tag:blogger.com,1999:blog-6112936277054198647.post5351998218944775867..comments2024-03-12T00:42:06.642+10:00Comments on Who-T: libinput touchpad pointer acceleration analysisPeter Huttererhttp://www.blogger.com/profile/17204066043271384535noreply@blogger.comBlogger16125tag:blogger.com,1999:blog-6112936277054198647.post-63481447312634281902017-03-17T15:55:17.202+10:002017-03-17T15:55:17.202+10:00From the perspective of someone coming from Chrome...From the perspective of someone coming from Chromebooks -- I haven't had a laptop running Linux in several years:<br /><br />Picked up a ThinkPad X1 Carbon 4th Gen recently, and installed Antergos with Gnome. Compared to my old Toshiba Chromebook 2 (Swanky), the cursor feels extremely squirrely -- with the biggest issue being that I overshoot targets every single time. On my Chromebook, cursor comes to a screeching halt exactly where I want it, after both long and short movements. My ThinkPad's touchpad feels pretty horrible by comparison, to be frank, and I've been using it for about a week now. (As a separate matter, palm detection doesn't seem to actually work).<br /><br />I know this isn't the place for bug reports, just thought I would offer my impressions coming from Chrome OS -- not a comparison to previous versions of libinput. It's almost maddening.<br /><br />I wish I could use the Chrome OS input stack. PDXWebhttps://www.blogger.com/profile/06131932135264826321noreply@blogger.comtag:blogger.com,1999:blog-6112936277054198647.post-85126206694890819162017-01-19T08:21:06.287+10:002017-01-19T08:21:06.287+10:00Thanks a lot! Thanks a lot! André Braithttps://www.blogger.com/profile/00080795097898444484noreply@blogger.comtag:blogger.com,1999:blog-6112936277054198647.post-79674251612379720282017-01-19T07:04:47.984+10:002017-01-19T07:04:47.984+10:00https://wayland.freedesktop.org/libinput/doc/lates...https://wayland.freedesktop.org/libinput/doc/latest/reporting_bugs.htmlPeter Huttererhttps://www.blogger.com/profile/17204066043271384535noreply@blogger.comtag:blogger.com,1999:blog-6112936277054198647.post-51095878525387943002017-01-19T00:18:10.177+10:002017-01-19T00:18:10.177+10:00Where should one report bugs related to libinput?Where should one report bugs related to libinput?André Braithttps://www.blogger.com/profile/00080795097898444484noreply@blogger.comtag:blogger.com,1999:blog-6112936277054198647.post-83494154839084337852017-01-16T13:01:24.290+10:002017-01-16T13:01:24.290+10:00MatMaul: file a bug please, this blog's format...MatMaul: file a bug please, this blog's format isn't useful when it comes to debugging things.Peter Huttererhttps://www.blogger.com/profile/17204066043271384535noreply@blogger.comtag:blogger.com,1999:blog-6112936277054198647.post-3932402952309643562017-01-13T20:51:31.743+10:002017-01-13T20:51:31.743+10:00Hi Peter, I installed master on my Fedora 25 to te...Hi Peter, I installed master on my Fedora 25 to test the new code. Seems fine, but I find the max speed too slow for me. While it was ok before with the gnome setting at around 80%, I am now struggling even at 100%. Is it possible to change a multiplier somewhere to be able to reach higher speed ? Thanks !MatMaulhttps://www.blogger.com/profile/00781874457960516859noreply@blogger.comtag:blogger.com,1999:blog-6112936277054198647.post-36145939372574332952017-01-09T11:44:15.527+10:002017-01-09T11:44:15.527+10:00Shark: probably not until I can cut a 1.6 RC, it&#...Shark: probably not until I can cut a 1.6 RC, it's too many patches to pull in, sorry. But at least that one shouldn't be too far off.Peter Huttererhttps://www.blogger.com/profile/17204066043271384535noreply@blogger.comtag:blogger.com,1999:blog-6112936277054198647.post-83737594000217592682017-01-07T21:11:25.256+10:002017-01-07T21:11:25.256+10:00Hi there, Peter! Nice work! Any expectations as to...Hi there, Peter! Nice work! Any expectations as to when it might land in Fedora's repos?André Braithttps://www.blogger.com/profile/00080795097898444484noreply@blogger.comtag:blogger.com,1999:blog-6112936277054198647.post-23369921767121988532017-01-01T20:19:41.974+10:002017-01-01T20:19:41.974+10:00reading