https://bugs.freedesktop.org/show_bug.cgi?id=106457
--- Comment #6 from Peter Hutterer <peter.hutte...@who-t.net> ---
(In reply to fd from comment #5)
> Well, Now I run evemu-record and I can confirm that
> # EV_KEY / BTN_TOOL_PEN 1
> on stylus proximity, but get nothing when it leaves proximity.
>
> After edit /lib/udev/hwdb.d/90-libinput-model-quirks.hwdb i get the # EV_KEY
> / BTN_TOOL_PEN 0 entry but not always.
in the evemu recording log? we don't write that event to the device node, so
you shouldn't see that event except in the libinput debug log. If the event
comes up in the evemu log then we're dealing with a different issue here.
> Each time that get BTN_TOOL_PEN 0, then touchscreen gestures works. but
> each time that proximity remain enabled, the finger gestures are disabled.
yeah, that's expected. when you're using the pen, you're likely to touch the
screen and cause fake pointer motion. This is called touch arbitration and an
explicit feature (used to be done in the kernel by some drivers). But that
obviously requires a working proximity out.
> I get only 5 successfully
> BTN_TOOL_PEN 0
> in 100 trials. So I can't confirm if the edition of
> /lib/udev/hwdb.d/90-libinput-model-quirks.hwdb is related.
does the property show up in the udevadm info output and did you restart
libinput, like the link suggested? libinput should (internally) fake the event
after 50ms of inactivity, i.e. no events from the pen device.
> sorry, for my english.
meh, neither of us is a native english speaker, so...
--
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
wayland-bugs mailing list
wayland-bugs@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/wayland-bugs
