On Thu, Jun 09, 2011 at 04:19:14PM +0100, Daniel Stone wrote:
> > a bit of a side-issue with this approach is that the scroll events lose
> > valuator information. we hit that one with the wacom driver where one of my
> > patches started submitting button events with first/num_valuators as 0/0. it
> > is then impossible for the client to get valuator information without
> > tracking motion events (which may be quite costly and I'm not sure all
> > clients do).
> > 
> > so we went back to submitting the current valuator state with _every_ button
> > event. This patch suffers from the same issue. If you scroll, there is no
> > information about the other axes. Which for relative devices doesn't matter
> > much but for absolute devices it certainly does.
> 
> Hmph, so is that just a matter of dropping the current x and y in the
> mask then?

it pretty much means that you have to pass the valuator mask as provided by
the driver into each GPE for emulated scroll events.

Cheers,
  Peter
_______________________________________________
[email protected]: X.Org development
Archives: http://lists.x.org/archives/xorg-devel
Info: http://lists.x.org/mailman/listinfo/xorg-devel

Reply via email to