Lots of array-handling refactoring (Mmm refactored code...)

Quite a bit more work on the array-handling code in OpenGL-ctypes. It seems to be working properly now, and the number of OpenGLContext demos that are running is growing substantially (even have the first bitmap-based demo working (though it's a pretty simple demo, no PIL involved)).

Something wonky is going on with the GLUT bitmap font pointers. Under the covers these are actually c_void_p values retrieved from the DLL, but for some silly reason using the values is causing a core dump.

Also found a weird case where glRenderMode is returning 0 if a select rendermode pass fails. Not sure if this is a problem with OpenGLContext trying to call glRenderMode( GL_RENDER ) when in GL_RENDER or OpenGL-ctypes doing something wonky. Oh well.

Biggest thing I still want to do with the auto-generation is to provide a way to easily declare an argument "output". That should simplify and reduce a lot of code in the various modules. The wrapper code is already written to support this kind of operation, it's just a matter of factoring the common pattern out.

After that, the next biggest item to tackle is image handling. Brian is going to be looking at the extension loading/handling. GLU is still pretty broken; tessellation and nurbs don't work at all. GLE hasn't even been auto-wrapped yet, let alone tested.

Anyway, think I'm done for the night. Though I've had so much sugar I probably need to do some exercise to work it off before I try to sleep.


Comments are closed.


Pingbacks are closed.