On Mon, 15 May 2023, Lucio Chiappetti wrote:

I'll be reading the libinput synaptics and synclient manpages.

I've done this exercise and some tests

ELAN076D: 00 04F3:3244 Mouse               id=10   [slave  pointer  (2)]
ELAN076D: 00 04F3:3244 Touchpad            id=11   [slave  pointer  (2)]

(not sure why it reports twice the same device as touchpad and mouse, there is no physical mouse attached)

I've no idea, I suppose they are both loaded in case a mouse is attached.

 Synaptics driver and libinput (probably the correct, "modern" option).

If I do xinput --list-props for device 11 (touchpad) it returns Synaptics and for device 10 (Mouse) it returns libinput.

So I am assuming I have to use synclient to get my status.

The current one has just the touchpad, but the lower part of it is "depressable".

I believe this is what is technically called a "clickpad".

The synaptics man page says that in principle one could set EmulateMidButtonTime to non-zero to do MB2 with simultaneous MB1-MB3 (as I was used on the old touchpad-with-buttons) but ClickPads do not support middle mouse button emulation ... so I gabe up on that.

However looking at the synclient output I discovered parameters

    RTCornerButton          = 2
    RBCornerButton          = 3
    LTCornerButton          = 0
    LBCornerButton          = 0

so this mean that TAPPING (not pressing-releasing as for the lower area) in the right-top corner of the touchpad is MB2.

Since I found somehmoiw difficult to locate the exact area of such corner, and the tap somehow "hard" I decided to ...

... simply SWAP the actions for MB2 and MB3 with

xinput set-button-map 11 1 3 2 4 5 6 7 8 9 10 11 12

This way I can click in the lower right corner to do MB2 (which I use often in any window to do quick paste, and more seldom in the root window to call IconMan), while I tap in the upper right corner to do MB3 (which I use seldom, I can think only of a browser window to get a context dependent menu)

I'll keep this along with PointerKey assignment to right-control (which works except in xterms). If I'm satisfied I'll make it permanent.

Thanks in oarticular to Mandar Mitra for addressing me to synaptics and xinput.

--
Lucio Chiappetti

Reply via email to