[RFC XI 2.1] current issues

[Peter Hutterer]

On Tue, Dec 07, 2010 at 11:46:55PM +0000, Daniel Stone wrote:
> Hi,
> 
> On Tue, Dec 07, 2010 at 10:44:12PM +0100, Denis Dzyubenko wrote:
> > It would be nice to expose which of the touch sequences result in the
> > mouse event.
> > 
> > With dependent touch it is pretty clear - if there is only one touch,
> > that's the one that triggers the consequtive mouse events, and if
> > there is more than one, there will be no mouse movements.
> > 
> > However with direct touch this might not be the case and the mouse
> > cursor might be moving even if you move two fingers on the screen. I
> > would like to be able to know which one of those touch sequences
> > trigger mouse events.
> 
> That's still a TODO for now; in either case, I'm very loathe to put it
> in the spec, and would like to have the emulation pointer-dependent.
> The resulting pointer events will have the TouchPointer (although this
> will probably be renamed to something like PointerFromTouch to avoid
> ambiguity with touch event flags) flag set, indicating that they came
> from a touch event.  Would you like the touch events to have a flag set
> indicating that they are also generating pointer events?
> 
> To be honest, I haven't really thought too much about pointer emulation
> - Chase or Peter might have some more thoughts on this.

so far our assumption was that if you're capable of handling touch, you
should just ignore pointer events generated by touches. It'd be easy enough
to add a flag to touch events to indicate they're generating pointer events
but that still doesn't tell you which touch event has generated which
pointer event. which is the case if mouse + touchscreen usage starts
overlapping.

Cheers,
  Peter