
Message-ID: <CAKpSqLDYaBgt2=89avzrE1dUUicku-h9a7Frct4Yhouu=4ueHw@mail.gmail.com>

Please take me off the list

Thanks

On Thu, Sep 25, 2014 at 12:00 PM,  <users-request at racket-lang.org> wrote:
> Send users mailing list submissions to
>         users at racket-lang.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>         http://lists.racket-lang.org/users/listinfo
> or, via email, send a message with subject or body 'help' to
>         users-request at racket-lang.org
>
> You can reach the person managing the list at
>         users-owner at racket-lang.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of users digest..."
>
>
> [Racket Users list:
>  http://lists.racket-lang.org/users ]
>
>
> Today's Topics:
>
>    1.  Unquoting a variable without unquoting its contents? (J Arcane)
>    2. Re: macro question (Jay McCarthy)
>    3. Re: macro question (Alejandro Zamora)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Thu, 25 Sep 2014 16:52:47 +0300
> From: J Arcane <jarcane at gmail.com>
> To: Racket Users <users at racket-lang.org>
> Subject: [racket]  Unquoting a variable without unquoting its
>         contents?
> Message-ID:
>         <CA+HVF0Ddf_WybAgJAAnJBvYeVoYMXhcN=4Kp4YF1_v-a9TSwGg at mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
>
> http://try-racket.org continues to splutter to life, but there's still
> something of a glaring error in that it doesn't really like list or quote
> at all.
>
> AlexKnauth has narrowed the problem down here:
> https://github.com/jarcane/try-racket/issues/4#issuecomment-56751603
>
> It seems that unquoting a variable does funny things to the variable's
> contents when they contain a bare quoted list or symbol, so we get weird
> results and errors. Quotes get lost, and then the eval finds itself looking
> at mangled forms.
>
> The best I can think of to do is maybe make a simple macro rather than
> using quasiquote? Any ideas?
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140925/e97f0d3f/attachment-0001.html>
>
> ------------------------------
>
> Message: 2
> Date: Thu, 25 Sep 2014 10:17:49 -0400
> From: Jay McCarthy <jay.mccarthy at gmail.com>
> To: Alejandro Zamora <terefv at ltu.sld.cu>
> Cc: Racket Users <users at racket-lang.org>
> Subject: Re: [racket] macro question
> Message-ID:
>         <CAJYbDakR7Sxa55rRiZoFinREGEdZyT7V4d+QRD0f_G1CA4j-GA at mail.gmail.com>
> Content-Type: text/plain; charset=UTF-8
>
> I won't answer the question about R5RS, but in Racket you can do this
> fairly succinctly:
>
> #lang racket/base
> (require (for-syntax racket/base
>                      racket/syntax))
>
> (define-syntax (vars stx)
>   (syntax-case stx ()
>     [(_ id count val)
>      (with-syntax
>          ([(id_i ...)
>            (for/list ([i (in-range 1 (add1 (syntax->datum #'count)))])
>              (format-id #'id "~a~a" #'id i))])
>      (syntax/loc stx
>        (begin (define id_i val)
>               ...)))]))
>
> (vars id 5 null)
>
> (vector id1 id2 id3 id4 id5)
>
> 2014-09-24 22:35 GMT-04:00 Alejandro Zamora <terefv at ltu.sld.cu>:
>> Hi Racket people:
>>         I have a (maybe too simple) question:
>>         How I create one macro like this (but using define-macro &
>>         R5RS language only)?:
>>
>>         (vars sym var-cant val-of-vars)
>>
>>         Example:
>>         (vars id 5 null) -> (prog (define id1 null)
>>                                   (define id2 null)
>>                                   (define id3 null)
>>                                   (define id4 null)
>>                                   (define id5 null))
>>
>> --
>> Nunca digas nunca, di mejor: gracias, permiso, disculpe.
>>
>> Este mensaje le ha llegado mediante el servicio de correo electronico que ofrece Infomed para respaldar el cumplimiento de las misiones del Sistema Nacional de Salud. La persona que envia este correo asume el compromiso de usar el servicio a tales fines y cumplir con las regulaciones establecidas
>>
>> Infomed: http://www.sld.cu/
>>
>> ____________________
>>   Racket Users list:
>>   http://lists.racket-lang.org/users
>
>
>
> --
> Jay McCarthy
> http://jeapostrophe.github.io
>
>            "Wherefore, be not weary in well-doing,
>       for ye are laying the foundation of a great work.
> And out of small things proceedeth that which is great."
>                           - D&C 64:33
>
>
>
> ------------------------------
>
> Message: 3
> Date: Thu, 25 Sep 2014 10:49:15 -0400
> From: Alejandro Zamora <terefv at ltu.sld.cu>
> To: Jay McCarthy <jay.mccarthy at gmail.com>
> Cc: Racket Users <users at racket-lang.org>
> Subject: Re: [racket] macro question
> Message-ID: <20140925104915.59291dc0 at alejandro-H61H2-CM>
> Content-Type: text/plain; charset=UTF-8
>
> Thanks Jay!!
> nice solution!!
> I'm asking for define-macro & R5RS because I'm trying to use one
> R5RS Embedded Scheme Interpreter on Python that have only define-macro
> construction for make macros and I want to use it.
>
> greetings!
>
>  El Thu, 25 Sep 2014 10:17:49 -0400
> Jay McCarthy <jay.mccarthy at gmail.com> escribi?:
>> I won't answer the question about R5RS, but in Racket you can do this
>> fairly succinctly:
>>
>> #lang racket/base
>> (require (for-syntax racket/base
>>                      racket/syntax))
>>
>> (define-syntax (vars stx)
>>   (syntax-case stx ()
>>     [(_ id count val)
>>      (with-syntax
>>          ([(id_i ...)
>>            (for/list ([i (in-range 1 (add1 (syntax->datum #'count)))])
>>              (format-id #'id "~a~a" #'id i))])
>>      (syntax/loc stx
>>        (begin (define id_i val)
>>               ...)))]))
>>
>> (vars id 5 null)
>>
>> (vector id1 id2 id3 id4 id5)
>>
>> 2014-09-24 22:35 GMT-04:00 Alejandro Zamora <terefv at ltu.sld.cu>:
>> > Hi Racket people:
>> >         I have a (maybe too simple) question:
>> >         How I create one macro like this (but using define-macro &
>> >         R5RS language only)?:
>> >
>> >         (vars sym var-cant val-of-vars)
>> >
>> >         Example:
>> >         (vars id 5 null) -> (prog (define id1 null)
>> >                                   (define id2 null)
>> >                                   (define id3 null)
>> >                                   (define id4 null)
>> >                                   (define id5 null))
>> >
>> > --
>> > Nunca digas nunca, di mejor: gracias, permiso, disculpe.
>> >
>> > Este mensaje le ha llegado mediante el servicio de correo
>> > electronico que ofrece Infomed para respaldar el cumplimiento de
>> > las misiones del Sistema Nacional de Salud. La persona que envia
>> > este correo asume el compromiso de usar el servicio a tales fines y
>> > cumplir con las regulaciones establecidas
>> >
>> > Infomed: http://www.sld.cu/
>> >
>> > ____________________
>> >   Racket Users list:
>> >   http://lists.racket-lang.org/users
>>
>>
>>
>
>
> --
> Nunca digas nunca, di mejor: gracias, permiso, disculpe.
>
> Este mensaje le ha llegado mediante el servicio de correo electronico que ofrece Infomed para respaldar el cumplimiento de las misiones del Sistema Nacional de Salud. La persona que envia este correo asume el compromiso de usar el servicio a tales fines y cumplir con las regulaciones establecidas
>
> Infomed: http://www.sld.cu/
>
>
>
>
> End of users Digest, Vol 109, Issue 56
> **************************************


From moshedeutsch115 at gmail.com  Mon Sep 29 16:44:05 2014
From: moshedeutsch115 at gmail.com (Moshe Deutsch)
Date: Mon, 29 Sep 2014 16:44:05 -0400
Subject: [racket] users Digest, Vol 109, Issue 55
In-Reply-To: <mailman.440.1411649827.679.users@racket-lang.org>
References: <mailman.440.1411649827.679.users@racket-lang.org>
Message-ID: <CAKpSqLDjXF=02wWUG=vD5OojAr2yYjyihBLfPtC1GkEnOQAMRw@mail.gmail.com>

Please take me off the list

Thanks

On Thu, Sep 25, 2014 at 8:57 AM,  <users-request at racket-lang.org> wrote:
> Send users mailing list submissions to
>         users at racket-lang.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>         http://lists.racket-lang.org/users/listinfo
> or, via email, send a message with subject or body 'help' to
>         users-request at racket-lang.org
>
> You can reach the person managing the list at
>         users-owner at racket-lang.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of users digest..."
>
>
> [Racket Users list:
>  http://lists.racket-lang.org/users ]
>
>
> Today's Topics:
>
>    1. macro question (Alejandro Zamora)
>    2. Re: ranking items by their appearance in sets (David T. Pierson)
>    3. Switching from editor to repl (C K Kashyap)
>    4. Re: Switching from editor to repl (Jay McCarthy)
>    5. Re: Switching from editor to repl (C K Kashyap)
>    6. Re: Switching from editor to repl (Robby Findler)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Wed, 24 Sep 2014 22:35:49 -0400
> From: Alejandro Zamora <terefv at ltu.sld.cu>
> To: Racket Users <users at racket-lang.org>
> Subject: [racket] macro question
> Message-ID: <20140924223549.7f15c912 at alejandro-H61H2-CM>
> Content-Type: text/plain; charset=US-ASCII
>
> Hi Racket people:
>         I have a (maybe too simple) question:
>         How I create one macro like this (but using define-macro &
>         R5RS language only)?:
>
>         (vars sym var-cant val-of-vars)
>
>         Example:
>         (vars id 5 null) -> (prog (define id1 null)
>                                   (define id2 null)
>                                   (define id3 null)
>                                   (define id4 null)
>                                   (define id5 null))
>
> --
> Nunca digas nunca, di mejor: gracias, permiso, disculpe.
>
> Este mensaje le ha llegado mediante el servicio de correo electronico que ofrece Infomed para respaldar el cumplimiento de las misiones del Sistema Nacional de Salud. La persona que envia este correo asume el compromiso de usar el servicio a tales fines y cumplir con las regulaciones establecidas
>
> Infomed: http://www.sld.cu/
>
>
>
> ------------------------------
>
> Message: 2
> Date: Wed, 24 Sep 2014 23:07:13 -0400
> From: "David T. Pierson" <dtp at mindstory.com>
> To: Matthias Felleisen <matthias at ccs.neu.edu>
> Cc: users at racket-lang.org
> Subject: Re: [racket] ranking items by their appearance in sets
> Message-ID: <1411613699.19584%dtp at mindstory.com>
> Content-Type: text/plain; charset=us-ascii
>
> On Wed, Sep 24, 2014 at 10:44:45AM -0400, Matthias Felleisen wrote:
>> Style: Here is my rewrite with style suggestions. I liked to have a
>> single point that tests all versions of a function when I conduct such
>> design experiments. I also want to have all my test code in the test
>> submodule so that it doesn't get deployed. That's why I am using a
>> syntactic abstraction that unfolds into a submodule rather than a
>> function that calls things.
>
> Thanks for the style suggestions.  I am reminded I need to reread the
> style guide.
>
>> Performance: I have not checked their performance but I cannot imagine
>> the first version to be performant. See my annotations.
>
> Good point.  I had put performance out of my mind because my real data
> is relatively small and I just wanted to play a bit with the methods,
> but I definitely appreciate the notes.
>
>> My choice: Your second version is what I would have written.
>
> Glad to know I was on the right track.
>
> Thank you.
>
> David
>
>
> ------------------------------
>
> Message: 3
> Date: Thu, 25 Sep 2014 11:47:57 +0530
> From: C K Kashyap <ckkashyap at gmail.com>
> To: users at racket-lang.org
> Subject: [racket] Switching from editor to repl
> Message-ID:
>         <CAGdT1gqUp-0hdF15FYo2T_Qpm+qf9AkKMak_iLV9HTC4WkDrtQ at mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
>
> Hi,
> I am new to racket (but am really excited by the idea - everything is a
> program) - need a quick help with Dr Racket -
>
> What's the keyboard shortcut to switch between editor and repl?
>
> I got to know from a video on youtube that I could use C-e to toggle the
> repl visibility but not able to find out how to get to the rpl using the
> keyboard.
> Regards,
> Kashyap
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140925/39d4382b/attachment-0001.html>
>
> ------------------------------
>
> Message: 4
> Date: Thu, 25 Sep 2014 02:33:55 -0400
> From: Jay McCarthy <jay.mccarthy at gmail.com>
> To: C K Kashyap <ckkashyap at gmail.com>
> Cc: users <users at racket-lang.org>
> Subject: Re: [racket] Switching from editor to repl
> Message-ID:
>         <CAJYbDamt+_RvG1utwzjJ1KW-m5ynRDxPr2xVivTOGQj8Y2R3DA at mail.gmail.com>
> Content-Type: text/plain; charset=UTF-8
>
> I use C-e and C-d to do this and I just hit it twice, once hides the
> side and the second shows it and focuses it.
>
> On Thu, Sep 25, 2014 at 2:17 AM, C K Kashyap <ckkashyap at gmail.com> wrote:
>> Hi,
>> I am new to racket (but am really excited by the idea - everything is a
>> program) - need a quick help with Dr Racket -
>>
>> What's the keyboard shortcut to switch between editor and repl?
>>
>> I got to know from a video on youtube that I could use C-e to toggle the
>> repl visibility but not able to find out how to get to the rpl using the
>> keyboard.
>> Regards,
>> Kashyap
>>
>> ____________________
>>   Racket Users list:
>>   http://lists.racket-lang.org/users
>>
>
>
>
> --
> Jay McCarthy
> http://jeapostrophe.github.io
>
>            "Wherefore, be not weary in well-doing,
>       for ye are laying the foundation of a great work.
> And out of small things proceedeth that which is great."
>                           - D&C 64:33
>
>
> ------------------------------
>
> Message: 5
> Date: Thu, 25 Sep 2014 13:13:58 +0530
> From: C K Kashyap <ckkashyap at gmail.com>
> To: Jay McCarthy <jay.mccarthy at gmail.com>
> Cc: users <users at racket-lang.org>
> Subject: Re: [racket] Switching from editor to repl
> Message-ID:
>         <CAGdT1gqYYRz1Kk11U1==oUgP93ELFR0PFPsdG5v8C_tYjwqH=g at mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
>
> C-d works for me ... thanks Jay.
> I was wondering though if I could've found this out in the documentation
> somewhere. I could not find it in the guide/google.
>
> Regards,
> Kashyap
>
> On Thu, Sep 25, 2014 at 12:03 PM, Jay McCarthy <jay.mccarthy at gmail.com>
> wrote:
>
>> I use C-e and C-d to do this and I just hit it twice, once hides the
>> side and the second shows it and focuses it.
>>
>> On Thu, Sep 25, 2014 at 2:17 AM, C K Kashyap <ckkashyap at gmail.com> wrote:
>> > Hi,
>> > I am new to racket (but am really excited by the idea - everything is a
>> > program) - need a quick help with Dr Racket -
>> >
>> > What's the keyboard shortcut to switch between editor and repl?
>> >
>> > I got to know from a video on youtube that I could use C-e to toggle the
>> > repl visibility but not able to find out how to get to the rpl using the
>> > keyboard.
>> > Regards,
>> > Kashyap
>> >
>> > ____________________
>> >   Racket Users list:
>> >   http://lists.racket-lang.org/users
>> >
>>
>>
>>
>> --
>> Jay McCarthy
>> http://jeapostrophe.github.io
>>
>>            "Wherefore, be not weary in well-doing,
>>       for ye are laying the foundation of a great work.
>> And out of small things proceedeth that which is great."
>>                           - D&C 64:33
>>
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <http://lists.racket-lang.org/users/archive/attachments/20140925/93aa6ef8/attachment-0001.html>
>
> ------------------------------
>
> Message: 6
> Date: Thu, 25 Sep 2014 07:56:53 -0500
> From: Robby Findler <robby at eecs.northwestern.edu>
> To: C K Kashyap <ckkashyap at gmail.com>
> Cc: Jay McCarthy <jay.mccarthy at gmail.com>, users
>         <users at racket-lang.org>
> Subject: Re: [racket] Switching from editor to repl
> Message-ID:
>         <CAL3TdONC6=U8EibFfxLYDNVVW0+SEo9+Do5ZuMSu2bLQRm8y8A at mail.gmail.com>
> Content-Type: text/plain; charset=UTF-8
>
> The guide focuses on the language, not the DrRacket IDE. This is a
> link to the IDE docs that has a list of keyboard shortcuts.
>
> You might also try looking at the menu items; things are a little more
> discoverable that way too.
>
> http://docs.racket-lang.org/drracket/Keyboard_Shortcuts.html
>
> hth,
> Robby
>
> On Thu, Sep 25, 2014 at 2:43 AM, C K Kashyap <ckkashyap at gmail.com> wrote:
>> C-d works for me ... thanks Jay.
>> I was wondering though if I could've found this out in the documentation
>> somewhere. I could not find it in the guide/google.
>>
>> Regards,
>> Kashyap
>>
>> On Thu, Sep 25, 2014 at 12:03 PM, Jay McCarthy <jay.mccarthy at gmail.com>
>> wrote:
>>>
>>> I use C-e and C-d to do this and I just hit it twice, once hides the
>>> side and the second shows it and focuses it.
>>>
>>> On Thu, Sep 25, 2014 at 2:17 AM, C K Kashyap <ckkashyap at gmail.com> wrote:
>>> > Hi,
>>> > I am new to racket (but am really excited by the idea - everything is a
>>> > program) - need a quick help with Dr Racket -
>>> >
>>> > What's the keyboard shortcut to switch between editor and repl?
>>> >
>>> > I got to know from a video on youtube that I could use C-e to toggle the
>>> > repl visibility but not able to find out how to get to the rpl using the
>>> > keyboard.
>>> > Regards,
>>> > Kashyap
>>> >
>>> > ____________________
>>> >   Racket Users list:
>>> >   http://lists.racket-lang.org/users
>>> >
>>>
>>>
>>>
>>> --
>>> Jay McCarthy
>>> http://jeapostrophe.github.io
>>>
>>>            "Wherefore, be not weary in well-doing,
>>>       for ye are laying the foundation of a great work.
>>> And out of small things proceedeth that which is great."
>>>                           - D&C 64:33
>>
>>
>>
>> ____________________
>>   Racket Users list:
>>   http://lists.racket-lang.org/users
>>
>
>
> End of users Digest, Vol 109, Issue 55
> **************************************


From matthias at ccs.neu.edu  Mon Sep 29 17:15:20 2014
From: matthias at ccs.neu.edu (Matthias Felleisen)
Date: Mon, 29 Sep 2014 17:15:20 -0400
Subject: [racket] Cyclic data structures in Typed Racket
In-Reply-To: <CAAtAoRoBwwg=DDMVKL49SxAJUNLzqRfcB5TSdbpxR0MFeTRODQ@mail.gmail.com>
References: <21545.33909.430329.240145@Konrad-Hinsens-MacBook-Pro-2.local>
 <CAK=HD+Y2x3d1e=caisU9xayyFqRvwm+EHT+Rg66yDCpEAhNGtA@mail.gmail.com>
 <21545.35298.582248.45625@Konrad-Hinsens-MacBook-Pro-2.local>
 <CAK=HD+Ywfnfv-mjQPpo9GmOO8u+aV7f6ue-4motks_GkVnNRaQ@mail.gmail.com>
 <3597E44F-0EF5-4837-AA2D-E06E728B8B71@ccs.neu.edu>
 <CAAtAoRoBwwg=DDMVKL49SxAJUNLzqRfcB5TSdbpxR0MFeTRODQ@mail.gmail.com>
Message-ID: <076C047A-DC0A-49DE-A32A-1C64551DBF9B@ccs.neu.edu>


Why don't you see whether it jives with TR? That would be a nice reproducability study. 


On Sep 29, 2014, at 4:32 PM, Benjamin Greenman <blg59 at cornell.edu> wrote:

> Could Andrew Myer's work on masked types (to eliminate nulls for object initialization) be relevant?
> http://www.cs.cornell.edu/Projects/jmask/
> 
> On Mon, Sep 29, 2014 at 3:30 PM, Matthias Felleisen <matthias at ccs.neu.edu> wrote:
> 
> I guess the real problem is to create suitable default values so
> that the pass the type checker or to figure out a cast that forces
> TR to accept some other value as the default value (and making sure
> you never use the object).
> 
> Undefined not completely solved. Argh. -- Matthias
> 
> 
> 
> 
> 
> On Sep 29, 2014, at 12:52 PM, Sam Tobin-Hochstadt <samth at cs.indiana.edu> wrote:
> 
> > Ah, if you want to create truly cyclic structure with no base case
> > like this, then you'll have to do more work. Either you'll need to
> > expand the type of the `x` field in `foo` to allow an initial value,
> > or you'll need to create a dummy `foo` with that extended type, and
> > then copy the cyclic data once you've created it to a new version of
> > `foo`, using hash tables to track cycles. That's basically how
> > `shared` works, although I haven't tried implementing it in TR and it
> > might not be possible.
> >
> > Sam
> >
> > On Mon, Sep 29, 2014 at 12:33 PM, Konrad Hinsen
> > <konrad.hinsen at fastmail.net> wrote:
> >> Sam Tobin-Hochstadt writes:
> >>
> >>> I recommend doing the mutation yourself, and not using `shared`.
> >>> That's the most obvious solution.
> >>
> >> Not for me, unfortunately, but probably I am missing something
> >> obvious.
> >>
> >> Here's explicit mutation in untyped Racket:
> >>
> >>   #lang racket
> >>
> >>   (struct foo (x) #:mutable)
> >>   (struct bar (x))
> >>
> >>   (define f (foo (void)))
> >>   (define b (bar f))
> >>   (set-foo-x! f b)
> >>
> >>   (eq? (bar-x b) f)
> >>   (eq? (foo-x f) b)
> >>
> >> That works fine. Moving to Typed Racket, this is rejected by the type
> >> checker because (void) is of type Void.  Since I need a bar to make a
> >> foo and a foo to make a bar, I don't see how I can ever initialize my
> >> foo structure.
> >>
> >>> The reason that your `require/typed` doesn't work is that it creates
> >>> new versions of the placeholder functions, but those aren't the ones
> >>> that `shared` uses internally, so Typed Racket doesn't use the types
> >>> you've given.
> >>
> >> That makes sense, thanks for the explanation!
> >>
> >> Konrad.
> > ____________________
> >  Racket Users list:
> >  http://lists.racket-lang.org/users
> 
> 
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
> 



From nickmshelley at gmail.com  Mon Sep 29 17:21:34 2014
From: nickmshelley at gmail.com (Nick Shelley)
Date: Mon, 29 Sep 2014 15:21:34 -0600
Subject: [racket] users Digest, Vol 109, Issue 70
In-Reply-To: <CAKpSqLBEwno4bLuEhXsq552WszgRF-i23h0wQMQEWSB6U5aAqQ@mail.gmail.com>
References: <mailman.478.1411936198.679.users@racket-lang.org>
 <CAKpSqLBEwno4bLuEhXsq552WszgRF-i23h0wQMQEWSB6U5aAqQ@mail.gmail.com>
Message-ID: <CAPrQK3Am_sJYCnPGugiGfLEJs=oeQo5v-kpYCM6wAzPTf-0n4A@mail.gmail.com>

I'm not certain, but I'm pretty sure following the instructions at the top
of the emails you are receiving would be a faster way to get what you want,
rather than spamming the whole list for every email you receive. I'll quote
them for your convenience (copied from the email you are replying to):

To subscribe or unsubscribe via the World Wide Web, visit
         http://lists.racket-lang.org/users/listinfo
 or, via email, send a message with subject or body 'help' to
        users-request at racket-lang.org



On Mon, Sep 29, 2014 at 2:41 PM, Moshe Deutsch <moshedeutsch115 at gmail.com>
wrote:

> Please take me off the list
>
> Thanks
>
> On Sun, Sep 28, 2014 at 4:29 PM,  <users-request at racket-lang.org> wrote:
> > Send users mailing list submissions to
> >         users at racket-lang.org
> >
> > To subscribe or unsubscribe via the World Wide Web, visit
> >         http://lists.racket-lang.org/users/listinfo
> > or, via email, send a message with subject or body 'help' to
> >         users-request at racket-lang.org
> >
> > You can reach the person managing the list at
> >         users-owner at racket-lang.org
> >
> > When replying, please edit your Subject line so it is more specific
> > than "Re: Contents of users digest..."
> >
> >
> > [Racket Users list:
> >  http://lists.racket-lang.org/users ]
> >
> >
> > Today's Topics:
> >
> >    1. Re: Help debugging a ffi crash (Eric Dobson)
> >    2. How to document a field? (Roman Klochkov)
> >    3. Re: proof assistants, DrRacket and Bootstrap (Bill Richter)
> >    4. Re: aws/glacier: credential scope (Greg Hendershott)
> >    5. Re: aws/glacier: credential scope (Norman Gray)
> >    6. Re: aws/glacier: credential scope (Greg Hendershott)
> >
> >
> > ----------------------------------------------------------------------
> >
> > Message: 1
> > Date: Sun, 28 Sep 2014 11:15:33 -0700
> > From: Eric Dobson <eric.n.dobson at gmail.com>
> > To: Matthew Flatt <mflatt at cs.utah.edu>
> > Cc: "users at racket-lang.org" <users at racket-lang.org>
> > Subject: Re: [racket] Help debugging a ffi crash
> > Message-ID:
> >         <
> CAAEHQ5tWAKzi5AizADUH9smVsNXpYTbmxtO3teGm9Zj8ZDUnLg at mail.gmail.com>
> > Content-Type: text/plain; charset=UTF-8
> >
> > That is almost surely it, thanks for the second pair of eyes. I had
> > similar issues with a different type that I thought was a pointer but
> > was actually a struct, but that one I couldn't even get a single call
> > to work so it was much more obvious something was up.
> >
> > On Sun, Sep 28, 2014 at 11:08 AM, Matthew Flatt <mflatt at cs.utah.edu>
> wrote:
> >> Looking at
> >>
> >>   http://clang.llvm.org/doxygen/CXString_8h_source.html
> >>
> >> it seems that CXString as returned by clang_getCursorSpelling() is not
> >> a pointer:
> >>
> >>  typedef struct {
> >>    const void *data;
> >>    unsigned private_flags;
> >>  } CXString;
> >>
> >> If that's right, I'm a little surprised that `cursor-spelling` works
> >> --- but when you get representation wrong, strange things can happen,
> >> including something working when it shouldn't.
> >>
> >> Am I looking at the right library/definitions?
> >>
> >> At Sun, 28 Sep 2014 10:48:06 -0700, Eric Dobson wrote:
> >>> I'm trying to debug an FFI crash that I'm seeing, and because it is
> >>> dealing with C code the error just presents as a segfault. I believe I
> >>> have tracked down what is causing the problem, but don't understand
> >>> how it could be doing so.
> >>>
> >>> I have two racket functions which take a "cursor" (the foreign
> >>> libraries object) and return a string representation of it, which I'm
> >>> trying to use for debugging.
> >>>
> >>> (define raw-clang-get-cstring
> >>>   (get-ffi-obj "clang_getCString" lib-clang
> >>>     (_fun _pointer -> _string)))
> >>>
> >>> (define raw-cursor-spelling
> >>>  (get-ffi-obj "clang_getCursorSpelling" lib-clang
> >>>    (_fun _CXCursor -> _pointer)))
> >>>
> >>> (define (cursor-spelling c)
> >>>   (raw-clang-get-cstring (raw-cursor-spelling c)))
> >>>
> >>> (define cursor-spelling2
> >>>  (get-ffi-obj "clang_getCursorSpelling" lib-clang
> >>>    (_fun _CXCursor -> (make-ctype _pointer values (? (v)
> >>> (raw-clang-get-cstring v))))))
> >>>
> >>> If I use cursor-spelling, I have not been able to trigger a crash. But
> >>> if I use cursor-spelling2 I can reliably trigger a crash.
> >>>
> >>> Is there anything obvious on how these functions are different?
> >>> Because they look to me like they should be doing the same thing. If
> >>> it would be helpful I can try to get my code in a portable enough
> >>> shape so that it will work/crash on another machine.
> >>>
> >>> ____________________
> >>>   Racket Users list:
> >>>   http://lists.racket-lang.org/users
> >
> >
> >
> > ------------------------------
> >
> > Message: 2
> > Date: Sun, 28 Sep 2014 22:39:54 +0400
> > From: Roman Klochkov <kalimehtar at mail.ru>
> > To: racket users list <users at racket-lang.org>
> > Subject: [racket] How to document a field?
> > Message-ID: <1411929594.923492574 at f67.i.mail.ru>
> > Content-Type: text/plain; charset="utf-8"
> >
> >  At  http://docs.racket-lang.org/scribble/doc-classes.html I see
> classes, interfaces, methods, but don't see fields.
> >
> > How to document them? Via @defthing or is there some syntax with
> (get-field ..) (set-field! ...) template like @defparam ?
> >
> >
> > --
> > Roman Klochkov
> > -------------- next part --------------
> > An HTML attachment was scrubbed...
> > URL: <
> http://lists.racket-lang.org/users/archive/attachments/20140928/ac478807/attachment-0001.html
> >
> >
> > ------------------------------
> >
> > Message: 3
> > Date: Sun, 28 Sep 2014 13:51:26 -0500
> > From: Bill Richter <richter at math.northwestern.edu>
> > Cc: users at racket-lang.org
> > Subject: Re: [racket] proof assistants, DrRacket and Bootstrap
> > Message-ID:
> >         <201409281851.s8SIpQMx006764 at poisson.math.northwestern.edu>
> >
> > Matthias & Prabhakar, I think I see what you're saying now.  I need the
> camlp parser to turn my dialect-expressions into trees.  Then I turn my
> tree into standard expressions just as one does in Scheme.  The only
> differences is that in Scheme we have the quote function that skips the
> parsing step.  So your ML code
> >
> > MF>   type ast = VAR of String | LAM of String * ast | APP of ast * ast
> | CONST of int | ADD of ast * ast
> > MF>   type val = BASIC of int | FUNCTION of val -> val | ...
> >
> > looks a lot like the HOL Light definition of preterms:
> >
> > type preterm = Varp of string * pretype       (* Variable           - v
>     *)
> >              | Constp of string * pretype     (* Constant           - c
>     *)
> >              | Combp of preterm * preterm     (* Combination        - f
> x    *)
> >              | Absp of preterm * preterm      (* Lambda-abstraction -
> \x. t  *)
> >              | Typing of preterm * pretype;;  (* Type constraint    - t
> : ty *)
> >
> > and so we can see one of these preterms in a standard term:
> >
> > # preterm_of_term `x + y`;;
> > val it : preterm =
> >   Combp
> >    (Combp
> >      (Constp ("+",
> >        Ptycon ("fun",
> >         [Ptycon ("num", []);
> >          Ptycon ("fun", [Ptycon ("num", []); Ptycon ("num", [])])])),
> >      Varp ("x", Ptycon ("num", []))),
> >    Varp ("y", Ptycon ("num", [])))
> >
> > So the Scheme story works fine except that I haven't understood camlp or
> how HOL Light turns expressions `[...]` into preterms.  But I feel better
> about investing the time into learning camlp now that I see the Scheme
> story remains intact in HOL Light. And you're right  about my ignorance:
> >
> > MF> 1. HtDP explains the above in a couple of sections.
> >
> > Right, there's something I seem not to have understood in either Scheme
> or ML, how when we create and evaluate our trees we retain the values of
> our variables.  I can figure that out, and HtDP sounds like a good place to
> start.
> >
> > PR> A Racketeer would probably tell you to use macros to define your
> > PR> dialect, instead of using an eval hack. The corresponding tool for
> > PR> OCaml is camlp4, and it sounds as if it would be worth your time
> > PR> to learn it thoroughly.
> >
> > That's an interesting comparison, and since macros, maybe I can learn to
> like camlp.  In fact, maybe I ought to learn something about racket macros
> first.
> >
> > --
> > Best,
> > Bill
> >
> >
> > ------------------------------
> >
> > Message: 4
> > Date: Sun, 28 Sep 2014 16:01:20 -0400
> > From: Greg Hendershott <greghendershott at gmail.com>
> > To: Norman Gray <norman at astro.gla.ac.uk>
> > Cc: users at racket-lang.org
> > Subject: Re: [racket] aws/glacier: credential scope
> > Message-ID:
> >         <
> CAGspUn0tqgSVUaDuk1C3Wb6hm6s6nQ2ps-j42pkJJSY2UESL0A at mail.gmail.com>
> > Content-Type: text/plain; charset=UTF-8
> >
> > Hi, Norman.
> >
> > I logged an issue for this:
> >
> >   https://github.com/greghendershott/aws/issues/32
> >
> > I see the problem (or at least the main problem) and will push a fix.
> >
> > The bug is embarrassing, not just because it's such a silly mistake,
> > but it's something a unit test could have caught. (I could say that
> > doing unit tests for Glacier is challenging, because the retrieval
> > process can take hours. Although that's true, I could have done more
> > to test _some_ operations working among various regions.)
> >
> > On Sat, Sep 27, 2014 at 4:14 PM, Norman Gray <norman at astro.gla.ac.uk>
> wrote:
> >>
> >> Greetings.
> >>
> >> I'm trying to use the aws/glacier package, and running into a problem
> where I'm being told:
> >>
> >>     Credential should be scoped to a valid region, not 'eu-west-1'
> >>
> >> I'm following the instructions at <
> https://github.com/greghendershott/aws/blob/master/aws/manual.md>
> >>
> >> My test code is:
> >>
> >>     % cat glacier.rkt
> >>     #lang racket/base
> >>
> >>     (require aws/glacier
> >>              aws/keys)
> >>
> >>     (define vault "testvault")
> >>     (region "eu-west-1")
> >>     (read-keys "aws-zbu-credentials") ; local file
> >>
> >>     (module+ main
> >>       (printf "region=~a~%" (region))
> >>       (printf "Vaults: ~s~%" (list-vaults))
> >>       (printf "...specifically: ~s~%" (describe-vault vault)))
> >>
> >> Running this produces:
> >>
> >>     % racket glacier.rkt
> >>     region=eu-west-1
> >>     aws: HTTP/1.1 403 Forbidden
> >>     x-amzn-RequestId: Un3-L2zlaJBPyrIVKJrWuQcqtMMYQAr34gYUOSScg6Qepc4
> >>     Content-Type: application/json
> >>     Content-Length: 129
> >>     Date: Sat, 27 Sep 2014 18:35:50 GMT
> >>
> >>      {"message":"Credential should be scoped to a valid region, not
> 'eu-west-1'. ","code":"InvalidSignatureException","type":"Client"}
> >>     HTTP 403 "Forbidden". AWS Code="InvalidSignatureException"
> Message="Credential should be scoped to a valid region, not 'eu-west-1'. "
> >>       context...:
> >>        check-response
> >>        /Users/norman/Library/Racket/6.1/pkgs/aws/aws/glacier.rkt:97:22:
> temp68
> >>        request/redirect/uri
> >>        (submod /checkouts/me/code/zbu/glacier.rkt main): [running body]
> >>
> >> Things I thought of:
> >>
> >>   * Printing (public-key)/(private-key) indicates that the credentials
> are being read correctly.
> >>   * When I change the argument of (region) to "us-west-1", that's the
> region that appears in the error message.
> >>   * My "testvault" vault is in eu-west-1 (and this is indeed one of the
> valid regions for glacier, reported in <
> http://docs.aws.amazon.com/general/latest/gr/rande.html> and which does
> have a host at http://glacier.eu-west-1.amazonaws.com
> >>   * As far as I can see, credentials are _not_ scoped, but are all at
> us-east-1.
> >>   * <
> http://docs.aws.amazon.com/general/latest/gr/signature-v4-troubleshooting.html>
> says that "IAM [...] accepts only us-east-1 as its region specification",
> so I'm taking it that (region) is for setting the _vault_'s region.
> >>   * I'm not a great AWS expert, so I could have something in my setup
> broken; but if so, I've no clue what.
> >>
> >> If, however, I change the (region) argument to "us-east-1", I get a
> different error message "User: arn:aws:iam::786725553169:user/zbu is not
> authorized to perform: glacier:ListVaults on resource:
> arn:aws:glacier:us-east-1:786725553169:vaults/"  That makes sense, since
> there's no such vault, but it's interesting that it gets _further_ when the
> (region) matches the region for the IAM service.
> >>
> >> I don't see any other (region) equivalents for the other services
> supported by the package.  Is that because all of the other services
> supported by the package are supported by all the AWS regions, or am I
> missing a configuration?
> >>
> >> Thanks for any pointers.
> >>
> >> All the best,
> >>
> >> Norman
> >>
> >>
> >> --
> >> Norman Gray  :  http://nxg.me.uk
> >> SUPA School of Physics and Astronomy, University of Glasgow, UK
> >>
> >>
> >> ____________________
> >>   Racket Users list:
> >>   http://lists.racket-lang.org/users
> >
> >
> >
> > ------------------------------
> >
> > Message: 5
> > Date: Sun, 28 Sep 2014 21:29:03 +0100
> > From: Norman Gray <norman at astro.gla.ac.uk>
> > To: Greg Hendershott <greghendershott at gmail.com>
> > Cc: users at racket-lang.org
> > Subject: Re: [racket] aws/glacier: credential scope
> > Message-ID: <FE320180-07AE-43E8-AE51-3AA566329E67 at astro.gla.ac.uk>
> > Content-Type: text/plain; charset=us-ascii
> >
> >
> > Greg, hello.
> >
> > On 2014 Sep 28, at 21:01, Greg Hendershott <greghendershott at gmail.com>
> wrote:
> >
> >> I logged an issue for this:
> >>
> >>  https://github.com/greghendershott/aws/issues/32
> >
> > Ah: I wondered if the problem might be related to that.
> >
> >> The bug is embarrassing, not just because it's such a silly mistake,
> >> but it's something a unit test could have caught. (I could say that
> >> doing unit tests for Glacier is challenging, because the retrieval
> >> process can take hours.
> >
> > I can imagine how protracted that would be ... *shudder*.
> >
> > Thanks for looking at this.
> >
> > Also, Frank:
> >
> >> What happens if you read-keys before setting the region?
> >
> > Thanks for this suggestion: I did try that just a little earlier, but it
> didn't make any difference.
> >
> > All the best,
> >
> > Norman
> >
> >
> > --
> > Norman Gray  :  http://nxg.me.uk
> > SUPA School of Physics and Astronomy, University of Glasgow, UK
> >
> >
> >
> >
> > ------------------------------
> >
> > Message: 6
> > Date: Sun, 28 Sep 2014 16:29:12 -0400
> > From: Greg Hendershott <greghendershott at gmail.com>
> > To: Norman Gray <norman at astro.gla.ac.uk>
> > Cc: users at racket-lang.org
> > Subject: Re: [racket] aws/glacier: credential scope
> > Message-ID:
> >         <
> CAGspUn2au5xydsH+o9E2Y3JTA1iQbPRX58zRcitywAPWh5Djng at mail.gmail.com>
> > Content-Type: text/plain; charset=UTF-8
> >
> > I pushed a fix, including running the tests across a few different AWS
> > regions. They pass.
> >
> > I clicked "update" on pkgs.racket-lang.org. But it seems to be slower
> > than usual to refresh. After it does, you can `raco pkg update aws` to
> > get the fix.
> >
> > On Sun, Sep 28, 2014 at 4:01 PM, Greg Hendershott
> > <greghendershott at gmail.com> wrote:
> >> Hi, Norman.
> >>
> >> I logged an issue for this:
> >>
> >>   https://github.com/greghendershott/aws/issues/32
> >>
> >> I see the problem (or at least the main problem) and will push a fix.
> >>
> >> The bug is embarrassing, not just because it's such a silly mistake,
> >> but it's something a unit test could have caught. (I could say that
> >> doing unit tests for Glacier is challenging, because the retrieval
> >> process can take hours. Although that's true, I could have done more
> >> to test _some_ operations working among various regions.)
> >>
> >> On Sat, Sep 27, 2014 at 4:14 PM, Norman Gray <norman at astro.gla.ac.uk>
> wrote:
> >>>
> >>> Greetings.
> >>>
> >>> I'm trying to use the aws/glacier package, and running into a problem
> where I'm being told:
> >>>
> >>>     Credential should be scoped to a valid region, not 'eu-west-1'
> >>>
> >>> I'm following the instructions at <
> https://github.com/greghendershott/aws/blob/master/aws/manual.md>
> >>>
> >>> My test code is:
> >>>
> >>>     % cat glacier.rkt
> >>>     #lang racket/base
> >>>
> >>>     (require aws/glacier
> >>>              aws/keys)
> >>>
> >>>     (define vault "testvault")
> >>>     (region "eu-west-1")
> >>>     (read-keys "aws-zbu-credentials") ; local file
> >>>
> >>>     (module+ main
> >>>       (printf "region=~a~%" (region))
> >>>       (printf "Vaults: ~s~%" (list-vaults))
> >>>       (printf "...specifically: ~s~%" (describe-vault vault)))
> >>>
> >>> Running this produces:
> >>>
> >>>     % racket glacier.rkt
> >>>     region=eu-west-1
> >>>     aws: HTTP/1.1 403 Forbidden
> >>>     x-amzn-RequestId: Un3-L2zlaJBPyrIVKJrWuQcqtMMYQAr34gYUOSScg6Qepc4
> >>>     Content-Type: application/json
> >>>     Content-Length: 129
> >>>     Date: Sat, 27 Sep 2014 18:35:50 GMT
> >>>
> >>>      {"message":"Credential should be scoped to a valid region, not
> 'eu-west-1'. ","code":"InvalidSignatureException","type":"Client"}
> >>>     HTTP 403 "Forbidden". AWS Code="InvalidSignatureException"
> Message="Credential should be scoped to a valid region, not 'eu-west-1'. "
> >>>       context...:
> >>>        check-response
> >>>
> /Users/norman/Library/Racket/6.1/pkgs/aws/aws/glacier.rkt:97:22: temp68
> >>>        request/redirect/uri
> >>>        (submod /checkouts/me/code/zbu/glacier.rkt main): [running body]
> >>>
> >>> Things I thought of:
> >>>
> >>>   * Printing (public-key)/(private-key) indicates that the credentials
> are being read correctly.
> >>>   * When I change the argument of (region) to "us-west-1", that's the
> region that appears in the error message.
> >>>   * My "testvault" vault is in eu-west-1 (and this is indeed one of
> the valid regions for glacier, reported in <
> http://docs.aws.amazon.com/general/latest/gr/rande.html> and which does
> have a host at http://glacier.eu-west-1.amazonaws.com
> >>>   * As far as I can see, credentials are _not_ scoped, but are all at
> us-east-1.
> >>>   * <
> http://docs.aws.amazon.com/general/latest/gr/signature-v4-troubleshooting.html>
> says that "IAM [...] accepts only us-east-1 as its region specification",
> so I'm taking it that (region) is for setting the _vault_'s region.
> >>>   * I'm not a great AWS expert, so I could have something in my setup
> broken; but if so, I've no clue what.
> >>>
> >>> If, however, I change the (region) argument to "us-east-1", I get a
> different error message "User: arn:aws:iam::786725553169:user/zbu is not
> authorized to perform: glacier:ListVaults on resource:
> arn:aws:glacier:us-east-1:786725553169:vaults/"  That makes sense, since
> there's no such vault, but it's interesting that it gets _further_ when the
> (region) matches the region for the IAM service.
> >>>
> >>> I don't see any other (region) equivalents for the other services
> supported by the package.  Is that because all of the other services
> supported by the package are supported by all the AWS regions, or am I
> missing a configuration?
> >>>
> >>> Thanks for any pointers.
> >>>
> >>> All the best,
> >>>
> >>> Norman
> >>>
> >>>
> >>> --
> >>> Norman Gray  :  http://nxg.me.uk
> >>> SUPA School of Physics and Astronomy, University of Glasgow, UK
> >>>
> >>>
> >>> ____________________
> >>>   Racket Users list:
> >>>   http://lists.racket-lang.org/users
> >
> >
> >
> > End of users Digest, Vol 109, Issue 70
> > **************************************
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140929/9c3a8818/attachment-0001.html>

