[racket-dev] [plt] Push #24958: master branch updated

From: Robby Findler (robby at eecs.northwestern.edu)
Date: Wed Jul 11 23:20:28 EDT 2012

That all makes sense to me (except, of course, that the futures
visualizer has nothing to do with drracket. Were the docs not clear on
this point somehow?)

On Wed, Jul 11, 2012 at 10:06 PM, Eli Barzilay <eli at barzilay.org> wrote:
> Several randomly collected replies:
>
> * Re "perf" -- I dislike it because it's easy to confuse with
>   the very different "pref"...
>
> * Re my reply about a "future profiler tool" -- to clarify, I meant
>   "future" in the English sense, ie, some proper drracket tool that
>   will visualize the results of the statistical profiler.
>
> * Re moving things in there -- like Vincent said, some of these tools
>   have an interface that is independent of drracket or of any gui, so
>   it's less appealing.
>
> And that last point leads to what I think is the main issue here: the
> name of such a collection is not really important since its intended
> for drracket tools -- so it's not something that you'll actually need
> to type in.  I think that this clarifies the organization too: there's
> not much point in a toplevel collection that is dedicated to drr tools
> since then it turns into a place where unrelated things live.  In
> other words, it becomes yet another place to split code into in a unix
> filesystem kind of way:
>
>   foo/, tests/foo, scribblings/foo, and now: tools/foo
>
> vs having all of these be subdirectories of foo.  I think that in
> terms of maintaining and managing code, having all of foo's code in a
> single place is ultimately better.  The management point is in making
> it easy to package the complete foo code, keep it in its own
> repository, etc; the maintenance point is -- for example -- in making
> it clear that foo's author is responsible for the code rather than
> tools/foo where it implicitly becomes "more shared" with Robby.
>
> Given all of this, I think that the best option is still a new
> toplevel collection for the new tool, and the decision on the name is
> something that James and Robby (wearing the futures hat, not the
> drracket one) should consider as the home of additional functionality
> that might be added there.
>
> --
>           ((lambda (x) (x x)) (lambda (x) (x x)))          Eli Barzilay:
>                     http://barzilay.org/                   Maze is Life!

Posted on the dev mailing list.