the gnuplot docs and your sourcecode right...reading the gnuplot docs and your sourcecode right now. thanks ledge. Ninehttps://www.blogger.com/profile/00259687923444285464noreply@blogger.comtag:blogger.com,1999:blog-6112936277054198647.post-31897028890527388452017-01-01T19:17:37.504+10:002017-01-01T19:17:37.504+10:00Nine: they're just gnuplot graphs. I used prin...Nine: they're just gnuplot graphs. I used printfs in the libinput code, then a python script to convert from the raw data into the bits I needed. The python script writes out a bunch of gnuplot files (plus matching data files). the source to reproduce is here https://github.com/whot/libinput/commits/wip/touchpad-pointer-accel-v3Peter Huttererhttps://www.blogger.com/profile/17204066043271384535noreply@blogger.comtag:blogger.com,1999:blog-6112936277054198647.post-20684720885335483682017-01-01T14:21:04.541+10:002017-01-01T14:21:04.541+10:00I've been messing around with the acceleromete...I've been messing around with the accelerometers on my phone in sailfish OS, I was wondering what tool you used to render your graphs, and what your scripts look like, because it looks heaps cleaner and smarter than what I'm doing (printfs and wild guesses).Ninehttps://www.blogger.com/profile/00259687923444285464noreply@blogger.comtag:blogger.com,1999:blog-6112936277054198647.post-41610331871756902812016-12-30T09:47:58.405+10:002016-12-30T09:47:58.405+10:00Hugh: I'm not treating it as completely invari...Hugh: I'm not treating it as completely invariant but you'll find that when you move your finger on a touchpad, you have a fairly clear idea of what's supposed to happen. If that doesn't happen then you get higher cognitive load because now you have to correct for it. This is particularly true for movements with a pre-selected target, e.g. an icon close to the pointer.<br /><br />Personally, I also notice this a lot more on a touchpad than a mouse, the feedback loop on a mouse seems to be tighter than on a touchpad. But that's based on a sample size of 1 (i.e. me :)Peter Huttererhttps://www.blogger.com/profile/17204066043271384535noreply@blogger.comtag:blogger.com,1999:blog-6112936277054198647.post-75198200503745107332016-12-28T00:05:25.974+10:002016-12-28T00:05:25.974+10:00Thanks very much for working on this.
I'm a l...Thanks very much for working on this.<br /><br />I'm a little confused. Surely a human will adjust their touchpad action based on screen pointer feedback from the system. How can you treat touchpad sequences as invariant even when the feedback is being changed?<br /><br />(I recently read a book about Doug Englebart (the inventor of the mouse). His goal was to have users and systems "co-evolve": each changing and improving together. In the case of the touchpad, users will evolve to handle whatever the touchpad software does.)<br /><br />Perhaps the first requirement is consistency. I've sometimes felt touchpad response a little inconsistent on my Linux systems, especially on ones that are overloaded. That problem might be in Firefox rather than libinput.Hugh Redelmeierhttps://www.blogger.com/profile/04906181582604638650noreply@blogger.comtag:blogger.com,1999:blog-6112936277054198647.post-8849232321918378212016-12-21T23:13:50.898+10:002016-12-21T23:13:50.898+10:00Touchpad on my Dell Vostro 3360 is now much more u...Touchpad on my Dell Vostro 3360 is now much more usable. Thank you!Artihttps://www.blogger.com/profile/13985274277426287809noreply@blogger.comtag:blogger.com,1999:blog-6112936277054198647.post-48931358342356549182016-12-20T23:18:28.683+10:002016-12-20T23:18:28.683+10:00AwesomeAwesomeLapohttps://www.blogger.com/profile/17105228766270388766noreply@blogger.comtag:blogger.com,1999:blog-6112936277054198647.post-25757209459923119092016-12-20T20:53:43.323+10:002016-12-20T20:53:43.323+10:00That's a really huge improvement for my thouch...That's a really huge improvement for my thouchpad, thank you very much!Shibahttps://www.blogger.com/profile/03119834443247933506noreply@blogger.com