Hello :) Thank you for reply.
> The nvidia libGL libraries should override the Mesa ones as they are > conflicting, ie provide the same functions but are not HW accelerated. aha :) Ok than the diversion seems to be justified. But still I do not see why the setiathome decreeses the 3D acceleration so much. > at which priority (nice-level) does Seti run on your machine? Does 'nice > -n0 glxgears' change anything? (Or perheaps 'nice -n-20 glxgears' as > root I run setiathome from wmsetimon applet (wmsetimon 0.5-2). It is somehow broken and gives it a nice value 0 even if I specify nice value 20 in .wmsetimonrc. But anyway, if I : $renice +20 8691 8691: old priority 0, new priority 19 $ and after that: $ glxgears 935 frames in 5.0 seconds = 187.000 FPS 947 frames in 5.0 seconds = 189.400 FPS 998 frames in 5.0 seconds = 199.600 FPS 1345 frames in 5.0 seconds = 269.000 FPS 950 frames in 5.0 seconds = 190.000 FPS But if I stop setiathome and than run $ glxgears 15898 frames in 5.0 seconds = 3179.600 FPS 29568 frames in 5.0 seconds = 5913.600 FPS 29803 frames in 5.0 seconds = 5960.600 FPS 28815 frames in 5.0 seconds = 5763.000 FPS 33334 frames in 5.0 seconds = 6666.800 FPS 34156 frames in 5.0 seconds = 6831.200 FPS 25737 frames in 5.0 seconds = 5147.400 FPS 13591 frames in 5.0 seconds = 2718.200 FPS So renice change the speed. So I still wonder what is the distribution of the load between cpu and video chip. Evidently, cpu does a lot of work here :) Best regards Michal -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
