[racket] racket/place seems to be incompatible with DrRacket

When I add a step "2b" to save the file, then I get the correct
behavior (using a git build from today). Do you see that?

Robby

On Tue, Aug 30, 2011 at 9:54 AM, David Mitchell
<dave at fallingcanbedeadly.com> wrote:
> Steps:
> 1. Open DrRacket
> 2. In the definitions window, paste in
>
> #lang racket
> (define (super-simple)
>   (define p
>     (place ch
>            (place-channel-put ch 5)))
>   (place-channel-get p))
>
> 3. Click the "Run" button in the toolbar
> 4. In the REPL, type "(super-simple)"
> The call to provide is necessary when loading code in the Terminal, but it
> doesn't seem necessary with DrRacket. In fact, based on the comment I saw in
> IRC about DrRacket reloading the module, I wonder if that has something to
> do with it. (although I'm new enough to Racket to not put much stock in my
> intuitions)
> — Dave
>
> On Tue, Aug 30, 2011 at 7:19 AM, Robby Findler <robby at eecs.northwestern.edu>
> wrote:
>>
>> I'm not seeing bad behavior so probably I'm doing something slightly
>> differently than you, which is making the difference. If you have the energy
>> to send a precise set of steps I'm happy to try them on my machine and maybe
>> that'll help us sort it out.
>>
>> Robby
>>
>> On Tuesday, August 30, 2011, David Mitchell <dave at fallingcanbedeadly.com>
>> wrote:
>> > I tried saving the module and putting a require in the DrRacket window,
>> > and that works correctly. I think this will be an acceptable workaround for
>> > the time being, but I'd really like to know more about what's going on.
>> >
>> > -- Dave (on the iPhone)
>> >
>> > On Aug 30, 2011, at 4:44 AM, Robby Findler <robby at eecs.northwestern.edu>
>> > wrote:
>> >
>> >> I'm sorry; I was wrong.
>> >>
>> >> I don't see the behavior you're seeing, tho. If I add parens around
>> >> 'all-defined-out', and save the file in tmp.rkt, then this is the
>> >> behavior I see at the command-line racket:
>> >>
>> >> Welcome to Racket v5.1.3.6.
>> >>> (require "tmp.rkt")
>> >>> (super-simple)
>> >> 5
>> >>
>> >> and I see similar behavior in DrRacket:
>> >>
>> >> Welcome to DrRacket, version 5.1.3.6--2011-08-25(7bb8c6c4/d) [3m].
>> >> Language: racket; memory limit: 128 MB.
>> >>> (super-simple)
>> >> 5
>> >>>
>> >>
>> >> FWIW, in 5.1.3, I get an error that's surprising to me, as I don't see
>> >> how the sandbox gets involved in this program, but here it is:
>> >>
>> >> car: expects argument of type <pair>; given '#((module program
>> >> racket/base))
>> >>
>> >> === context ===
>> >> /Applications/Racket v5.1.3/collects/racket/sandbox.rkt:572:17: ns
>> >> /Applications/Racket v5.1.3/collects/racket/sandbox.rkt:584:20
>> >> /Applications/Racket v5.1.3/collects/racket/sandbox.rkt:334:0:
>> >> call-with-limits
>> >> /Applications/Racket v5.1.3/collects/racket/sandbox.rkt:564:0:
>> >> evaluate-program
>> >> /Applications/Racket
>> >> v5.1.3/collects/racket/private/more-scheme.rkt:151:2:
>> >> call-with-break-parameterization
>> >> /Applications/Racket v5.1.3/collects/racket/sandbox.rkt:704:2:
>> >> user-process
>> >> /Applications/Racket
>> >> v5.1.3/collects/mred/private/wx/common/queue.rkt:430:6
>> >> /Applications/Racket
>> >> v5.1.3/collects/mred/private/wx/common/queue.rkt:470:32
>> >>
>> >> Robby
>> >>
>> >> On Mon, Aug 29, 2011 at 8:55 PM, David Mitchell
>> >> <dave at fallingcanbedeadly.com> wrote:
>> >>> You might have to unpack that a little bit more for me, as I don't see
>> >>> how that's the case. Additionally, why would it work from Terminal and not
>> >>> DrRacket?
>> >>>
>> >>> -- Dave (on the iPhone)
>> >>>
>> >>> On Aug 29, 2011, at 6:32 PM, Robby Findler
>> >>> <robby at eecs.northwestern.edu> wrote:
>> >>>
>> >>>> I think what's happening is that you're creating a place whose first
>> >>>> action is to create that same place again (over and over).
>> >>>>
>> >>>> Robby
>> >>>>
>> >>>> On Mon, Aug 29, 2011 at 8:17 PM, David Mitchell
>> >>>> <dave at fallingcanbedeadly.com> wrote:
>> >>>>> I've been trying to experiment with racket/place, but it seems to
>> >>>>> have
>> >>>>> significant incompatibilities with DrRacket. I've defined a module
>> >>>>> based on
>> >>>>> a code snippet in
>> >>>>> http://www.cs.utah.edu/plt/publications/dls11-tsffd.pdf
>> >>>>> that looks like this:
>> >>>>>
>> >>>>> #lang racket
>> >>>>> (provide all-defined-out)
>> >>>>> (define (fib n)
>> >>>>>   (define (fib-inner n-remaining i i+1)
>> >>>>>     (if (= n-remaining 1)
>> >>>>>         i+1
>> >>>>>         (fib-inner (- n-remaining 1) i+1 (+ i i+1))))
>> >>>>>   (if (<= n 0)
>> >>>>>       0
>> >>>>>       (fib-inner n 0 1)))
>> >>>>> (define (start-fib n)
>> >>>>>   (define p
>> >>>>>     (place ch
>> >>>>>            (define n (place-channel-get ch))
>> >>>>>            (place-channel-put ch (fib n))))
>> >>>>>   (place-channel-put p n)
>> >>>>>   p)
>> >>>>> (define (super-simple)
>> >>>>>   (define p
>> >>>>>     (place ch
>> >>>>>            (place-channel-put ch 5)))
>> >>>>>   (place-channel-get p))
>> >>>>>
>> >>>>> If I run racket in my Terminal and load this module, I'm able to use
>> >>>>> start-fib and super-simple as expected. However, if I define this
>> >>>>> module in
>> >>>>> the DrRacket editor and access the functions through the repl,
>> >>>>> attempts to
>> >>>>> rec
>
> _________________________________________________
>  For list-related administrative tasks:
>  http://lists.racket-lang.org/listinfo/users
>