Join GitHub today
GitHub is home to over 28 million developers working together to host and review code, manage projects, and build software together.
Sign upone shot modifier event occurs after the key to be modified in libinput-1.9.3 #2081
Comments
This comment has been minimized.
Show comment
Hide comment
This comment has been minimized.
pushcx
Dec 4, 2017
I also have this bug on a Planck using a custom layout - same behavior that keys on the same row as the one-shot key work but not others. Happens with both OSM and OSL. Downgrading to Arch's libinput 1.9.2 worked around it for me.
The most likely related lines in the release notes are about debouncing, which has three threads: 1 2 3. If this is libinput's bug, it has not been reported on their tracker.
Hope this helps. I'd try to run this down further, but I know basically nothing of qmk or libinput and my C is very rusty.
pushcx
commented
Dec 4, 2017
•
|
I also have this bug on a Planck using a custom layout - same behavior that keys on the same row as the one-shot key work but not others. Happens with both The most likely related lines in the release notes are about debouncing, which has three threads: 1 2 3. If this is libinput's bug, it has not been reported on their tracker. Hope this helps. I'd try to run this down further, but I know basically nothing of qmk or libinput and my C is very rusty. |
This comment has been minimized.
Show comment
Hide comment
This comment has been minimized.
drashna
Mar 25, 2018
Collaborator
This should work in libinput 1.9.4-1 now. Closing.
If it doesn't, please reopen
|
This should work in If it doesn't, please reopen |
tim-tx commentedNov 30, 2017
I recently upgraded my system to libinput-1.9.3 from libinput-1.9.2. My one shot shift key had worked as expected before the upgrade, but now only works for the bottom row (ZXCVBNM... in qwerty). Hitting the one shot shift and a key on one of the other rows just produces a lowercase character. Xev shows that the shift modifier actually registers after second key for these other rows, but in the correct order for the bottom row. The log is attached, I hit n, m, h, j, OSM(MOD_LSFT) + n, osm + m, osm + h, osm + j. You see the right order for osm + {n,m}, but shift registering second for osm + {h,j}. Downgrading to libinput-1.9.2 fixed the issue. This is on an Ergodox EZ, my layout is here.
xev.log