You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
first_frame test inactive_minimize_and_restore fails on main #11
Why: found while measuring #7. The test fails on both origin/main and PR #10, so it's not caused by the lag fix.
`inactive_minimize_and_restore_keep_x64_and_x86_windows_protected` (`src-tauri/tests/first_frame.rs` or hmw-core tests, line ~307) fails on a focus assertion: `HMW_NO_FOCUS` vs `HMW_FOCUSED`.
`quick_launch_arms_x64_and_x86_before_first_show_and_stops_on_failure` (line ~681) failed once on main, passed on the fix branch; may be flaky.
Command: `cargo test --locked -p hmw-core --test first_frame -- --test-threads=1`, run on Windows 11 on a loaded machine (98% CPU).
Checklist
Find whether the focus failure is the test environment (background desktop can't take foreground) or a real regression
Fix the test or the code; make quick_launch deterministic if it's flaky
3 consecutive green runs of first_frame on Windows
Worker started · hmw-firstframe-11 (25bdff4d) · medium tier · folder .worktrees/hmw-feedback
Objective: Make the hmw-core first_frame tests pass reliably on Windows by finding and fixing why inactive_minimize_and_restore fails (focus assertion) and whether quick_launch is flaky.
Worker report 4 · hmw-firstframe-11
State: partial - local run still in the background, results not in yet
Needs decision: none
Links: draft PR #12 (opened by my manager), issue #11
Cause: the fixture's focus command called SetForegroundWindow once. Windows foreground-lock rules refuse that call from a background process, so the fixture reported HMW_NO_FOCUS. Fix (a570aa9, PR #12): tap Alt and retry up to 20 times, 50 ms apart. Protection assertions are unchanged. Proof: first_frame 4/4 on 3 consecutive CI runs on windows-latest, run 39112333442 attempts 1-3.
Worker started · hmw-first-frame-test (43045fda) · medium tier · branch hmw-first-frame-test in .worktrees/hmw-first-frame-test
Objective: Make the hmw-core first_frame integration tests reliably green on Windows.
Worker report 1 · hmw-first-frame-test
State: Tests are running; waiting for the result.
Needs decision: If you plan no follow-up for hmw-first-frame-test and it is clean, leave it idle: delegate reuses it and auto-stop removes it after the warm period.
Links:
Worker report 3 · hmw-first-frame-test
State: PR #13 is open. CI and the local 3× run are in progress; I'll report when they finish.
Needs decision: If you plan no follow-up for hmw-first-frame-test and it is clean, leave it idle: delegate reuses it and auto-stop removes it after the warm period.
Links:
Activity
MehdiMamas commentedon Oct 18, 2026
Worker started · hmw-firstframe-11 (25bdff4d) · medium tier · folder .worktrees/hmw-feedback
Objective: Make the hmw-core first_frame tests pass reliably on Windows by finding and fixing why inactive_minimize_and_restore fails (focus assertion) and whether quick_launch is flaky.
MehdiMamas commentedon Oct 18, 2026
Worker report 3 · hmw-firstframe-11
State: partial - fix committed locally, build and test running in the background
Needs decision: none
Links: none
MehdiMamas commentedon Oct 18, 2026
Worker report 4 · hmw-firstframe-11
State: partial - local run still in the background, results not in yet
Needs decision: none
Links: draft PR #12 (opened by my manager), issue #11
MehdiMamas commentedon Oct 18, 2026
Cause: the fixture's focus command called SetForegroundWindow once. Windows foreground-lock rules refuse that call from a background process, so the fixture reported HMW_NO_FOCUS. Fix (a570aa9, PR #12): tap Alt and retry up to 20 times, 50 ms apart. Protection assertions are unchanged. Proof: first_frame 4/4 on 3 consecutive CI runs on windows-latest, run 39112333442 attempts 1-3.
MehdiMamas commentedon Oct 18, 2026
Worker started · hmw-first-frame-test (43045fda) · medium tier · branch hmw-first-frame-test in .worktrees/hmw-first-frame-test
Objective: Make the hmw-core first_frame integration tests reliably green on Windows.
MehdiMamas commentedon Oct 18, 2026
Worker report 1 · hmw-first-frame-test
State: Tests are running; waiting for the result.
Needs decision: If you plan no follow-up for hmw-first-frame-test and it is clean, leave it idle: delegate reuses it and auto-stop removes it after the warm period.
Links:
MehdiMamas commentedon Oct 18, 2026
Worker report 3 · hmw-first-frame-test
State: PR #13 is open. CI and the local 3× run are in progress; I'll report when they finish.
Needs decision: If you plan no follow-up for hmw-first-frame-test and it is clean, leave it idle: delegate reuses it and auto-stop removes it after the warm period.
Links: