Opened 4 days ago
Closed 37 hours ago
#19238 closed bug (fixed)
Holding a key causes mouse button presses to be dropped
Reported by: | Hanicef | Owned by: | nobody |
---|---|---|---|
Priority: | normal | Milestone: | R1/beta6 |
Component: | Add-Ons/Input Filters | Version: | R1/Development |
Keywords: | Cc: | ||
Blocked By: | Blocking: | ||
Platform: | All |
Description
If a key is held down, mouse button presses is not always registered in the program. You can try this on the clock program, where clicking normally changes the clock theme as usual, but holding down the f key on the keyboard while clicking the clock causes it to only change themes seemingly at random.
Note that the clock is not the only program affected by this, all programs seem to exhibit the same weird behavior.
Change History (10)
comment:1 by , 4 days ago
comment:2 by , 4 days ago
Maybe this has something to do with key repeat? I tried timing it relative to a text box, and the amount of time it takes for mouse input to go unresponsive is about the same as the amount of time it takes for key repeat to kick in.
comment:4 by , 4 days ago
Component: | Servers/input_server → Add-Ons/Input Filters |
---|
comment:6 by , 3 days ago
No, but I don't think it's reasonable for mouse input to block, then pass, then block again, touchpad or no.
comment:7 by , 3 days ago
If you don't have a touchpad, then really PadBlocker shouldn't be active at all, so that's a bug here.
comment:8 by , 3 days ago
Fair enough. I still consider key repeats blocking input to be a bug, but I can try to look into disabling PadBlocker for everything except touchpads and send in another patch that fixes that.
comment:9 by , 3 days ago
FYI, got a patch for preventing non-trackpad devices from being blocked: https://review.haiku-os.org/c/haiku/+/8557
comment:10 by , 37 hours ago
Milestone: | Unscheduled → R1/beta6 |
---|---|
Resolution: | → fixed |
Status: | new → closed |
Fixed in hrev58319.
add-ons/input_server/filters/padblocker
is the culprit. A similar issue is reported as #17821. As a workaround, you can blocklist the padblocker...