Testers wanted: Synaptics acceleration patch

[Simon Thum]

Hi,

this patch fixes the long-standing issue of synaptics being accelerated
twice, once by itself and once by the dix, based on different settings.
This typically makes synaptics feel awkward, since one needs a careful
setup to avoid the issue.

This patch fixes that by taking most of acceleration out of the driver,
utilizing the accel framework in dix. It should behave quite the same,
except that the double-accel is now missing, so you may need to adapt to
that.

Unfortunately, I don't have matching a HW+SW to test at the moment. A
previous version of that patch did work though. So I ask for volunteers,
and the maintainers to pick it up when feedback is positive.

Prerequisites:
X.Org Server 1.7.0+ or git master
xf86-input-synaptics from git (1.2.0 may be fine too) plus patch


What to look out for:

If you are fine with your synaptics, and it stays like that after issuing

  xset m 1 1

(or selecting the none (-1) accel profile using xinput, for that matter)
then this patch should not change much for you, because your setup is
already good. A small change in behaviour is acceptable, and should be
for the better. In particular, pressure-sensitive acceleration should
remain working, but AFAIK that feature isn't on by default.

If the xset m ... changed synaptics behaviour, then the patch should
change it likewise. You can readjust as usual. In the end, it should
behave a bit more predictable.

Anyone who feels qualified, please test it and give feedback!

-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: 0001-Setup-pointer-acceleration-for-synaptics.patch
URL: <http://lists.x.org/archives/xorg/attachments/20100116/1c72fabe/attachment.ksh>