As long as it doesn't interfere with the development of Ren'Py, I'd like to
contribute as much as possible of the technology stack that underlies
Ren'Py to the pygame project.


Pygame_sdl2 is the obvious choice for inclusion, as it was really meant to
be an API-compatible successor to Pygame. I think it would benefit from
being merged with pygame proper, as much of the missing functionality is
based on C code found in the Pygame project.

The rapt and renios modules - which are used for Android support and iOS
support, respectively - are also good candidates. Both are already capable
of packaging arbitrary pygame(_sdl2) programs, at least in theory. A
problem that needs to be solved is how to include arbitrary python modules
in the distribution.

There's also a bit of code in Ren'Py that should really belong in pygame,
that I've been avoiding adding because it's a completely new API. In-app
purchase support is one of these, and achievement support is another -
although the latter may be complicated by licensing issues.


A project I've been contemplating is extending Ren'Py's launcher to support
arbitrary pygame code. For those not familiar, the Ren'Py launcher supports:

- Automatically downloading and installing text editors and other
applications needed for development.
- One-click packaging for Windows, Mac, and Linux - including all three in
a standalone zip file.
- Automatic installation of rapt and the various android tools, and
creation of Android apps.
- Automatic installation of renios and creation of iOS apps.

While the launcher is part of Ren'Py, I'd be happy to add pygame-specific
code to Ren'Py.


My one worry is that Linux distros would wait for stable releases of
pygame_sdl2 before packaging Ren'Py, leading to substantial holdups. My
hope is that could be solved by trying to keep master in a releasable
state, and making it clear to distros that it's acceptable to package
master up for releases.






On Sat, Jul 11, 2015 at 12:45 AM Daniel Foerster <pydsig...@gmail.com>
wrote:

>
> On 07/10/2015 09:47 PM, Peter Shinners wrote:
> > I haven't been paying close attention to Pygame, but it doesn't seem
> > controversial to say things have stalled.
> Not at all, unfortunately.
> > * Getting 1.9.2 actually released
> At the least we need a stable release so that Python 3 support can land
> in official repositories.
> > * Moving on to "Pygame 2", whatever that means
> I don't know how the rest of the mailing list feels, but  looking at SDL
> 2 (perhaps along the lines of Ren'Py's pygame_sdl2 project?) seems like
> a decent starting point.
> > * Catch up on the Bitbucket pull requests
> > It seems there are still many great people involved with the Pygame
> > project. Perhaps I can help by getting those people the control they
> > need to make progress.
> We're suffering from a lack of people who have the time and power to
> review and approve the PRs as much as anything.
> > * Website replacement and love
> It's definitely overdue for an overhaul, I'd be willing to work on that
> if I can get some information on what server-side tools are available
> (time for some modern Python web frameworks and templating abilities!).
> > * Migrate forum to Reddit (or community forum)
> I would suggest checking out Discourse, it seems like a pretty well
> engineered solution, much better than phpBB would be at the least!
> > I'm completely detached from things at this point, so I don't have any
> > context to jump in and try to change anything. What parts of the
> > project are going well these days?
> People are still writing cool stuff in Pygame and bugs seem to still be
> getting fixed.
>
>

Reply via email to