From neil at neilvandyke.org  Mon Sep 29 17:34:55 2014
From: neil at neilvandyke.org (Neil Van Dyke)
Date: Mon, 29 Sep 2014 17:34:55 -0400
Subject: [racket] users Digest, Vol 109, Issue 79
In-Reply-To: <CAKpSqLDBQh=92z9jovMmVgDb5fbE1FuD0Bt2o_5g29-9-bM6Rg@mail.gmail.com>
References: <mailman.496.1412022767.679.users@racket-lang.org>
 <CAKpSqLDBQh=92z9jovMmVgDb5fbE1FuD0Bt2o_5g29-9-bM6Rg@mail.gmail.com>
Message-ID: <5429D07F.1070408@neilvandyke.org>

I had privately pointed him to the URL to unsubscribe himself, but he 
says it didn't work.  It appears that tensions have escalated. :)

Neil V.


From nickmshelley at gmail.com  Mon Sep 29 18:32:27 2014
From: nickmshelley at gmail.com (Nick Shelley)
Date: Mon, 29 Sep 2014 16:32:27 -0600
Subject: [racket] users Digest, Vol 109, Issue 79
In-Reply-To: <5429D07F.1070408@neilvandyke.org>
References: <mailman.496.1412022767.679.users@racket-lang.org>
 <CAKpSqLDBQh=92z9jovMmVgDb5fbE1FuD0Bt2o_5g29-9-bM6Rg@mail.gmail.com>
 <5429D07F.1070408@neilvandyke.org>
Message-ID: <CAPrQK3BApxuvqfGMqDS23+fFoFB9=-it=a+epJhcMi732DK61A@mail.gmail.com>

I did so privately as well, but he said he doesn't have his password I
guess. I replied to that and pointed him to the users-request email option.
Hopefully that will help.

On Mon, Sep 29, 2014 at 3:34 PM, Neil Van Dyke <neil at neilvandyke.org> wrote:

> I had privately pointed him to the URL to unsubscribe himself, but he says
> it didn't work.  It appears that tensions have escalated. :)
>
> Neil V.
>
>
> ____________________
>  Racket Users list:
>  http://lists.racket-lang.org/users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140929/544003ef/attachment.html>

From eli at barzilay.org  Mon Sep 29 19:14:28 2014
From: eli at barzilay.org (Eli Barzilay)
Date: Mon, 29 Sep 2014 19:14:28 -0400
Subject: [racket] users Digest, Vol 109, Issue 79
In-Reply-To: <CAPrQK3BApxuvqfGMqDS23+fFoFB9=-it=a+epJhcMi732DK61A@mail.gmail.com>
References: <mailman.496.1412022767.679.users@racket-lang.org>
 <CAKpSqLDBQh=92z9jovMmVgDb5fbE1FuD0Bt2o_5g29-9-bM6Rg@mail.gmail.com>
 <5429D07F.1070408@neilvandyke.org>
 <CAPrQK3BApxuvqfGMqDS23+fFoFB9=-it=a+epJhcMi732DK61A@mail.gmail.com>
Message-ID: <CALO-gutvnQV1U=4E_6kyWw3-84dCooiUV6pNvcX89x4MhvqWCA@mail.gmail.com>

On Mon, Sep 29, 2014 at 3:34 PM, Neil Van Dyke <neil at neilvandyke.org> wrote:
> I had privately pointed him to the URL to unsubscribe himself, but he
> says it didn't work.  It appears that tensions have escalated. :)

He eventually did unsubscribe.


