[racket-dev] Release Announcement for v5.3, third draft

From: Eli Barzilay (eli at barzilay.org)
Date: Thu Aug 2 21:59:27 EDT 2012

[Aggregated reply.]

10 hours ago, Ryan Culpepper wrote:
> * A new `json' library (derived from Dave Herman's PLaneT library) for
>    parsing and generating JSON comes with the distribution.

Please revert this to the original phrasing.  The code is "derived" in
a weak sense, since pretty much all of it was rewritten, and there are
differences in both function names and the type that gets used.

> * `racket/string' is extended with string manipulation functions that
>    avoid some basic regexp tweaks.

My original phrasing here wasn't great, but the above is worse: I
parse this sentence as saying that there were some regexp tweaks that
were needed for some reason, and that's fixed.  Maybe just this:

      `racket/string' is extended with a set of simplified string
      manipulation functions that are more convenient than using

> === DEPRECATION ==========================================

("Graphically" speaking, this separator is too bold.)

Also, the two semi-deprecations-by-moving-to-planet items fit this

10 hours ago, Sam Tobin-Hochstadt wrote:
> I think the below should be somewhat reordered, as follows:
> On Thu, Aug 2, 2012 at 11:16 AM, Ryan Culpepper <ryan at cs.utah.edu> wrote:
> > * Most error messages use a new multi-line format
> > * The new `racket/generic' library
> > * The contract library comes with support for
> > * The futures visualizer
> > * Optimization Coach
> > * Typed Racket supports function definitions with keyword arguments
> > * A new `json' library
> > * The `class' form supports declaring a method abstract.
> > * The new `images/flomap' library
> > * `racket/string' is extended

I'd change this with
- putting the json item higher since it's a popular library and since
  it is currently eye candy ^H^H^H buzzwordy enough to draw attention
  from quick skimmers (not schemers).
- raising the string item too, at least I was surprised by the amount
  of excited feedback to it.
- lower the generic item, since it's not something that many people
  will use (at least in this early stage).
- The error messages item should be relatively high since it's
  something that will make most people's first impression, but it
  shouldn't be at the top part since it's not (yet) a functionality

10 hours ago, Ryan Culpepper wrote:
> On 08/02/2012 11:16 AM, Ryan Culpepper wrote:
> > * There is now a very complete completion code for zsh.  It is not
> >    included in the distribution though, get it at:
> >    http://goo.gl/DU8JK (This script and the bash completions will
> >    be included in the standard installers in future versions.)
> Should we leave this item out and include it in the future, once the
> scripts are actually included in the distribution?

Please leave it in -- a major clientele for these announcements are
third-party packagers and sysadmins who often add such files as
patches.  Also, it would be good to have feedback on this before it's
distributed and installed by default.

In fact, I explicitly mentioned the bash completions file because it
wasn't done when it came out.  (And BTW, the code there is shaky
enough IMO that it could *really* use some testing and hopefully some
looking over from bash completion experts.)

10 hours ago, Sam Tobin-Hochstadt wrote:
> I attempted to order them by (my impression of) importance to Racket
> programmers.  [...]

If there is no clear agreement on the order then a possible option is
to do the voting thing that we used in the past where you order
your preferences and the result is the order that satisfies most
people.  (Yes, it'll be a minor hassle, but it seems to me like a good
idea since this stage is fertile bikeshedding stuff.)  Setting it up
will take about two minutes.

10 hours ago, Sam Tobin-Hochstadt wrote:
> First, the grouping was not at all obvious from the text of the
> release announcement.  Second, we (meaning the release manager)
> should make an effort to sort by importance.  Release announcements
> are one of the primary ways that we communicate the cool things
> we're doing to our users.  We should take advantage of that to help
> them use some of the really cool new things we've added, like better
> error messages and generics.

(An obvious plus one for the importance of the order, and a couple of
hundred minus ones for the implication that the order & grouping was
not carefully thought out.)

          ((lambda (x) (x x)) (lambda (x) (x x)))          Eli Barzilay:
                    http://barzilay.org/                   Maze is Life!

Posted on the dev mailing list.