SUMMARY Keyboard repeat is broken. Holding a key does not repeat it like it should. STEPS TO REPRODUCE 1. Open a wayland session 2. open a terminal 3. press and hold a key OBSERVED RESULT Only one key press is recorded and printed in the terminal EXPECTED RESULT as long as I hold it should repeat the key SOFTWARE/OS VERSIONS Linux/KDE Plasma: master (available in About System) KDE Plasma Version: master KDE Frameworks Version: master Qt Version: 5.15.2 ADDITIONAL INFORMATION It has to do with these changes: https://invent.kde.org/plasma/kwayland-server/-/commit/8a6d17f2cbfadee783d1664ec7e59e521075cfd5 https://invent.kde.org/plasma/kwin/-/commit/5e037570fe90d9883e887bb8f0e78d5bef8ec2f6 Reverting those two fixes the issue.
A possibly relevant merge request was started @ https://invent.kde.org/plasma/kwin/-/merge_requests/372
Git commit 733692f4391ebf82f0f9a44c1a9aadab95240e43 by Bhushan Shah. Committed on 21/10/2020 at 10:31. Pushed by bshah into branch 'master'. input: reconfigure after setting up the libinput Previously seat interface was keeping track of the repeat information, but now it is taken care by KeyboardInterface, which is created only if the setHasKeyboard is set to true. So, since we are setting up various capabilities on seat in libinput, re-parse configuration here, so that various information like key repeat is notified to kwayland-server M +1 -1 input.cpp https://invent.kde.org/plasma/kwin/commit/733692f4391ebf82f0f9a44c1a9aadab95240e43
Is it appropriate to backport this to 5.20?
>Is it appropriate to backport this to 5.20? It is not needed.
(In reply to Nate Graham from comment #3) > Is it appropriate to backport this to 5.20? No, the regression was from master branch, so only needed for master branch