On Mon, Sep 29, 2014 at 6:32 PM, Nick Shelley <nickmshelley at gmail.com> wrote:
> I did so privately as well, but he said he doesn't have his password I
> guess. I replied to that and pointed him to the users-request email
> option.  Hopefully that will help.

Note that every email has, at the bottom (or at the top for digest
emails), that link that lets you unsubscribe (and change other options).
There is no need for a password: if you don't have it just use the
unsubscribe button without it, the mailing list will send you an email
with a link to get you off.

(Plus, there's a button under that which will send you the password.
Overall, Mailman is doing its best to not require your password, and
make it easy to get it if you do need it.)

-- 
          ((lambda (x) (x x)) (lambda (x) (x x)))          Eli Barzilay:
                    http://barzilay.org/                   Maze is Life!

From mflatt at cs.utah.edu  Mon Sep 29 19:15:05 2014
From: mflatt at cs.utah.edu (Matthew Flatt)
Date: Mon, 29 Sep 2014 17:15:05 -0600
Subject: [racket] users Digest, Vol 109, Issue 79
In-Reply-To: <CAPrQK3BApxuvqfGMqDS23+fFoFB9=-it=a+epJhcMi732DK61A@mail.gmail.com>
References: <mailman.496.1412022767.679.users@racket-lang.org>
 <CAKpSqLDBQh=92z9jovMmVgDb5fbE1FuD0Bt2o_5g29-9-bM6Rg@mail.gmail.com>
 <5429D07F.1070408@neilvandyke.org>
 <CAPrQK3BApxuvqfGMqDS23+fFoFB9=-it=a+epJhcMi732DK61A@mail.gmail.com>
Message-ID: <20140929231507.22A69650197@mail-svr1.cs.utah.edu>

I unsubscribed him manually after the first few messages reached me (I
had been away from e-mail for a while), and I think the rest were in
the out queue already. Thanks for your help!

At Mon, 29 Sep 2014 16:32:27 -0600, Nick Shelley wrote:
> I did so privately as well, but he said he doesn't have his password I
> guess. I replied to that and pointed him to the users-request email option.
> Hopefully that will help.
> 
> On Mon, Sep 29, 2014 at 3:34 PM, Neil Van Dyke <neil at neilvandyke.org> wrote:
> 
> > I had privately pointed him to the URL to unsubscribe himself, but he says
> > it didn't work.  It appears that tensions have escalated. :)
> >
> > Neil V.
> >
> >
> > ____________________
> >  Racket Users list:
> >  http://lists.racket-lang.org/users
> >
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users

From emanonhere at gmail.com  Mon Sep 29 21:52:17 2014
From: emanonhere at gmail.com (rrandom)
Date: Tue, 30 Sep 2014 09:52:17 +0800
Subject: [racket] Is there any way to get the bit depth of a bitmap?
In-Reply-To: <20140929152408.F0E42650192@mail-svr1.cs.utah.edu>
References: <54297763.3020407@gmx.com>
 <20140929152408.F0E42650192@mail-svr1.cs.utah.edu>
Message-ID: <542A0CD1.6080507@gmail.com>

sorry, I do not understand, Does that mean I must set the bit depth of a 
bitmap manually?


? 2014/9/29 23:24, Matthew Flatt ??:
> Racket libraries can read and write images of various formats and
> pixels depths, but a `bitmap%` object always represents an image using
> 1 or 32 bits per pixel in memory.
>
> At Mon, 29 Sep 2014 23:14:43 +0800, rrandom wrote:
>> Hi,I am recently working in the book Digital Image Processing with
>> racket. I want to know is there any way to get the bit depth or color
>> depth of a input bitmap in Racket? I found the get-depth method in
>> racket/draw,bitmap%, but it only return 1 and 32, that's not what i
>> really want.I input a bitmap whose bit depth is 8 but i got 32.
>> following is the code:
>>
>>> (send (read-bitmap "Fig0305(a)(DFT_no_log).tif")
>> get-depth)
>> 32
>>
>> Is there anything I am wrong? Thank you!
>> ____________________
>>    Racket Users list:
>>    http://lists.racket-lang.org/users


From alex.mclin at gmail.com  Mon Sep 29 22:44:36 2014
From: alex.mclin at gmail.com (Alexander McLin)
Date: Mon, 29 Sep 2014 22:44:36 -0400
Subject: [racket] users Digest, Vol 109, Issue 55
In-Reply-To: <CAKpSqLDjXF=02wWUG=vD5OojAr2yYjyihBLfPtC1GkEnOQAMRw@mail.gmail.com>
References: <mailman.440.1411649827.679.users@racket-lang.org>
 <CAKpSqLDjXF=02wWUG=vD5OojAr2yYjyihBLfPtC1GkEnOQAMRw@mail.gmail.com>
Message-ID: <CAKeBiN5-N4HZ5yMCutuN1hNyJQB4GpXL0iFVtY_YxBY-B3MQBA@mail.gmail.com>

Good grief, was it necessary to spam the list with your requests? You could
have just sent one email instead of replying to like 20 different threads?

And it would have been nice if you'd read the basic instructions which
tells you exactly how to unsubscribe yourself without flooding my inbox.

On Mon, Sep 29, 2014 at 4:44 PM, Moshe Deutsch <moshedeutsch115 at gmail.com>
wrote:

> Please take me off the list
>
> Thanks
>
> On Thu, Sep 25, 2014 at 8:57 AM,  <users-request at racket-lang.org> wrote:
> > Send users mailing list submissions to
> >         users at racket-lang.org
> >
> > To subscribe or unsubscribe via the World Wide Web, visit
> >         http://lists.racket-lang.org/users/listinfo
> > or, via email, send a message with subject or body 'help' to
> >         users-request at racket-lang.org
> >
> > You can reach the person managing the list at
> >         users-owner at racket-lang.org
> >
> > When replying, please edit your Subject line so it is more specific
> > than "Re: Contents of users digest..."
> >
> >
> > [Racket Users list:
> >  http://lists.racket-lang.org/users ]
> >
> >
> > Today's Topics:
> >
> >    1. macro question (Alejandro Zamora)
> >    2. Re: ranking items by their appearance in sets (David T. Pierson)
> >    3. Switching from editor to repl (C K Kashyap)
> >    4. Re: Switching from editor to repl (Jay McCarthy)
> >    5. Re: Switching from editor to repl (C K Kashyap)
> >    6. Re: Switching from editor to repl (Robby Findler)
> >
> >
> > ----------------------------------------------------------------------
> >
> > Message: 1
> > Date: Wed, 24 Sep 2014 22:35:49 -0400
> > From: Alejandro Zamora <terefv at ltu.sld.cu>
> > To: Racket Users <users at racket-lang.org>
> > Subject: [racket] macro question
> > Message-ID: <20140924223549.7f15c912 at alejandro-H61H2-CM>
> > Content-Type: text/plain; charset=US-ASCII
> >
> > Hi Racket people:
> >         I have a (maybe too simple) question:
> >         How I create one macro like this (but using define-macro &
> >         R5RS language only)?:
> >
> >         (vars sym var-cant val-of-vars)
> >
> >         Example:
> >         (vars id 5 null) -> (prog (define id1 null)
> >                                   (define id2 null)
> >                                   (define id3 null)
> >                                   (define id4 null)
> >                                   (define id5 null))
> >
> > --
> > Nunca digas nunca, di mejor: gracias, permiso, disculpe.
> >
> > Este mensaje le ha llegado mediante el servicio de correo electronico
> que ofrece Infomed para respaldar el cumplimiento de las misiones del
> Sistema Nacional de Salud. La persona que envia este correo asume el
> compromiso de usar el servicio a tales fines y cumplir con las regulaciones
> establecidas
> >
> > Infomed: http://www.sld.cu/
> >
> >
> >
> > ------------------------------
> >
> > Message: 2
> > Date: Wed, 24 Sep 2014 23:07:13 -0400
> > From: "David T. Pierson" <dtp at mindstory.com>
> > To: Matthias Felleisen <matthias at ccs.neu.edu>
> > Cc: users at racket-lang.org
> > Subject: Re: [racket] ranking items by their appearance in sets
> > Message-ID: <1411613699.19584%dtp at mindstory.com>
> > Content-Type: text/plain; charset=us-ascii
> >
> > On Wed, Sep 24, 2014 at 10:44:45AM -0400, Matthias Felleisen wrote:
> >> Style: Here is my rewrite with style suggestions. I liked to have a
> >> single point that tests all versions of a function when I conduct such
> >> design experiments. I also want to have all my test code in the test
> >> submodule so that it doesn't get deployed. That's why I am using a
> >> syntactic abstraction that unfolds into a submodule rather than a
> >> function that calls things.
> >
> > Thanks for the style suggestions.  I am reminded I need to reread the
> > style guide.
> >
> >> Performance: I have not checked their performance but I cannot imagine
> >> the first version to be performant. See my annotations.
> >
> > Good point.  I had put performance out of my mind because my real data
> > is relatively small and I just wanted to play a bit with the methods,
> > but I definitely appreciate the notes.
> >
> >> My choice: Your second version is what I would have written.
> >
> > Glad to know I was on the right track.
> >
> > Thank you.
> >
> > David
> >
> >
> > ------------------------------
> >
> > Message: 3
> > Date: Thu, 25 Sep 2014 11:47:57 +0530
> > From: C K Kashyap <ckkashyap at gmail.com>
> > To: users at racket-lang.org
> > Subject: [racket] Switching from editor to repl
> > Message-ID:
> >         <
> CAGdT1gqUp-0hdF15FYo2T_Qpm+qf9AkKMak_iLV9HTC4WkDrtQ at mail.gmail.com>
> > Content-Type: text/plain; charset="utf-8"
> >
> > Hi,
> > I am new to racket (but am really excited by the idea - everything is a
> > program) - need a quick help with Dr Racket -
> >
> > What's the keyboard shortcut to switch between editor and repl?
> >
> > I got to know from a video on youtube that I could use C-e to toggle the
> > repl visibility but not able to find out how to get to the rpl using the
> > keyboard.
> > Regards,
> > Kashyap
> > -------------- next part --------------
> > An HTML attachment was scrubbed...
> > URL: <
> http://lists.racket-lang.org/users/archive/attachments/20140925/39d4382b/attachment-0001.html
> >
> >
> > ------------------------------
> >
> > Message: 4
> > Date: Thu, 25 Sep 2014 02:33:55 -0400
> > From: Jay McCarthy <jay.mccarthy at gmail.com>
> > To: C K Kashyap <ckkashyap at gmail.com>
> > Cc: users <users at racket-lang.org>
> > Subject: Re: [racket] Switching from editor to repl
> > Message-ID:
> >         <
> CAJYbDamt+_RvG1utwzjJ1KW-m5ynRDxPr2xVivTOGQj8Y2R3DA at mail.gmail.com>
> > Content-Type: text/plain; charset=UTF-8
> >
> > I use C-e and C-d to do this and I just hit it twice, once hides the
> > side and the second shows it and focuses it.
> >
> > On Thu, Sep 25, 2014 at 2:17 AM, C K Kashyap <ckkashyap at gmail.com>
> wrote:
> >> Hi,
> >> I am new to racket (but am really excited by the idea - everything is a
> >> program) - need a quick help with Dr Racket -
> >>
> >> What's the keyboard shortcut to switch between editor and repl?
> >>
> >> I got to know from a video on youtube that I could use C-e to toggle the
> >> repl visibility but not able to find out how to get to the rpl using the
> >> keyboard.
> >> Regards,
> >> Kashyap
> >>
> >> ____________________
> >>   Racket Users list:
> >>   http://lists.racket-lang.org/users
> >>
> >
> >
> >
> > --
> > Jay McCarthy
> > http://jeapostrophe.github.io
> >
> >            "Wherefore, be not weary in well-doing,
> >       for ye are laying the foundation of a great work.
> > And out of small things proceedeth that which is great."
> >                           - D&C 64:33
> >
> >
> > ------------------------------
> >
> > Message: 5
> > Date: Thu, 25 Sep 2014 13:13:58 +0530
> > From: C K Kashyap <ckkashyap at gmail.com>
> > To: Jay McCarthy <jay.mccarthy at gmail.com>
> > Cc: users <users at racket-lang.org>
> > Subject: Re: [racket] Switching from editor to repl
> > Message-ID:
> >         <CAGdT1gqYYRz1Kk11U1==oUgP93ELFR0PFPsdG5v8C_tYjwqH=
> g at mail.gmail.com>
> > Content-Type: text/plain; charset="utf-8"
> >
> > C-d works for me ... thanks Jay.
> > I was wondering though if I could've found this out in the documentation
> > somewhere. I could not find it in the guide/google.
> >
> > Regards,
> > Kashyap
> >
> > On Thu, Sep 25, 2014 at 12:03 PM, Jay McCarthy <jay.mccarthy at gmail.com>
> > wrote:
> >
> >> I use C-e and C-d to do this and I just hit it twice, once hides the
> >> side and the second shows it and focuses it.
> >>
> >> On Thu, Sep 25, 2014 at 2:17 AM, C K Kashyap <ckkashyap at gmail.com>
> wrote:
> >> > Hi,
> >> > I am new to racket (but am really excited by the idea - everything is
> a
> >> > program) - need a quick help with Dr Racket -
> >> >
> >> > What's the keyboard shortcut to switch between editor and repl?
> >> >
> >> > I got to know from a video on youtube that I could use C-e to toggle
> the
> >> > repl visibility but not able to find out how to get to the rpl using
> the
> >> > keyboard.
> >> > Regards,
> >> > Kashyap
> >> >
> >> > ____________________
> >> >   Racket Users list:
> >> >   http://lists.racket-lang.org/users
> >> >
> >>
> >>
> >>
> >> --
> >> Jay McCarthy
> >> http://jeapostrophe.github.io
> >>
> >>            "Wherefore, be not weary in well-doing,
> >>       for ye are laying the foundation of a great work.
> >> And out of small things proceedeth that which is great."
> >>                           - D&C 64:33
> >>
> > -------------- next part --------------
> > An HTML attachment was scrubbed...
> > URL: <
> http://lists.racket-lang.org/users/archive/attachments/20140925/93aa6ef8/attachment-0001.html
> >
> >
> > ------------------------------
> >
> > Message: 6
> > Date: Thu, 25 Sep 2014 07:56:53 -0500
> > From: Robby Findler <robby at eecs.northwestern.edu>
> > To: C K Kashyap <ckkashyap at gmail.com>
> > Cc: Jay McCarthy <jay.mccarthy at gmail.com>, users
> >         <users at racket-lang.org>
> > Subject: Re: [racket] Switching from editor to repl
> > Message-ID:
> >         <CAL3TdONC6=
> U8EibFfxLYDNVVW0+SEo9+Do5ZuMSu2bLQRm8y8A at mail.gmail.com>
> > Content-Type: text/plain; charset=UTF-8
> >
> > The guide focuses on the language, not the DrRacket IDE. This is a
> > link to the IDE docs that has a list of keyboard shortcuts.
> >
> > You might also try looking at the menu items; things are a little more
> > discoverable that way too.
> >
> > http://docs.racket-lang.org/drracket/Keyboard_Shortcuts.html
> >
> > hth,
> > Robby
> >
> > On Thu, Sep 25, 2014 at 2:43 AM, C K Kashyap <ckkashyap at gmail.com>
> wrote:
> >> C-d works for me ... thanks Jay.
> >> I was wondering though if I could've found this out in the documentation
> >> somewhere. I could not find it in the guide/google.
> >>
> >> Regards,
> >> Kashyap
> >>
> >> On Thu, Sep 25, 2014 at 12:03 PM, Jay McCarthy <jay.mccarthy at gmail.com>
> >> wrote:
> >>>
> >>> I use C-e and C-d to do this and I just hit it twice, once hides the
> >>> side and the second shows it and focuses it.
> >>>
> >>> On Thu, Sep 25, 2014 at 2:17 AM, C K Kashyap <ckkashyap at gmail.com>
> wrote:
> >>> > Hi,
> >>> > I am new to racket (but am really excited by the idea - everything
> is a
> >>> > program) - need a quick help with Dr Racket -
> >>> >
> >>> > What's the keyboard shortcut to switch between editor and repl?
> >>> >
> >>> > I got to know from a video on youtube that I could use C-e to toggle
> the
> >>> > repl visibility but not able to find out how to get to the rpl using
> the
> >>> > keyboard.
> >>> > Regards,
> >>> > Kashyap
> >>> >
> >>> > ____________________
> >>> >   Racket Users list:
> >>> >   http://lists.racket-lang.org/users
> >>> >
> >>>
> >>>
> >>>
> >>> --
> >>> Jay McCarthy
> >>> http://jeapostrophe.github.io
> >>>
> >>>            "Wherefore, be not weary in well-doing,
> >>>       for ye are laying the foundation of a great work.
> >>> And out of small things proceedeth that which is great."
> >>>                           - D&C 64:33
> >>
> >>
> >>
> >> ____________________
> >>   Racket Users list:
> >>   http://lists.racket-lang.org/users
> >>
> >
> >
> > End of users Digest, Vol 109, Issue 55
> > **************************************
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140929/2b185982/attachment-0001.html>

