libinput touchscreen maxTouchPoints

[Peter Hutterer]

On Tue, Jan 30, 2018 at 04:18:26PM +1000, Peter Hutterer wrote:
> On Wed, Jan 24, 2018 at 10:47:10AM +0100, Johannes Pointner wrote:
> > On Wed, Jan 24, 2018 at 10:18 AM, Peter Hutterer
> > <peter.hutterer at who-t.net> wrote:
> > > On Wed, Jan 24, 2018 at 10:04:52AM +0100, Johannes Pointner wrote:
> > >> Hello,
> > >>
> > >> I have noticed that since we are using xf86-input-libinput all our
> > >> touchscreens (resistive and pcap) reporting a value of 15 for
> > >> maxTouchPoints.
> > >> After further looking into this I saw that xf86-input-libinput sets
> > >> this value fix to 15. I also checked if it is possible to get this
> > >> info from libinput but I wasn't able to find a possibility.
> > >>
> > >> Is this intentionally?
> > >
> > > mostly, it's largely a matter of not exporting information that doesn't need
> > > to be exported. at least with touchpads libinput does a few tricks and
> > > changes the number of touches available (not yet at runtime, but who knows)
> > > - once we start exporting this we're a lot more limited in what we can do
> > > internally to work around broken devices.
> > >
> > > Main question is though: what do you need this value for?
> > We use this value which is available through the browser
> > (https://developer.mozilla.org/en-US/docs/Web/API/Navigator/maxTouchPoints)
> > to determine if it is a multitouch device or not.
> > The same goes for Qt
> > (http://doc.qt.io/qt-5/qtouchdevice.html#maximumTouchPoints) which has
> > a libinput backend.
> 
> Actually, this is something we could fix in the driver by checking for
> ABS_MT_SLOT and forcing it to 1 when missing. that wouldn't fix Qt though,
> so it may be better to have this exported by libinput after all...
> Need to think of the various special cases here.

for reference, I filed this bug:
https://bugs.freedesktop.org/show_bug.cgi?id=104867

Cheers,
   Peter