opengl and dx

[jjayb]

We've seen the Ati demos and the Nv30 screenshots. Yes I know the Nv30 screenshots might not actually be Nv30 screenshots, please don't turn this thread into a debate on that. It looks like you can do some pretty amazing stuff with dx9. Is any of this possible with opengl? If not now, will you be able to do this with opengl 2.0? When is opengl 2.0 supposed to be finished? I know absolutely nothing about programing. I'm just a gamer who's really excited about what I've seen with these demos. Basically, I'm hoping that with the amazing things we've seen with dx9, we'll see even more amazing things with opengl 2.0. Just being greedy and wanting to know if I'll get more of my eye candy fix.

While on the topic of opengl and dx, what are the merits of each api? Why do some developers choose to use dx, while others use opengl? Ease of use? Can do better effects with one over the other? One faster than the other?

 

[demalion]

According to ATi, they have a functioning extension concurrent with 9700 release that will expose the PS/VS 2.0 functionality for OpenGL. Whether that is very useful at all at this point I don't know.

 

[Dave Baumann]

We've already heard mention of 'version 2' of NV's vertex shader and pixels shader OGL extensions as well, and they are likely to expose all of NV30's functionality through that (in fact if DX9 is limited to PS/VS2.0 then its their best shot at getting the functionality exposed in the short term).

 

[tEd]

it'll probably very useful for developers. How do you think should carmack take advantage of the new features if there are no extensions?

 

[demalion]

it'll probably very useful for developers. How do you think should carmack take advantage of the new features if there are no extensions?

I mean in the sense of it not being standardized when OpenGL is at the point where a standard is being adopted. In other words, the ARB should decide on the interim 1.4 (or have they already... was that really the official 1.4 announcement a while ago...I had the impression it was premature...) until 2.0 comes out, and that an enhanced set of extensions by either vendor is of questionable usefulness unless 2.0 is going to be significantly delayed (but Carmack has already tested a 2.0 implementation...is it just the HLSL that is going to hold up release?)

 

[MfA]

OpenGL 2.0 would need a couple extensions too to properly expose DX9 hardware.

DirectX sets the standard, OpenGL tries to be different ... and extensions close the gap, and add some bells and whistles m$ refused.