From konrad.hinsen at fastmail.net  Tue Sep 30 03:32:58 2014
From: konrad.hinsen at fastmail.net (Konrad Hinsen)
Date: Tue, 30 Sep 2014 09:32:58 +0200
Subject: [racket] Cyclic data structures in Typed Racket
In-Reply-To: <3597E44F-0EF5-4837-AA2D-E06E728B8B71@ccs.neu.edu>
References: <21545.33909.430329.240145@Konrad-Hinsens-MacBook-Pro-2.local>
 <CAK=HD+Y2x3d1e=caisU9xayyFqRvwm+EHT+Rg66yDCpEAhNGtA@mail.gmail.com>
 <21545.35298.582248.45625@Konrad-Hinsens-MacBook-Pro-2.local>
 <CAK=HD+Ywfnfv-mjQPpo9GmOO8u+aV7f6ue-4motks_GkVnNRaQ@mail.gmail.com>
 <3597E44F-0EF5-4837-AA2D-E06E728B8B71@ccs.neu.edu>
Message-ID: <21546.23722.89229.703426@Ordinateur-de-Catherine-Konrad.local>

Matthias Felleisen writes:

 > I guess the real problem is to create suitable default values so 
 > that the pass the type checker or to figure out a cast that forces
 > TR to accept some other value as the default value (and making sure
 > you never use the object). 
 > 
 > Undefined not completely solved. Argh. -- Matthias

A related issue is the requirement to make data structures mutable
in order to allow cyclic data. That's a lot of risk to take for
a one-time use.

Could single-assignment cells be part of a solution to both issues?
Instead of mutable cells in a struct (or list ...) , one would have
single-assignment cells. They are initialized to undefined, and
assignment is only allowed from undefined to some other value, which
means it can happen only once.

A special code block (Racket's share) would allow the creation and
initialization and guarantee that everything is fully initialized
in the end, meaning that the type checker doesn't have to worry
about possible undefined values in the rest of the code.

A variant on this idea is Clojure's transients: data structures that
are created mutable, but are locked explicitly after initialization to
become immutable. That's much more general, but probably also much
harder to trace by a type checker.

Konrad.

From konrad.hinsen at fastmail.net  Tue Sep 30 03:38:14 2014
From: konrad.hinsen at fastmail.net (Konrad Hinsen)
Date: Tue, 30 Sep 2014 09:38:14 +0200
Subject: [racket] typed racket, filters, and polymorphism
In-Reply-To: <CAL3TdOPrHreseE2RxJW6x5jm5kSnKVQkjvAt-A8RuiKaarCZZw@mail.gmail.com>
References: <1FB59E1F-0D4F-4276-BAB7-C6328866CE70@knauth.org>
 <1411920801296.55e3aa5@Nodemailer>
 <B849BCC0-7F5C-402E-B98C-01F97E1762D8@knauth.org>
 <2B2B37BB-377F-4D28-8E69-EFEFCF8181A2@ccs.neu.edu>
 <B52D6D46-C9C1-46D6-9F1E-46800DB7B23C@knauth.org>
 <C3486009-3F00-4717-BD42-E18FF074FD4B@ccs.neu.edu>
 <21545.7176.252047.613404@Ordinateur-de-Catherine-Konrad.local>
 <CAL3TdONRKwUJ0pV3WhXDQvRgMT6mhSp6D6QGr4=W4y7g7DsK4w@mail.gmail.com>
 <CAL3TdOPrHreseE2RxJW6x5jm5kSnKVQkjvAt-A8RuiKaarCZZw@mail.gmail.com>
Message-ID: <21546.24038.422390.24048@Ordinateur-de-Catherine-Konrad.local>

Robby Findler writes:

 > Whoops, sorry: this is the same thing Konrad was talking about! Duh.
 > Well, hopefully the link is useful.

It is, thanks! There's a lot of material explaining the background of
F#'s unit checker.

The problem is highly relevant in science and engineering, where
dimensional analysis is the equivalent of type checking in
programming, with the added benefit that no one questions its utility
or importance.

There is also some interesting material on the Haskell Wiki:

   http://www.haskell.org/haskellwiki/Physical_units

Unfortunately it is hard to see what exactly is going on there, since
every package on that list requires some arcane GHC extensions whose
precise impact on the type system is known only to Haskell insiders.

Konrad.

From ckkashyap at gmail.com  Tue Sep 30 03:53:55 2014
From: ckkashyap at gmail.com (C K Kashyap)
Date: Tue, 30 Sep 2014 13:23:55 +0530
Subject: [racket] rsound docs
In-Reply-To: <20140929152815.C87E6650194@mail-svr1.cs.utah.edu>
References: <CAGdT1gq=E_-O9JcnaMnOEQkEK51vWxdPsFHHNSudzRoJCwHm2g@mail.gmail.com>
 <20140929152815.C87E6650194@mail-svr1.cs.utah.edu>
Message-ID: <CAGdT1gpx=2UVJTzThe1RCfuPfj83=DKtNXdLNeGRw_NuS60RCQ@mail.gmail.com>

Thank you so much Matthew!

