Re: [PyOpenGL-Users] glVertexAttribPointer(); loading platform-specific functions

[Mike C. F. <mcf...@vr...>]

Joshua Davis wrote:
>>> I don't see any obvious problems with the attribute-based version. 
>>> I've
...
> The two versions of the program are attached below. They differ only
> in the initializeShaders() and handleDisplayEvent() functions. But I
> suppose the better diagnostic tool is the fact that the shader_4.py
> tutorial doesn't seem to generate any fragments on my machine (see my
> previous post).
Okay, on 2 Linux AMD64 machines (with a Radeon Mobility 3650 and a
Geforce 7600), the two scripts produce exactly the same output, these
machines have core shader/vbo support, but the script also works if I
disable use of the core libraries and use the ARB extensions on the
Radeon machine.  Hypotheses:

    * we could be looking at an error in the ARB versions on the Mac
      driver (extremely unlikely, this is bog-standard stuff)
    * could be a bug in how we're wrapping the ARB versions
          o might be a difference between accelerated/unaccelerated
            PyOpenGL, but doesn't appear to be (I've tested with both on
            Linux with this script)
          o might be a platform-specific requirement for the wrapper
            that we're missing (would be surprising, given that the rest
            of the extensions are working)
    * could be some subtle usage bug that Linux ignores but Mac rejects
          o I'd be surprised there, this is pretty straight-forward code

The fact that shader_4 demonstrates the same effect would suggest that
we've got something wrong in the wrapper, but without a Mac to test on
I'm somewhat limited here in tracking down the failure.  I may be able
to have a friend try the script at PyGTA this evening.

Good luck,
Mike

-- 
________________________________________________
  Mike C. Fletcher
  Designer, VR Plumber, Coder
  http://www.vrplumber.com
  http://blog.vrplumber.com