The deprecated keys are kept in isolated ".deprecated" suffixed schemas.
On plugin startup, the settings-daemon paths are opened with these legacy
schemas, and all user-modified keys are dumped to the new location and
reset.

This ensures the transition just happens once per-key, and ensures the
dconf database is left clean.

https://bugzilla.gnome.org/show_bug.cgi?id=742593

Most settings have been moved there, so make this plugin use the new
schemas for the handled settings, and remain using the settings-daemon
schema for settings that can just be honored on X11.

In order to simplify the situation for the future, the management of
"middle-button-enabled", "horiz-scroll-enabled" and "disable-while-typing"
settings have been removed, and the sane default is applied on non-libinput
devices on X11.

For all other settings, the mouse plugin has been made to ignore devices
created through the libinput driver, as mutter handles these.

Additionally, the entire plugin is disabled on Wayland, the settings
remaining in g-s-d schemas are not a thing there.

https://bugzilla.gnome.org/show_bug.cgi?id=742593

These are not a windowing/desktop thing on wayland, so just ignore these
there at the moment.

https://bugzilla.gnome.org/show_bug.cgi?id=742593

With patches from bug #739397 applied. This is managed by mutter on X11
too, so no need to do it here again.

https://bugzilla.gnome.org/show_bug.cgi?id=742593

This function pokes a libinput-driver-specific property, and can be used to
find out whether a device uses the libinput driver or not.

This will be used to determine whether a device is to be managed by mutter,
or by g-s-d itself.

https://bugzilla.gnome.org/show_bug.cgi?id=742593

This is part now of org.gnome.desktop.peripherals.[tablet|touchscreen],
in use now by g-s-d.

https://bugzilla.gnome.org/show_bug.cgi?id=742593

The direct usage of GsdDeviceMapper allows us to remove some code here.

https://bugzilla.gnome.org/show_bug.cgi?id=742593

This has brought some changes in the xrandr and wacom modules, which now
use GsdDevice for device mapping.

The use of GdkDevices and XInput properties remain for these settings not
yet in the gsettings-desktop-schemas peripherals settings in the wacom case,
and those bits have been made to run only on X11 as expected.

https://bugzilla.gnome.org/show_bug.cgi?id=742593

This should be eventually done by the compositor. In the mean time, restrict
it to X11, and remove the GsdDeviceMapper usage, as it's not strictly
necessary.

https://bugzilla.gnome.org/show_bug.cgi?id=742593

This will be used on X11 environments, poking device capabilities
through device properties. This will be the default backend on
X11.

https://bugzilla.gnome.org/show_bug.cgi?id=742593

This will be used on wayland environments in order to figure out
the plugged devices and their capabilities.

https://bugzilla.gnome.org/show_bug.cgi?id=742593

This is a GUdev-based object that keeps track of the plugged in input
devices. Meant to be used across g-s-d wherever some knowledge about these
must remain.

On X11, some API is still offered to access the gdk/X devices (which may not
map 1:1 with the udev devices).

For both platforms, the relevant /org/gnome/desktop/peripherals/... settings
can be accessed, in order to interoperate with the settings that mutter
applies on libinput devices.

https://bugzilla.gnome.org/show_bug.cgi?id=742593

gnome-shell does not allow to grab any particular key combo more
than once, even when the corresponding keybinding modes don't
overlap, so the FOO_KEY_NO_DIALOG shortcuts introduced in commit
b7827c13 never actually worked.
Instead, use the new 'action-mode' parameter that was added to the
'Activated' signal to determine whether to a power action should
be interactive or not.

https://bugzilla.gnome.org/show_bug.cgi?id=741762

This will allow GTK+ to detect ssh forwarded display situations
where app menus won't work, and handle it better.

https://bugzilla.gnome.org/show_bug.cgi?id=741740

The AcceleratorActivated signal now uses a dictionary for additional
parameters to make future additions easier.

The type was renamed in gnome-shell to account for usage besides
keybindings (namely gestures), so do the same here to keep the
headers in sync.

Signed-off-by: Trần Ngọc Quân <vnwildman@gmail.com>

Otherwise we risk being called and not have all out internal state
ready (e.g. priv->rw_screen being NULL) leading to crashes.

We also need to disown our well known name on stop() since we'll
reaquire it if start() gets called again.

https://bugzilla.gnome.org/show_bug.cgi?id=740151

They won't leak if the plugin is stopped and started again but they do
if we reach finalize() which isn't really a problem but there's no
reason to keep them around anyway.

https://bugzilla.gnome.org/show_bug.cgi?id=740151

This is currently used by the brightness change OSD.

https://bugzilla.gnome.org/show_bug.cgi?id=704086

When StepUp() or StepDown() are called to change the screen brightness,
also return the output id where the brightness transition happened (if
any).
This is useful so that the shell can show the OSD only on the relevant
monitor, which will be done in the next commit.

https://bugzilla.gnome.org/show_bug.cgi?id=704086

We're going to use this later.

https://bugzilla.gnome.org/show_bug.cgi?id=704086

Outputs being connected/disconnected should update the input devices that
are mapped to these through config, so they are remapped/unmapped correctly
after the change.

https://bugzilla.gnome.org/show_bug.cgi?id=739891

This ensures that either an output is guessed for the device again, or for
display-less tablets, the default matrix spanning full desktop extents is
reinstaurated.

https://bugzilla.gnome.org/show_bug.cgi?id=739891

The code that handled this when creating a GsdInputInfo and whenever
the settings changed were almost the same, put it in a common place.

https://bugzilla.gnome.org/show_bug.cgi?id=739891

If the monitor reports a width/height that looks suspiciously like an
aspect ratio (16/9 or 16/10) don't check for hi-dpi. We can assume that
makers of devices that do support hi-dpi aren't so careless.

See http://cgit.freedesktop.org/~daniels/xserver/commit/?h=lodpi

https://bugzilla.gnome.org/show_bug.cgi?id=738748

For example, turning off a tablet's touchscreen when the screenlock
kicks in.

Note that this is currently broken due to bugs in the kernel.
For the X.org tracking bug, see
https://bugs.freedesktop.org/show_bug.cgi?id=81781

And a possible patch exists at:
https://lkml.org/lkml/2014/9/30/588

https://bugzilla.gnome.org/show_bug.cgi?id=720295

If the input device isn't known to the device mapper (say a keyboard for
example), we'd crash trying to check which output it was attached to.

https://bugzilla.gnome.org/show_bug.cgi?id=720295

Some trackballs don't have a separate scroll wheel, so one of the
buttons needs to be repurposed so pressing it will turn the main ball
into a wheel.

https://bugzilla.gnome.org/show_bug.cgi?id=645666