Hi Paul, thanks for the feedback.
With respect to tapping, I'm already running out of hypotheses that can be tested without fine-grained debugging. You might check whether pressure thresholds play a role, but I wouldn't be too optimistic about it. You could clear them as follows: # wsconsctl mouse.tp.param=2:0,3:0 and see whether it helps. This command restores the defaults: # wsconsctl mouse.tp.param=2:25,3:30 As to your observations concerning click-and-drag, I might have thought of that, it concerns some basics: If you want to move the pointer while there are two or more contacts on the touchpad, which one should control the movement? I think that's not totally trivial. Linux gives a simple, but somewhat problematic answer: The "oldest" touch controls the pointer. But if a driver does nothing else about it, this has the effect that the outcomes of your two-finger gestures can depend on the order in which you make the contacts. You can still observe that on some Linux laptops when your are scrolling with two fingers, but leave one finger simply resting on the surface. In your case, it's the firmware, not the driver, that does this work, but the principle seems to be the same. But how are users supposed to know it? It's not evident. The synaptics driver adds up the differences of the coordinates when it receives multi-touch data, and moves the pointer by those "accumulated" values. Older versions applied the method to click-and-drag operations, newer versions to both click-and-drag and two-finger scrolling. The wsmouse driver implements a different approach: when it receives multi-touch data - up to now, from Apple or Elantech-4 models - it assigns pointer-control to touches that are moving (if such a touch exists). Regards, Ulf On 08/07/2017 09:48 AM, Paul de Weerd wrote: > On Mon, Aug 07, 2017 at 01:13:22AM +0200, Ulf Brosziewski wrote: > | On 08/05/2017 11:10 PM, Paul de Weerd wrote: > | > Hi Ulf, > | > > | > On Fri, Aug 04, 2017 at 11:26:12PM +0200, Ulf Brosziewski wrote: > | > | Hi Paul, > | > | > | > | thanks for your help. Does tapping work when you use > | > | the synaptics driver? > | > > | > Nope, it doesn't. > | > > | > | which probably means there is either something happening that > | our hardware driver doesn't cover, or there is a hardware/firmware > | bug. Anyhow, it's strange because the drivers only need very > | basic data to identify a tap: the start of a contact, its > | end, and the duration. Have you checked - with the synaptics > | driver - whether a higher tap timeout helps? If not, would you > | mind to make a short test? Could you increase the tap timeout > | to a very high value, say, two seconds, and test whether a tap > | works (with a slight delay)? For the wsmouse-internal driver, > | the following command will set a two-second timeout: > | # wsconsctl mouse.tp.param=137:2000 > | Of course you could not work reasonably with such a timeout, > | you might want to check then whether something between 200 > | 350 milliseconds would do. The default is 180. > > I tried that: > > [weerd@drop] $ doas wsconsctl mouse.tp.param=137:2000 > mouse.tp.param -> 137:999 > > But I still can't tap. > > | > | > This doesn't work on my touchpad. Also, I can't click-and-drag (never > | > | > worked, in any combination I while playing with the driver settings). > | > > | > This 'click-and-drag' behaviour does work if I click, keep the button > | > depressed and then move that same finger around. [...] > | > | Does this also work if you put a second finger on the touchpad (which > | does nothing)? > > No, but you helped me find something useful / interesting: > > 1. click button (keep depressed) > 2. second finger doesn't do anything > > however > > 1. touch 'second finger' first > 2. click button (keep depressed); now the 2nd finger can drag (1st > finger doesn't work for dragging now) > > This in addition to > > 1. click button (keep depressed) > 2. move 1st finger around on small clickable area to drag > > So, with this new approach, I have more surface area to drag around > and I can release my second finger to scroll beyond the end of the > touchpad. Never knew this was an option, I feel like an idiot now :) > > Thanks! > > Paul >
