I've been trying to work on OpenGL-ctypes in the hours here and there. Mostly minor fixes. I've added the Demos directory from PyOpenGL. I've added and tested the Tk/Togl system with Togl 1.7 (system install (though there was no emerge for Gentoo and required re-emerging Python with Tkinter (yay, Tkinter isn't included in Python))) and it seems to work, but seems to have a different mechanism for calculating clipping planes (so certain Tk demos are "working" but not producing visible output).

I also started working (again) on the shader-compilation code. I'm working from a PyGame wiki example, but need to make a more general mechanism for handling the multi-string data-type. Basically I'm creating a List sub-class that has all strings and caches data-pointer-arrays and length-arrays.

Climbing onto the soap-box. This kind of thing is exactly why the ctypes version is useful. People can sit down and hack together a solution using Python and it's just a matter of refactoring the approach into the main code-base. Any Python programmer with an itch can extend the capabilities of the system and I can use their code as a guide for how to implement the functionality for the more general case with sample code I can use to test it.

I've been following the PyGame discussions about whether to make PyGame-ctypes the official PyGame or whether the current (C) implementation will continue to wear the PyGame top-level package name. My intention is still to make OpenGL-ctypes PyOpenGL-3.0. There may be a need for a few tiny extensions (though I'd like to have Python backups for those) for speed to make that happen, but I want the installation process to be dirt simple and standard (compared with PyOpenGL, which is a huge hacked-up set of distutils scripts that makes me cry when I work with it). We'll have to distribute a compiled version for Windows, of course, but it should be less of a PITA to create that compiled version.

BTW, people wanting to help out: someone who has a working Windows install (all of mine are dead!) and feels like debugging the port there would be very helpful in getting us to a first alpha release. We also need to start running applications and determining what functionality is and isn't working (and maybe fix that). People who want support for a particular extension could also start working on those extensions.