I apologies if I have mistakenly given you the impression that I have read every thread on this board in the last 2.5 years, and every post in all those threads.

I will definitely attempt to be...

Why can't you just resend calls to glVertexPointer(), glColorPointer(), etc, offsetting them by the required index offset?

You don't have to walk through the index list this way.

I have reported the bug in our internal bug tracking system.

Have a look at GL_ALPHA_TEST and glAlphaFunc().

Which is fine for bilinear and to a certain extent trilinear. It's a whole other matter to (efficiently) do aniso.

What's your monitor refresh rate? Sounds like you're waiting for vsync.

This step isn't (yet) hardware accelerated on any commercial GPUs.

You will need to either use the blending functionality, or render to texture (with copy) to get the effect you want.

The documentation for ADDSIGNED2X was probably written when GPUs didn't all use floating-point numbers for fragment processing.

I'm assuming you have a non-symmetric blend function.

If you think about your blend function, you're basically taking the background and mixing it in with the fragments you're rendering.

As...

How are you measuring your time? Are you sure that you're not synchronized to your monitor's vertical refresh signal?

How would instancing help you there? The GPU will still need to transform the same number of vertices and fetch the same number of attributes, with or without instancing.

The only thing you save...

Like I mentioned, I don't disagree that accelerated 1.4 is much better than software 1.1.

I'm much more worried about the new UI being turned off when the IHV's ICD is loaded, though. This is...

You should probably reread the spec.

The compiler thus cannot put a scalar varying in the w component (or any other unused component) of an attribute.

Perhaps, but then why not pick up GL 1.5? GL 1.5 only adds VBOs, occlusion querries and generalized shadow compare functions to the core.

Without either VBOs or IHV extensions for vertex arrays,...

Would you pay a premium to play your favorite games in Windows? Or is a worse user experience compared to (say) Windows XP acceptable?

Also note that a conformant implementation can (but need not to) implement glDrawRangeElements() as just glDrawElements() (with additional parameter checking, of course). glDrawRangeElements()...

I question your intent.

Backfacing polygons are normally culled by backface culling. That is, they don't get rasterized anyway.

However, you can rasterize back-facing polygons if you disable...

Our unified drivers work on everything from the latest GeForce back to the venerable RivaTNT2.

Release Notes for Forceware 71.89 . Look at pages 25 to 28.

You should use a type of GLhalfARB instead.

This should be:

glTexImage2D(GL_TEXTURE_2D, 0, GL_ALPHA16F_ARB, 256, 256, 0, GL_ALPHA, GL_HALF_FLOAT_ARB, tab);

My understanding of the issue is that the wording for the behavior of GL_CLAMP used to be such that doing the same thing as GL_CLAMP_TO_EDGE was acceptable. With OpenGL 1.2, the wording for GL_CLAMP...

No it doesn't.

Behold:
The pixel format descriptor is const.

What you actually need to do is check the return value, and if it's valid, call DescribePixelFormat() with that pixel format.
...

That would have to be a very specific case: You know the alpha values are all failing after the vertex program, and your fragment program does not modify alpha. This is not exactly a common case.

Obviously, not all possible applications can be written this way. Consider a function closer to OpenGL 1.0: glReadPixels().

What idr says is that some applications can be designed this way.

Cg will compile down to ARB_fragment_program and ARB_vertex_program, which (AFAIK) are supported by the ATI Radeon 9500 and later GPUs.

Cg won't generate the most optimal code for ATI hardware...

What's wrong with the linear/swizzled terminology again? Some of us actually use the term 'linear' to mean 'unswizzled'.