[racket-dev] proposal: `data' collection

On Jul  2, Robby Findler wrote:
> Those numbers seem pretty small in today's disk sizes,

Obviously -- but the issue is not diskspace.

And Jay McCarthy wrote:
> I feel like I routinely download programs and dev environments where
> the distribution is over 100MBs.

  winooski:~/mail eli> rpm -q --queryformat '%{SIZE} %{NAME}\n' tcl perl ghc js python ruby lua plt-scheme | sort -n
  595769 lua
  962834 js
  1441553 ghc
  1679403 ruby
  3669827 tcl
  22866733 python
  35175610 perl
  69558809 plt-scheme


Robby:
> but I do agree that there is value in being able to divide up the
> distribution and to be able to stratify things so we can better keep
> track of our dependencies.

Yes, that's exactly my point.

> (BTW, just a random question: have you thought about trying to
> visualize the collection-level dependencies with, say, dot?)

I didn't get to that yet.  I suspect that it's not what Petey did: it
should be a graph of dependencies between collections rather than
modules.  Cycles there should be much more alarming IMO than looking
at the module-level graph.


> It seems like you're after something that would allow multiple
> collections with the same name. Is that part of it, all of it, or
> mostly irrelevant to your main issue?

It is relevant, but inaccurate: what I'm after is a way to split
"packages" below the collection level.  Without that, the only way to
make an extensible `data' thing would be to have collections like
`data-list', `data-stack', etc.  (I'm *not* suggesting that as a
better alternative...)

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