Proposal: Python Info Collective

[Andrew M. Kuchling]

usenet-mm-yyyy at julesallen.com (Jules Allen) writes:
> One of the main problems I've had with understanding Python's depth
> is finding information and modules. 

Here's a bit of history, which may not be known to people who've
recently joined the Python community.

The issue of indexing modules is an old one, and I'm still astounded
that we've never converged on a solution, though everyone always
complains about this.  Originally you'd upload things to
ftp.python.org (and occasionally people still do this), or, if your
project had a Web page of its own, you could send a note to the
maintainers of python.org, and they'd add a link to
http://www.python.org/download/Contributed.html .

This centralized system has been problematic for a long time; the
python.org maintainers are all quite busy, & their time is better
spent hacking on Python code.  A decentralized system would obviously
scale better, and the Locator-SIG was started in 1995.  Looking
through the SIG archive is probably useful:

ftp://ftp.python.org/pub/python-unmirrored/maillist-archives/locator-sig/locator-sig.archive

Nothing was ever really completed; Paul Everitt's prototype never
became widely used.  The backbone of Trove
(http://www.tuxedo.org/~esr/trove/) is partially implemented, but
there's no user interface yet, ESR is too busy to work on it much,
and no one else is really helping with it.

IMHO this isn't a technical problem; it's not too hard to write the
scripts something like freshmeat.net or CPAN.  The problem is getting
acceptance from the community; people need to start using it to look
for things, and that acts as a strong incentive for people to enter
their new modules into it.  For example, in the Perl community it's
deeply ingrained that you always, *always* release things on CPAN, and
follow CPAN's naming scheme.  We just need to push the Python
community very hard to enter their data.

http://www.vex.net/~x/parnassus/ is *very* nice, and it might be the
only solution required.  Apparently a good deal of work has been put
into seeding the initial database, and that's important, because it
makes the index immediately useful.  Yet I hadn't heard about it until
now; is the site still beta?  Did I miss a c.l.py.announce article
about it?  

There are various issues with it -- I don't know how scalable the
underlying code is, the graphic design could be simpler, vex.net's
reliability and bandwidth are unknown, and we can argue about the
classification hierarchy *forever* -- but it's still valuable.

We need to just pick a solution and stick with it; Parnassus is as
good as any, and has the incalculable advatange of being implemented.
I'd like to encourage everyone who maintains Python code, or writes
Python tutorials and docs, to go to http://www.vex.net/~x/parnassus/ .
Is everything you've done in the index?  If not, add the missing
items.  Can we prominently link to it from python.org, so people begin
to use it?

-- 
A.M. Kuchling			http://starship.python.net/crew/amk/
I am the Head and I demand your shrunken experiences now!
    -- In ENIGMA #1: "The Lizard, The Head, The Enigma"