The OS X tool for Logitech's gaming mouse did not allow button-profiles on the
G500s, but DPI modes could be configured. This is an issue with the software,
not the device itself since the device supports muliple profiles. So the
original assumption that there are devices that have profiles for DPIs but not
buttons is likely wrong, the devices we looked at so far all have profiles
with a set of DPI modes in it, one of which is currently active.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
and relocate the raw value 0 of RATBAG_BUTTON_TYPE_NONE to be below 0x06.
The windows application sets 0x06 when disabling a button.
Signed-off-by: Benjamin Tissoires <benjamin.tissoires@gmail.com>
Problem: when a button is diverted, it stops sending its original button,
which is good, but then reports only on the HID++ long report collection.
That is a little bit embarrassing.
Signed-off-by: Benjamin Tissoires <benjamin.tissoires@gmail.com>
Etekcity mice have volume up/down mapped on the 2 buttons on the right
side of the mouse. They get easily pressed by the hand and can annoy the
user. This command switches on and off these buttons.
Signed-off-by: Benjamin Tissoires <benjamin.tissoires@gmail.com>
This allows to test ratbag_device_set_active_profile() and the
.write_profile() callback from the drivers.
Signed-off-by: Benjamin Tissoires <benjamin.tissoires@gmail.com>
When accessing the buttons of a profile, we should not unref it until
we are done (or there will be some illegal access to a free-ed pointer).
The last profile_unref should be called against the active_profile, not
the last profile.
Signed-off-by: Benjamin Tissoires <benjamin.tissoires@gmail.com>
Naming scheme is now:
* ratbag - the context
* ratbag_device - the device
* ratbag_profile
* ratbag_button
...
Only leftovers of "libratbag" are the header file name, #defines that are
exported through that header file and anything that has to do with the actual
library iself (e.g. the symbol export file).
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Looks like a command name before the current ratbag-command one, and since
this is a static method anyway just usage() will do.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Put the code in place for a multi-command tool, first command is
ratbag-command info /path/to/device
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Same as libinput. Since we need to open device files we need something that's
authorized to open them. So we punt that to the caller.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
We don't need the evdev fd, it's more flexible to start with a udev device.
This changes ratbag-command to take a /sys path as well, but for not it'll
do.
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
Problem is the current implementation requires to actually get the profile
before accessing the buttons. But it may be the simplest and safest way.
Signed-off-by: Benjamin Tissoires <benjamin.tissoires@gmail.com>
