[PATCH libinput 00/14] Lid switch interface

[Peter Hutterer]

The first four are James' patches, I think the only change I did to those
was force ID_INPUT_SWITCH in the udev rule (until systemd is ready) and a
removal of a double empty line.

Patches 5 and 11 are cleanup, I should've done those first but...

The others make sure that:
* lid switch is synchronised on startup, i.e. we get a "lid is closed" event
  if that is the case on start.
* clarify that a switch's neutral state is "off" and that we don't send
  events for that on startup. possibly controversial, but also sensible IMO
* to monitor keyboard events to reactivate if we get key events. This can
  happen if the lid switch state isn't reliable.

The default behaviour now is:
- if the lid switch is deemed reliable (on all laptops right now, see patch
  14) we sync it at startup
- otherwise, the lid switch is *not* synced on startup but will work as
  expected once we see the first close event. This conveniently filters all
  devices that pretend their switch is always on

Cheers,
  Peter