Regards,
Kashyap

On Mon, Sep 29, 2014 at 8:58 PM, Matthew Flatt <mflatt at cs.utah.edu> wrote:

> When you install the package, you also install documentation for
> RSound. You should be able to find it by selecting "Racket
> Documentation" from the "Help" menu in DrRacket, by right-clicking on
> an identifier in DrRacket that is provided by the package's library, or
> by running `raco docs` on the command line.
>
> You can also find documentation linked from the "rsound" entry at
> http://pkgs.racket-lang.org/, which goes to
>
>   http://pkg-build.racket-lang.org/doc/rsound/index.html
>
>
> At Mon, 29 Sep 2014 19:09:17 +0530, C K Kashyap wrote:
> > Hi,
> >
> > I saw this nice racketcon video (
> >
> https://www.youtube.com/watch?v=DkIVzHNjNEA&list=PLXr4KViVC0qI9t3lizitiFJ1cFIeN2
> > Gdh&index=9)
> > on rsound. I've installed the rsound package and have been able to run
> the
> > example from the video.
> >
> > Is there a place that has mode documentation/tutorial on rsound?
> >
> > Regards,
> > Kashyap
> > ____________________
> >   Racket Users list:
> >   http://lists.racket-lang.org/users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140930/8d244da3/attachment.html>

From konrad.hinsen at fastmail.net  Tue Sep 30 07:12:19 2014
From: konrad.hinsen at fastmail.net (Konrad Hinsen)
Date: Tue, 30 Sep 2014 13:12:19 +0200
Subject: [racket] Cyclic data structures in Typed Racket
In-Reply-To: <CAK=HD+Ywfnfv-mjQPpo9GmOO8u+aV7f6ue-4motks_GkVnNRaQ@mail.gmail.com>
References: <21545.33909.430329.240145@Konrad-Hinsens-MacBook-Pro-2.local>
 <CAK=HD+Y2x3d1e=caisU9xayyFqRvwm+EHT+Rg66yDCpEAhNGtA@mail.gmail.com>
 <21545.35298.582248.45625@Konrad-Hinsens-MacBook-Pro-2.local>
 <CAK=HD+Ywfnfv-mjQPpo9GmOO8u+aV7f6ue-4motks_GkVnNRaQ@mail.gmail.com>
Message-ID: <21546.36883.42112.592373@Konrad-Hinsens-MacBook-Pro-2.local>

Sam Tobin-Hochstadt writes:

 > Ah, if you want to create truly cyclic structure with no base case
 > like this, then you'll have to do more work. Either you'll need to
 > expand the type of the `x` field in `foo` to allow an initial value,
 > or you'll need to create a dummy `foo` with that extended type, and
 > then copy the cyclic data once you've created it to a new version

In my specific application, I ended up replacing the cyclic data
structure, which was a singleton, by a function that simulates it.
But for another project, which involves graphs, I'll stick to
untyped Racket for now.

Konrad.

From samth at cs.indiana.edu  Tue Sep 30 07:21:06 2014
From: samth at cs.indiana.edu (Sam Tobin-Hochstadt)
Date: Tue, 30 Sep 2014 07:21:06 -0400
Subject: [racket] Cyclic data structures in Typed Racket
In-Reply-To: <21546.36883.42112.592373@Konrad-Hinsens-MacBook-Pro-2.local>
References: <21545.33909.430329.240145@Konrad-Hinsens-MacBook-Pro-2.local>
 <CAK=HD+Y2x3d1e=caisU9xayyFqRvwm+EHT+Rg66yDCpEAhNGtA@mail.gmail.com>
 <21545.35298.582248.45625@Konrad-Hinsens-MacBook-Pro-2.local>
 <CAK=HD+Ywfnfv-mjQPpo9GmOO8u+aV7f6ue-4motks_GkVnNRaQ@mail.gmail.com>
 <21546.36883.42112.592373@Konrad-Hinsens-MacBook-Pro-2.local>
Message-ID: <CAK=HD+Z_oMxtoXmpbKu8-QM0ipjMxXykD_+beKtX0fq6+A8Bqw@mail.gmail.com>

On Sep 30, 2014 7:12 AM, "Konrad Hinsen" <konrad.hinsen at fastmail.net> wrote:
>
> Sam Tobin-Hochstadt writes:
>
>  > Ah, if you want to create truly cyclic structure with no base case
>  > like this, then you'll have to do more work. Either you'll need to
>  > expand the type of the `x` field in `foo` to allow an initial value,
>  > or you'll need to create a dummy `foo` with that extended type, and
>  > then copy the cyclic data once you've created it to a new version
>
> In my specific application, I ended up replacing the cyclic data
> structure, which was a singleton, by a function that simulates it.
> But for another project, which involves graphs, I'll stick to
> untyped Racket for now.

Robby reminds me that we wrote a paper about another trick that might be
useful here:
http://www.ece.northwestern.edu/~robby/pubs/papers/stop2009-tf.pdf

Sam
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140930/52aae4ef/attachment.html>

From konrad.hinsen at fastmail.net  Tue Sep 30 08:50:01 2014
From: konrad.hinsen at fastmail.net (Konrad Hinsen)
Date: Tue, 30 Sep 2014 14:50:01 +0200
Subject: [racket] Cyclic data structures in Typed Racket
In-Reply-To: <CAK=HD+Z_oMxtoXmpbKu8-QM0ipjMxXykD_+beKtX0fq6+A8Bqw@mail.gmail.com>
References: <21545.33909.430329.240145@Konrad-Hinsens-MacBook-Pro-2.local>
 <CAK=HD+Y2x3d1e=caisU9xayyFqRvwm+EHT+Rg66yDCpEAhNGtA@mail.gmail.com>
 <21545.35298.582248.45625@Konrad-Hinsens-MacBook-Pro-2.local>
 <CAK=HD+Ywfnfv-mjQPpo9GmOO8u+aV7f6ue-4motks_GkVnNRaQ@mail.gmail.com>
 <21546.36883.42112.592373@Konrad-Hinsens-MacBook-Pro-2.local>
 <CAK=HD+Z_oMxtoXmpbKu8-QM0ipjMxXykD_+beKtX0fq6+A8Bqw@mail.gmail.com>
Message-ID: <21546.42745.458056.221096@Konrad-Hinsens-MacBook-Pro-2.local>

Sam Tobin-Hochstadt writes:

 > Robby reminds me that we wrote a paper about another trick that might be useful here: 
 > http://www.ece.northwestern.edu/~robby/pubs/papers/stop2009-tf.pdf

Thanks, that's a nice approach. I had tried something similar (doing
the equivalent of a typecast), but hadn't thought of using the module
interface for this. I am also surprised to see that require/typed can
be used to import typed modules, but I am not complaining that it
works ;-)

It actually works fine when I use two modules in two separate files,
but when I tried using a submodule for the data definition, to keep
it with the client code, I ran into what is probably an unrelated
issue:

   #lang typed/racket

   (module cyclic typed/racket
     (provide (except-out (all-defined-out)) set-foo-x! set-bar-x!)
     (struct foo ([x : (U bar Void)]) #:mutable)
     (struct bar ([x : (U foo Void)]) #:mutable)
     (define f (foo (void)))
     (define b (bar (void)))
     (set-foo-x! f b)
     (set-bar-x! b f))

   (require/typed 'cyclic
     [#:struct foo ([x : bar])]
     [#:struct bar ([x : foo])]
     [f foo]
     [b bar])

   (eq? (bar-x b) f)
   (eq? (foo-x f) b)

require: unknown module
  module name: #<resolved-module-path:'cyclic>

If I convert the outer module to plain Racket and use a plain require,
this works fine. Are the rules for module references different in
Typed Racket?

Konrad.

From samth at cs.indiana.edu  Tue Sep 30 09:02:53 2014
From: samth at cs.indiana.edu (Sam Tobin-Hochstadt)
Date: Tue, 30 Sep 2014 09:02:53 -0400
Subject: [racket] Cyclic data structures in Typed Racket
In-Reply-To: <21546.42745.458056.221096@Konrad-Hinsens-MacBook-Pro-2.local>
References: <21545.33909.430329.240145@Konrad-Hinsens-MacBook-Pro-2.local>
 <CAK=HD+Y2x3d1e=caisU9xayyFqRvwm+EHT+Rg66yDCpEAhNGtA@mail.gmail.com>
 <21545.35298.582248.45625@Konrad-Hinsens-MacBook-Pro-2.local>
 <CAK=HD+Ywfnfv-mjQPpo9GmOO8u+aV7f6ue-4motks_GkVnNRaQ@mail.gmail.com>
 <21546.36883.42112.592373@Konrad-Hinsens-MacBook-Pro-2.local>
 <CAK=HD+Z_oMxtoXmpbKu8-QM0ipjMxXykD_+beKtX0fq6+A8Bqw@mail.gmail.com>
 <21546.42745.458056.221096@Konrad-Hinsens-MacBook-Pro-2.local>
Message-ID: <CAK=HD+a1dKRB6w9FcU3G7huF4_UZG2J1bpffuzsdxfdnGVCKVw@mail.gmail.com>

On Tue, Sep 30, 2014 at 8:50 AM, Konrad Hinsen
<konrad.hinsen at fastmail.net> wrote:
>
> If I convert the outer module to plain Racket and use a plain require,
> this works fine. Are the rules for module references different in
> Typed Racket?


No, there's a problem with submodules and Typed Racket. I recommend
just using separate files.

Sam

From tomi.pievilainen+racket at iki.fi  Tue Sep 30 11:11:49 2014
From: tomi.pievilainen+racket at iki.fi (Tomi =?iso-8859-1?Q?Pievil=E4inen?=)
Date: Tue, 30 Sep 2014 18:11:49 +0300
Subject: [racket] Preliminary support for Racket on EV3
Message-ID: <20140930151149.GW15446@shirio>

First apologies to Alexsandro Soares, if I'm intruding, but I wasn't
sure if you were on the mailing list and you were asking about this in
June.

I tested how well Racket runs on the Lego Mindstorms EV3 (with
ev3dev), and found out that there is enough memory on the device to
get a REPL running.  The loading time is somewhat annoying, but not
prohibitive. I also started creating a library for using the ev3dev
kernel exposed file based API. I tested that I can use the color and
ultrasonic sensors, and motors. The rest is simple from those.

I'm a bit stuck by trying to grok Racket better. I'm thinking about
possibility to have some kind of "typed literals" for distance, time
and angle, so that I could use (move-robot 11cm) or (move-robot 11s)
like calls, instead of having separate functions for time and distance
variables, or having to create the values with (cm 11) or such.

Not sure if that makes any sense, but I was thinking that it might
feel more natural and also I'm interested in seeing how easily it
could be done.

-- 
Tomi Pievil?inen, +358 400 487 504
A: Because it disrupts the natural way of thinking.
Q: Why is top posting frowned upon?


From deren.dohoda at gmail.com  Tue Sep 30 11:51:56 2014
From: deren.dohoda at gmail.com (Deren Dohoda)
Date: Tue, 30 Sep 2014 11:51:56 -0400
Subject: [racket] Preliminary support for Racket on EV3
In-Reply-To: <20140930151149.GW15446@shirio>
References: <20140930151149.GW15446@shirio>
Message-ID: <CAE-rtpn1g_qzftd-tYWMjWW7TPy95r4yH6V686d=KZ16i9RMfg@mail.gmail.com>

Straightforward to use a struct, or a a tagged value like (cons 'cm 10) and
then have your procedure parse the argument.

Deren
 On Sep 30, 2014 11:14 AM, "Tomi Pievil?inen" <
tomi.pievilainen+racket at iki.fi> wrote:

> First apologies to Alexsandro Soares, if I'm intruding, but I wasn't
> sure if you were on the mailing list and you were asking about this in
> June.
>
> I tested how well Racket runs on the Lego Mindstorms EV3 (with
> ev3dev), and found out that there is enough memory on the device to
> get a REPL running.  The loading time is somewhat annoying, but not
> prohibitive. I also started creating a library for using the ev3dev
> kernel exposed file based API. I tested that I can use the color and
> ultrasonic sensors, and motors. The rest is simple from those.
>
> I'm a bit stuck by trying to grok Racket better. I'm thinking about
> possibility to have some kind of "typed literals" for distance, time
> and angle, so that I could use (move-robot 11cm) or (move-robot 11s)
> like calls, instead of having separate functions for time and distance
> variables, or having to create the values with (cm 11) or such.
>
> Not sure if that makes any sense, but I was thinking that it might
> feel more natural and also I'm interested in seeing how easily it
> could be done.
>
> --
> Tomi Pievil?inen, +358 400 487 504
> A: Because it disrupts the natural way of thinking.
> Q: Why is top posting frowned upon?
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140930/049e54ba/attachment.html>

From jensaxel at soegaard.net  Tue Sep 30 12:31:18 2014
From: jensaxel at soegaard.net (=?UTF-8?Q?Jens_Axel_S=C3=B8gaard?=)
Date: Tue, 30 Sep 2014 18:31:18 +0200
Subject: [racket] Preliminary support for Racket on EV3
In-Reply-To: <20140930151149.GW15446@shirio>
References: <20140930151149.GW15446@shirio>
Message-ID: <CABefVgzZrei5LpOYx2yNik1cti35mCpU=YHjw-wSvSAn0ZUUog@mail.gmail.com>

Hej Tomi,

I have attached an example, in which 30s evaluates to (sec 30).

