Here's what I run.
Geforce, Nvidia 0.9-4, Xfree86 4.0.1
2.2.16 kernel w/ 2.4 USB patch so I can enable my Wingman Gaming mouse.
Now, there are two basic ways I can run the Mouse in Quake3 Arena.. DGA on or off. With DGA off, the input is extremely smooth, however there is an ingrained, severe acceleration which causes you to do about 2 360 degree turns (even with MOUSE_ACCEL_TYPE OFF and xset m 0 0).
Enabling DGA makes the sensitivity and degree of control feel just like windows, with one major exception: occasional choppy mouse performance. I can sit there holding down the right arrow key for hours and have it turn perfectly smooth, but when I move the mouse left to right and back again, it will occasionally stutter. I have this problem whether I use the PS/2 connector or USB. This problem occurs with either m_filter on or off (which most serious gamers leave off to give a more precise mouse movement, albeit slightly choppier).
This leads me to beleive that the problem is a result of some flaw in DGA 2.0, as the normal "pointer grabbing" mouse input technique gives smooth input but lower response time.
Slashdot Mirror
Here's what I run. Geforce, Nvidia 0.9-4, Xfree86 4.0.1 2.2.16 kernel w/ 2.4 USB patch so I can enable my Wingman Gaming mouse. Now, there are two basic ways I can run the Mouse in Quake3 Arena.. DGA on or off. With DGA off, the input is extremely smooth, however there is an ingrained, severe acceleration which causes you to do about 2 360 degree turns (even with MOUSE_ACCEL_TYPE OFF and xset m 0 0). Enabling DGA makes the sensitivity and degree of control feel just like windows, with one major exception: occasional choppy mouse performance. I can sit there holding down the right arrow key for hours and have it turn perfectly smooth, but when I move the mouse left to right and back again, it will occasionally stutter. I have this problem whether I use the PS/2 connector or USB. This problem occurs with either m_filter on or off (which most serious gamers leave off to give a more precise mouse movement, albeit slightly choppier). This leads me to beleive that the problem is a result of some flaw in DGA 2.0, as the normal "pointer grabbing" mouse input technique gives smooth input but lower response time.