OpenGL lighting fixes

[John W. Eaton]

On  4-Feb-2009, Kai Habel wrote:

| Michael Goffioul schrieb:
| > ...
| >
| >   
| >> 2.) The surface normals should be calculated when facelighting is set to
| >> 'flat' too.
| >>     
| >
| > That's not my understanding of Matlab documentation. When facelighting
| > is flat, the ligthening should be uniform across the each quad and equal
| > to the lightening of the first vertex. That's why I don't update the normal on
| > the usbsequent vertices.
| >
| >   
| > ...
| 
| Thanks,
| for the reply and brief explanation. I did some tests myself, and can
| confirm the behavior you have described above. So, I have reworked the
| patch that only the calculation of diffuse reflection and surface
| normals is modified.

I applied this changeset.

Thanks,

jwe