Some OpenGL-ctypes hacking (Including getting auto-generation working again...)


Added a GLX sub-package to OpenGL-ctypes today. It was created using the auto-generation mechanism, which had drifted away from the subclass I created a bit. Seem to be back to usable results now.

Tried to get the WGL sub-package created, but didn't have one of the MSVC libraries I needed to do the auto-generation. Also integrated a submitted patch for win32. There are a number of things the new auto-generator does that I need to fix (e.g. the _libraries[ ] lines at the top and the fact that libraries no longer have names (they have the _libraries[ ] lines instead, and there are asserts in the code that are incorrect when you build the code on a 64-bit architecture but run it on a 32-bit architecture).

Added preliminary support to allow for using lists/tuples as array input types even if there is no numpy module available. That code is pretty slow compared to the numpy version, but it does seem to work. I also looked into gl2ps support. It's LGPL, so wouldn't want to include the code in the distribution, but AFAICS it doesn't have a library form that the user could install separately. That is, there's no loadable .so available, it's just assumed to be statically linked into each program that uses it. Oh well.

Comments

Comments are closed.

Pingbacks

Pingbacks are closed.