ATI Radeon 9500/9700 demo compatibility
category: general [glöplog]
Gargaj, good idea. I'll try to download a DX compatible screen grabbing program. Although I don't really have any webspace to host the images.
lepper, the points you mentioned are how it should look :)
ryg, thanks for the info :)
You guys have any crazy DX9'ified demos in the works?
You guys have any crazy DX9'ified demos in the works?
I haven't had problem grabbing screen shots of apps (even full screen apps) under Win2K by simply using Alt+PrintScreen.
some people seem to have more than enough time
Scali:
Theory #1 is probably more or less right. ATIs drivers _are_ a lot more buggy than nVidia's, at least when it comes to OpenGL (I've never programmed D3D, so I can't say anything about that). HOWEVER, they have a _lot_ better developer support (at least if your name isn't John Carmack). The three bug reports I've ever sent nVidia have just gone by unnoticed, with no reply _ever_ except for the usual "your bug has been entered in our system" autoreplies. ATI actually _answers_, and _fixes_ the problem.
Theory #3 is also more or less right. Too many people think "oh, heck, it works on my GF2MX"...
Theory #4 is probably wrong -- I don't know many OpenGL demos for Linux, really. I don't really see why this should matter for ATI vs. nVidia anyhow, except for the fact that the nVidia Linux drivers are probably a lot more mature than ATIs drivers (which are getting there, rapidly, but still aren't that stable yet).
Oh, and when I say "is" or "are" I mean "I think" or "in my experience" or "it appears" or something, okay? ;-)
Theory #1 is probably more or less right. ATIs drivers _are_ a lot more buggy than nVidia's, at least when it comes to OpenGL (I've never programmed D3D, so I can't say anything about that). HOWEVER, they have a _lot_ better developer support (at least if your name isn't John Carmack). The three bug reports I've ever sent nVidia have just gone by unnoticed, with no reply _ever_ except for the usual "your bug has been entered in our system" autoreplies. ATI actually _answers_, and _fixes_ the problem.
Theory #3 is also more or less right. Too many people think "oh, heck, it works on my GF2MX"...
Theory #4 is probably wrong -- I don't know many OpenGL demos for Linux, really. I don't really see why this should matter for ATI vs. nVidia anyhow, except for the fact that the nVidia Linux drivers are probably a lot more mature than ATIs drivers (which are getting there, rapidly, but still aren't that stable yet).
Oh, and when I say "is" or "are" I mean "I think" or "in my experience" or "it appears" or something, okay? ;-)
Paraphrased from reply to bug report to nVidia:
"hey, nVidia, your projected textures on gf1 are broken when trying to do it on 2 stages at once."
"yes, this is known, but is under NDA, so please just....uh, work around the problem."
"hey, nVidia, your projected textures on gf1 are broken when trying to do it on 2 stages at once."
"yes, this is known, but is under NDA, so please just....uh, work around the problem."
the only major bug I know of at the moment is that de_profundis-by_kolor hardlocks the PC on the 9700. ATI are aware and looking into it.
oh yeah, use the catalyst driver feedback form that is available now to submit bugs. It saves the devrel guys a lot of paper work, and makes it easier to collate info on specific bugs! They will still get back to you if it's an interesting bug. The guys at ATI love scene demos and always respond to bug reports! :)
i guess i have to buy a new graphic-card soon, not many new demos work with my good ol' TNT2. and, i'd also like to play around with the pixelshaders like all you big guys do :)
