From: Johannes G. <joh...@gm...> - 2007-09-16 22:30:59
|
Dear developers, My current ATI driver has a bug that prevents the use of [projection] flag_use_gl_point_sprite =3D true As far as I know, also other people are affected. Now I have found a dirty workaround using a fragment program. Interestingly on my hardware (radeon express 1150, X300) using the fragment program greatly reduces the performance, when compared to the hardwired rendering. Now my question to you is this: Has anybody got a really modern, good, ATI card? If so, do you experience the color bug that comes with the point sprites? And if so, would you care to try my patch? Johannes |
From: Johannes G. <joh...@gm...> - 2007-09-17 22:06:41
|
I have checked ATI X1600: the point sprite color bug does not occur. Point sprites work without my patch, but seem to be not faster than traditional rendering. I think I will discard my patch. Johannes On 2007.09.17 00:30:46 CEST, Johannes Gajdosik wrote: > Dear developers, >=20 > My current ATI driver has a bug that prevents the use of > [projection] > flag_use_gl_point_sprite =3D true >=20 > As far as I know, also other people are affected. > Now I have found a dirty workaround using a fragment program. > Interestingly on my hardware (radeon express 1150, X300) using > the fragment program greatly reduces the performance, when > compared to the hardwired rendering. >=20 > Now my question to you is this: Has anybody got a really modern, > good, ATI card? If so, do you experience the color bug that comes > with the point sprites? And if so, would you care to try my patch? >=20 > Johannes |