Re: [pygame] PCR & other notes (long)

[Scott Russell <scott@pantastik.com>]

On Mon, 2003-03-24 at 04:18, Pete Shinners wrote:
> Robert Mark Waugh wrote:
> > With the abscence of anyone volunteering to help with the PCR I will 
> > offer my services.
> 
> robert, thanks for this offer, i think we'll be able to use your input. 

Absolutely - we can use as much input as possible.  And just as
importantly, I need that tarball you owe me, Pete -- or I'm going to
just scrape the data and rewrite the PCR from scratch. :)

> actually scott russel has volunteered to be the full time pcr maintainer 
> for awhile, but it looks like we'll need some help getting things 
> reorganized a little.

If anyone wants to yell at me, I'm usually in python and pygame IRC as
queuetue.

> your other notes are also very solid. PCR definitely needs more 
> 'snippets', and i think it would be excellent for these to get 
> automatically (by a maintainer) pulled from the mailing list and chat 
> channel as things come up.

I like the idea, but would be concerned about maintaining quality from
auto-gleaned snippets.  If anyone wants to produce an explanation (or
some code, even better!) of how this could work, I'm definitely
receptive.

In talking to Pete, I can see three distinct "phases" I want the PCR to
go through:

1) Triage.  Stabilize the current system, get it accepting new
submissions, and show to the community that it's open for business
again.

2) Improvement.  Pete and I both thing a rating system is needed (This
would help allay my concerns over quality of automated code gathering). 
I also want to think hard about a classification and searching system to
make finding pertinent code (and similar examples) functional.  It might
be nice if we could make this available to the greater python community
at some future date.  I envision something similar in feel to linking in
a wiki or using Everything2.  I think the PCR could become a living
best-practices model for python programming in general. 

3) Rethinking.  I'd like to find (or create) people interested in
pythonizing famous graphics and game tutorials and algorithms.  The NeHe
OpenGL tuts are examples of this effort.  Also, It's intersting to
consider some mechanism, standard and interface to make pcr snippits
immediately useable.  

The next two paragraphs are a slightly "out there" rant and not very
pygame-oriented.  Feel free to skip them. :)

I have an ulterior motive in looking to help the PCR - I think python
would be well-served if it had a ubiquitous and trustworthy
installation, caching, uninstallation distribution and componentization
system. 

Looking to Pypi, portage and CPAN as distribution models, running on
something like bittorrent, with avodogado-style trust metrics and ssl
for layer security. Reusable code, defined using the interfacing
standards shared by Zope and PEAK, and learning from the great work
SkunkWeb is doing with it's low-impact componentization infrastructure
of what are effectively code snippets.  A consistent event horizon is
forming around the OSS community's efforts and I think that python will
sit at the center of what is coalescing, with pygame, piddle and
wxwindows providing it's visible aspect and canvas. I see the PCR as one
of many early steps to get us there.  End Rant.

Let's not get distracted, though -- the *concrete* goal is to make PCR
the best pygame snippets directory we can make it.

Foundation libraries are a vitally important part of what pygame can be,
and would serve to grow the community (Are you ready for a population
explosion, Pete?) beyond us hobbyist game developers.  I hope that
Pygame, in one form or another, will one day be considered one of the
technologies that put python on the map for the average developer.

I don't necesarily think that management and coordination those
libraries fall under the PCR's charter, though.  Regardles sof my
high-falootin' stance from above, the PCR project needs to be about one
thing:  effortless access to quality code snippets to help people
understand "The pygame way".  It will feed those library efforts, as all
efforts will feed it, but I feel a need to maintain a wall between it
and other projects - the PCR conceptually merges so easily with other
projects that it would be too easy to lose all sense of what it is and
of it's own value.

This said, scroll back up above this goofy manifesto I've surprised even
myself with, and look at stages 1) and 2).  These should be our
immediate focus (or mine, at least!) and I would appreciate feedback as
soon as possible.  And that tarball, Pete. :) (He's such a busy boy!)

- Scott Russell