The idea is a follows:

   30s  is an identifier, therefore if 30s is unbound, then 30s is
  a reference to a top-level variable.

  References top top-level variables expand to (#%top . 30s).

  The standard #%top will raise an error (identifier unbound).
  The trick is now to redefined #%top to treat identifiers that
  end with s in a special way.

  See attached for details.

/Jens Axel




2014-09-30 17:11 GMT+02:00 Tomi Pievil?inen <tomi.pievilainen+racket at iki.fi>:
> First apologies to Alexsandro Soares, if I'm intruding, but I wasn't
> sure if you were on the mailing list and you were asking about this in
> June.
>
> I tested how well Racket runs on the Lego Mindstorms EV3 (with
> ev3dev), and found out that there is enough memory on the device to
> get a REPL running.  The loading time is somewhat annoying, but not
> prohibitive. I also started creating a library for using the ev3dev
> kernel exposed file based API. I tested that I can use the color and
> ultrasonic sensors, and motors. The rest is simple from those.
>
> I'm a bit stuck by trying to grok Racket better. I'm thinking about
> possibility to have some kind of "typed literals" for distance, time
> and angle, so that I could use (move-robot 11cm) or (move-robot 11s)
> like calls, instead of having separate functions for time and distance
> variables, or having to create the values with (cm 11) or such.
>
> Not sure if that makes any sense, but I was thinking that it might
> feel more natural and also I'm interested in seeing how easily it
> could be done.
>
> --
> Tomi Pievil?inen, +358 400 487 504
> A: Because it disrupts the natural way of thinking.
> Q: Why is top posting frowned upon?
>
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users



-- 
--
Jens Axel S?gaard
-------------- next part --------------
A non-text attachment was scrubbed...
Name: units-for-lego.rkt
Type: application/octet-stream
Size: 1680 bytes
Desc: not available
URL: <http://lists.racket-lang.org/users/archive/attachments/20140930/1721c83d/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: use-units-for-lego.rkt
Type: application/octet-stream
Size: 49 bytes
Desc: not available
URL: <http://lists.racket-lang.org/users/archive/attachments/20140930/1721c83d/attachment-0001.obj>

From jensaxel at soegaard.net  Tue Sep 30 13:02:28 2014
From: jensaxel at soegaard.net (=?UTF-8?Q?Jens_Axel_S=C3=B8gaard?=)
Date: Tue, 30 Sep 2014 19:02:28 +0200
Subject: [racket] problem with pict->bitmap and mrlib/write-animated-gif
In-Reply-To: <CAFrFfuG9GXu+N6Tp6Ccg1-n1QX6KFybMRpyj9p_Drk2+SRcrKw@mail.gmail.com>
References: <CAFrFfuHsyYBhWB01PnOf8E5HzwDM_69n0nAKfLyNn2UY1+nkhQ@mail.gmail.com>
 <CAFrFfuG9GXu+N6Tp6Ccg1-n1QX6KFybMRpyj9p_Drk2+SRcrKw@mail.gmail.com>
Message-ID: <CABefVgzfRg8XAAgtkub3n6_MPs-SLHV6G46JMGFXX8ncfwuPcg@mail.gmail.com>

Hi Martin,

2014-09-28 0:46 GMT+02:00 Martin DeMello <martindemello at gmail.com>:
> Solved, with help from Jens Soegaard; I had to modify mrlib/write-gifs to
> take a disposal argument, and use it in all calls to gif-add-control.

You can submit your changes here:

https://github.com/plt/racket/blob/master/pkgs/gui-pkgs/gui-lib/mrlib/gif.rkt

Click at the pen to the right of the history button.

/Jens Axel

From mb at mbtype.com  Tue Sep 30 15:24:01 2014
From: mb at mbtype.com (Matthew Butterick)
Date: Tue, 30 Sep 2014 12:24:01 -0700
Subject: [racket] Constraint satisfaction problems
Message-ID: <F0E2E41F-0005-4A46-B614-62E2439AA776@mbtype.com>

What are the options for modeling & solving CSPs in Racket?

Would Datalog be suitable? cKanren?

From jay.mccarthy at gmail.com  Tue Sep 30 15:28:32 2014
From: jay.mccarthy at gmail.com (Jay McCarthy)
Date: Tue, 30 Sep 2014 15:28:32 -0400
Subject: [racket] Constraint satisfaction problems
In-Reply-To: <F0E2E41F-0005-4A46-B614-62E2439AA776@mbtype.com>
References: <F0E2E41F-0005-4A46-B614-62E2439AA776@mbtype.com>
Message-ID: <CAJYbDampKFzMT9fEnCXExGNfGMfW8J1p6ai0kMPLcumKd5b4nw@mail.gmail.com>

cKanren would be best. datalog in particular isn't great because its
data is first-order, which can be awkward. A prolog (like racklog or
parenlog) might be good, but cKanren is nicer to program in.

Jay

On Tue, Sep 30, 2014 at 3:24 PM, Matthew Butterick <mb at mbtype.com> wrote:
> What are the options for modeling & solving CSPs in Racket?
>
> Would Datalog be suitable? cKanren?
> ____________________
>   Racket Users list:
>   http://lists.racket-lang.org/users



-- 
Jay McCarthy
http://jeapostrophe.github.io

           "Wherefore, be not weary in well-doing,
      for ye are laying the foundation of a great work.
And out of small things proceedeth that which is great."
                          - D&C 64:33

From spdegabrielle at gmail.com  Tue Sep 30 20:00:13 2014
From: spdegabrielle at gmail.com (Stephen De Gabrielle)
Date: Wed, 1 Oct 2014 01:00:13 +0100
Subject: [racket] canonical index of Racket courses? [was: Summer
 programs learning Racket for a student]
In-Reply-To: <CANwREeXtCqExVYCchvg4pvqFWs06tb4b=jtmU5zpdwTsAaM4Hg@mail.gmail.com>
References: <CANwREeXB1EWu2SjHBHuzhJKba4dHJFhdGHdg7SXafJqJ=v-nsw@mail.gmail.com>
 <CANwREeXVeAtEirzyXOGQedt8uU4EK--qYU--ZkNRGKWQ5hn8Ag@mail.gmail.com>
 <CADBC08E-D847-4AAE-87BE-D62368E91936@ccs.neu.edu>
 <CAK=HD+Yj==FD+N01Dr_hAzGfm+jieKO9k=F=2VnRpg0opgmV5Q@mail.gmail.com>
 <CANwREeXtCqExVYCchvg4pvqFWs06tb4b=jtmU5zpdwTsAaM4Hg@mail.gmail.com>
Message-ID: <CAGHj7-Jkmhw8VRVJs7U0_q9js2Xy+Y8m8vXZ=iw0oyxXko=xXQ@mail.gmail.com>

I'm sure there is a course at Middlesex university too... Can't find the
link. :(

On Monday, 29 September 2014, <jab at math.brown.edu> wrote:

> On Wed, Sep 17, 2014 at 10:51 AM, Sam Tobin-Hochstadt <
> samth at cs.indiana.edu
> <javascript:_e(%7B%7D,'cvml','samth at cs.indiana.edu');>> wrote:
>
>> I've now created a wiki page for this, with some initial content:
>> https://github.com/plt/racket/wiki/Courses-using-Racket
>>
>
> And now it's up to 22 revisions! Thanks for creating, Sam, and to everyone
> who added to it.
>
> On Tue, Sep 16, 2014 at 8:43 PM, Matthias Felleisen
>> <matthias at ccs.neu.edu
>> <javascript:_e(%7B%7D,'cvml','matthias at ccs.neu.edu');>> wrote:
>> > Do you imagine listing college courses such as Brown's 17, which uses
>> > DrRacket and the teaching languages? Or do you want Coursera courses
>> that
>> > everyone can access?
>>
>
> I think linking to any class that makes educational materials publicly
> available is valuable, but they should be divided into one section for
> classes that don't require being admitted into a larger institution to
> participate in (e.g. Bootstrap World, Coursera, summer programs for teens,
> etc.), and another section for those that do.
>
>
>> > And yes, we should probably create a wiki like thing for such an effort
>> or
>> > perhaps something like packages.racket-lang.org. I'll bring it up as
>> we meet
>> > in St Louis.
>>
>
> Thanks, Matthias, and hope you all had a great time in St Louis. Looks
> like it was a wonderful conference.
>
> Can the wiki page above be directly linked off racket-lang.org, perhaps a
> bit more prominently? I hadn't even noticed
> https://github.com/plt/racket/wiki before; now I see the link to it kind
> of buried under the Contributing column of the Community section. Perhaps
> it's not easy to find for others too.
>
> Thanks again for following up on this.
>


-- 
Sent from Gmail Mobile
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20141001/ebf52068/attachment-0001.html>

From jab at math.brown.edu  Tue Sep 30 22:09:10 2014
From: jab at math.brown.edu (jab at math.brown.edu)
Date: Tue, 30 Sep 2014 22:09:10 -0400
Subject: [racket] canonical index of Racket courses? [was: Summer
 programs learning Racket for a student]
In-Reply-To: <CAGHj7-Jkmhw8VRVJs7U0_q9js2Xy+Y8m8vXZ=iw0oyxXko=xXQ@mail.gmail.com>
References: <CANwREeXB1EWu2SjHBHuzhJKba4dHJFhdGHdg7SXafJqJ=v-nsw@mail.gmail.com>
 <CANwREeXVeAtEirzyXOGQedt8uU4EK--qYU--ZkNRGKWQ5hn8Ag@mail.gmail.com>
 <CADBC08E-D847-4AAE-87BE-D62368E91936@ccs.neu.edu>
 <CAK=HD+Yj==FD+N01Dr_hAzGfm+jieKO9k=F=2VnRpg0opgmV5Q@mail.gmail.com>
 <CANwREeXtCqExVYCchvg4pvqFWs06tb4b=jtmU5zpdwTsAaM4Hg@mail.gmail.com>
 <CAGHj7-Jkmhw8VRVJs7U0_q9js2Xy+Y8m8vXZ=iw0oyxXko=xXQ@mail.gmail.com>
Message-ID: <CANwREeU3ORGo96XzQMpNEGtGxyDy6MDRSBjeVX-kPBii6Zpp4g@mail.gmail.com>

First just want to reiterate that the new wiki page for Racket courses is
great, thanks to all who contributed!

Reread my initial post and noticed this bit:

On Wed, Apr 16, 2014 at 4:45 PM, <jab at math.brown.edu> wrote:

> (Similarly, linking to a canonical index of Racket meetup groups (? la
> http://clojure.meetup.com/ and
> https://wiki.python.org/moin/LocalUserGroups)  would be great too. Google
> turned up http://racket.meetup.com/ but 2 of the 3 results there are for
> racket sports.)
>

Taking a cue from Sam, I started a new wiki page for this at
https://github.com/plt/racket/wiki/Racket-Meetups, and added some initial
content following the template from
http://dev.clojure.org/display/community/Clojure+User+Groups. From there I
linked to another new page I started,
https://github.com/plt/racket/wiki/Start-a-Meetup, which for now is just a
stub suggesting http://clojure.org/start_group as prior art.

Hope it's okay I took the liberty of adding the new pages, that other
people will edit them liberally, and that even more people will find this
helpful.

If we can get the ball rolling here, would be great if we could then get
some links up to the new pages from racket-lang.org. Just might inspire
more people to start new Racket meetups. The power of suggestion! :)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.racket-lang.org/users/archive/attachments/20140930/81b42c56/attachment.html>