Closed vizv closed 2 years ago
NACK, if you have both a mouse and a touch, there is the possibility that your patch picks the mouse as input device instead of the touch.
And even if there is only a mouse, it is possible to "calibrate" the mouse with very funny results :O
You are right when you say that the logic behind the selection of the device by id or by name should not be constrained by the devices[].type field.
So I think that we should consider the following cases:
1) no --device-id or --device-name If there is only one eligible device (a xi_touchscreen one), pick this one Otherwise throw an error, and let the user to explicitly select a device
2) --device-id passed Pick the device that matches the passed device-id ignoring if it is a xi_touchscreen or not
2) --device-name passed Pick the device that matches the passed device-name ignoring if it is a xi_touchscreen or not
I have to point out that with the v0.9, if the "libinput Calibration Matrix" doesn't exist, the "Coordinate Transformation Matrix" is used.
Please try my devel branch. Now it is (should be) possible to select any device using --device-id and/or --device-name.
If this solve please let me know, so I will update the master branch
That works, thanks
I got a eGalax USB Touchscreen identified as mouse but has
libinput Calibration Matrix
property, it works with--device-id
mentioned in #2 but not--device-name
which gives "Unable to find Device" error.This PR fixed such issue, however it's nice to do a proper check for the existence of
libinput Calibration Matrix
property.