Repository navigation
omamobile-shell: the power-key flash test no longer races its own stand-in - #889
Merged
Merged
Conversation
…nd-in The stand-in for omarchy-brightness-display turned the fake panel off and then wrote "panel-off" to the call log. omamobile-keys polls the panel files every 50 ms, so on a loaded machine it could see the panel off and log "brightnessctl ... set 2083" before the stand-in's own line landed, and the test failed with "the saved brightness goes back only after the panel is off" on a PR that touched nothing in this package (CI run 44324012922, 2026-11-18). The stand-in now writes its line just before it flips the files. With a 0.2 s pause put between the two steps the old order fails every time with that message and the new one passes. The three fixed sleeps after a press in this block (1.5 s, 1 s, 1 s) are now waits for the line the press ends with, up to 10 s, so a slow machine is waited for and a fast one loses no time. Test only: nothing the package installs changes, so no version bump. Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
…nd-in again The previous commit turned this case's fixed 1 s sleep into a wait for the brightness line, which Hyprland's word produces at once. The stand-in's delayed half (0.4 s) was then still pending when the next case put the fake panel files back, and it turned that new panel off: with the power key as Android's, "locked and lit: a press turns the screen off" then saw a dark panel and woke it instead (CI run 44333336393). The sleep is back, before the check. Co-Authored-By: Claude Fable 5.2 <noreply@anthropic.com>
Owner
Author
|
The first push failed CI on a different line ("locked and lit: a press turns the screen off", with the power key as Android's), and that one was this PR's own doing: turning the no-panel-files case's 1 s sleep into a wait let the stand-in's delayed half fire after the next case had put the fake panel files back, so that panel went dark behind the test. The sleep is back in that one case. Both key modes pass here four times each, and with the 0.2 s gap widened. |
This was referenced Nov 18, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The power-key test in
packages/omamobile-shell/test/keys.shfailed CI on #850, which touches nothing in this package (run 44324012922): "flash: the saved brightness goes back only after the panel is off".Cause: the stand-in for
omarchy-brightness-displayturned the fake panel off and then wrotepanel-offto the call log.omamobile-keyspolls the panel files every 50 ms, so on a loaded runner it could see the panel off and log its own lines before the stand-in's line landed. The failed run's log endsomamobile-sleepd now, thenpanel-off.Fix: the stand-in writes its line just before it flips the files, and the three fixed sleeps after a press in this block become waits for the line the press ends with (up to 10 s).
Checked: with a 0.2 s pause put between the stand-in's two steps, the old order fails every time with the CI message and the new order passes. Six runs each under 42 busy loops passed for both, so plain load didn't reproduce it here; the widened gap did.
Test only: nothing the package installs changes, so no version bump.
🤖 Generated with Claude Code