Libinput 1.6 Released With New Touchpad Acceleration

[darkbasic]

On Arch, why isn't 1.6 in testing? I can only find 1.5 here. What's going on with Arch these days? We were behind Fedora and Suse Tumbleweed getting x-org-server 1.9. And now looks like we're going to have to wait to try out this fix for our trackpads.

Agree, core/linux is still at 4.8 while mainline already released 4.9.4. What's happening?? Edit: just read the post before.

[theghost]

As for Kernel 4.9... maybe the reports about unbootable systems are the reason for leaving it in testing. Just look at the bugtracker.

Right, looked at the kernel 4.9 bugs for i686 and wondered why there were no reports. x86_64 explains all :-)

[nanonyme]

To be fair, Fedora's Xorg adoption policy is not always welcome, for example they switched to Xorg 1.18 before it had nVidia proprietary driver support, which initiated some flaming from users. Remember that Arch is bleeding edge, rolling realease, but it is by no means meant to be unstable. Also you are free to package it yourself and contribute ;-)

To be fair, Fedora policy is proprietary packages cannot be part of Fedora, hence RPMFusion. Since they are not part of Fedora, they cannot be blockers

[andrebrait]

That doesnt sound like an acceleration issue but instead your trackpad is reporting an incorrect resolution.
That can only be addressed through bug reporting and updates to the libinput tweaks file.

Actually, it is exactly an acceleration issue. For one, 1.6 fixes it in my laptop, which reports the cprrect dimensions to the driver, and second, Peter explained how these things could be a result of the bad acceleration algorithms in the pre-1.6 libinput releases. Only issue left for me with libinput is regarding tap misdetection, and I already filled a bug report and provided recordings, etc.

[liam]

Actually, it is exactly an acceleration issue. For one, 1.6 fixes it in my laptop, which reports the cprrect dimensions to the driver, and second, Peter explained how these things could be a result of the bad acceleration algorithms in the pre-1.6 libinput releases. Only issue left for me with libinput is regarding tap misdetection, and I already filled a bug report and provided recordings, etc.

He did? I don't know how the accel algorithm could go that badly for you without the trackpad reporting bad info. If it was just the algorithm then everyone whose trackpad was detected correctly would have similar problems.
If you've got a link to a bugreport, or something by Peter that explains how this could be I'd appreciate it.

[andrebrait]

He did? I don't know how the accel algorithm could go that badly for you without the trackpad reporting bad info. If it was just the algorithm then everyone whose trackpad was detected correctly would have similar problems.
If you've got a link to a bugreport, or something by Peter that explains how this could be I'd appreciate it.

Is his blog entry ok? In it he aknowldged the awfulness of the old acceleration code and how imprecise it could make things and describe how he much improved it for 1.6.0. Anyway, like I said, my touchpad reported correct dimensions and was still a mess before 1.6.0.

http://who-t.blogspot.com.br/2016/12/libinput-touchpad-pointer-acceleration.html?m=1

[liam]

Is his blog entry ok? In it he aknowldged the awfulness of the old acceleration code and how imprecise it could make things and describe how he much improved it for 1.6.0. Anyway, like I said, my touchpad reported correct dimensions and was still a mess before 1.6.0.

http://who-t.blogspot.com.br/2016/12...ation.html?m=1

Hi Andre, thanks for the link.
Ok, I'd read that post some back (you can see my comment), but, assuming libinput has the correct dimensions and resolution for the to, it doesn't explain why some trackpads work fine and others (notably elantech) don't. It simply says that the acceleration curves that were the same for everyone we suboptimal. That's fine, but I don't know why that would cause a working trackpad to accelerate to maximum at the slightest movement.
The problem may simply be that we have different expectations regarding trackpad behavior and thus I just don't notice these issues.