Regarding your statement: "In fact most modern hardware (Radeon, Geforce) will prefer it if you pack all your vertex components together, as DirectX has always worked. This all to make more efficient use of memory bandwidth. "
For GeForce this is NOT true. The cache for DMA transfers of non-interleaved arrays makes it just as good as interleaved arrays (or packed components). The primary exception here would be if padding were necessary for an individual array (eg GL_SHORT normals).
In general, it's much more important to write sequentially to AGP memory since the memory is uncached. This usually means that static vertex attributes should be *not* be interleaved with dynamic ones.
It bothers me to hear people chime in and complain about nVidia not supporting the DRI being anti-open-source. The DRI itself does not require that chip specifications be open-source for the very purpose of allowing IHVs to protect their IP. It seems that problem that Frank LaMonica has with nVidia is that they are not using "his" infrastructure. That may be a reasonable complaint in terms of multimon support and other interoperability issues associated with multiple heterogeneous graphics accelerators -- that issue will certainly need to be addressed. I don't think the answer should necessarily be that every IHV must use the DRI.
What is really disappointing to me is that nobody seems to take notice that nVidia is one of the few companies actually putting heavy resources into STRONG OpenGL drivers. They provide excellent documentation, examples, and whitepapers, for the extensions they support and useful rendering techniques. Other companies have generally let their OpenGL drivers fall by the wayside, instead devoting most of their time to Direct3D drivers. They farm out the OpenGL driver development to external parties or the open-source community because it is cost effective. If they had a genuine commitment to doing the "right thing" then they would have been devoting real resources to OpenGL all along.
Don't let the distinction between whose driver infrastructure is being used confuse the issue of which companies are actively supporting open standards, open-source products, and which ones are just releasing their specs and letting other people do their work.
I'm not knocking the companies that have released their specs! If they don't have the resources to develop OpenGL drivers for their hardware, I'm ecstatic that they're letting somebody do it for them.
Slashdot Mirror
Regarding your statement: "In fact most modern hardware (Radeon, Geforce) will prefer it if you pack all your vertex components together, as DirectX has always worked. This all to make more efficient use of memory bandwidth. "
For GeForce this is NOT true. The cache for DMA transfers of non-interleaved arrays makes it just as good as interleaved arrays (or packed components). The primary exception here would be if padding were necessary for an individual array (eg GL_SHORT normals).
In general, it's much more important to write sequentially to AGP memory since the memory is uncached. This usually means that static vertex attributes should be *not* be interleaved with dynamic ones.
Thanks -
Cass
It bothers me to hear people chime in and complain about nVidia not supporting the DRI being anti-open-source. The DRI itself does not require that chip specifications be open-source for the very purpose of allowing IHVs to protect their IP. It seems that problem that Frank LaMonica has with nVidia is that they are not using "his" infrastructure. That may be a reasonable complaint in terms of multimon support and other interoperability issues associated with multiple heterogeneous graphics accelerators -- that issue will certainly need to be addressed. I don't think the answer should necessarily be that every IHV must use the DRI.
What is really disappointing to me is that nobody seems to take notice that nVidia is one of the few companies actually putting heavy resources into STRONG OpenGL drivers. They provide excellent documentation, examples, and whitepapers, for the extensions they support and useful rendering techniques. Other companies have generally let their OpenGL drivers fall by the wayside, instead devoting most of their time to Direct3D drivers. They farm out the OpenGL driver development to external parties or the open-source community because it is cost effective. If they had a genuine commitment to doing the "right thing" then they would have been devoting real resources to OpenGL all along.
Don't let the distinction between whose driver infrastructure is being used confuse the issue of which companies are actively supporting open standards, open-source products, and which ones are just releasing their specs and letting other people do their work.
I'm not knocking the companies that have released their specs! If they don't have the resources to develop OpenGL drivers for their hardware, I'm ecstatic that they're letting somebody do it for them.
Cass Everitt
-----
http://www.r3.nu/