37 lines (30 loc) · 1.27 KB

Result Log: 2026-03-10 15:57:45 CST

Goal

Run a hard recovery benchmark that deliberately induces wrong focus, then compare App.switch against a visible taskbar fallback.

Steps

Observed Settings as the foreground window.
Induced wrong focus with one taskbar click on Chrome.
Verified Chrome foreground.
Attempted semantic recovery with App.switch("Notepad").
Re-observed after the failure.
Recovered with one taskbar click on the visible Notepad button.
Typed a marker into Notepad and re-observed.

Results

Wrong-focus induction

pass
Chrome became the foreground window as expected.

Semantic recovery

fail closed
App.switch("Notepad") returned a failure message
the next snapshot still showed Chrome foreground

Taskbar fallback recovery

pass
one taskbar click on the visible Notepad button restored Notepad foreground

Post-recovery task completion

pass
Type(loc="500,400", text=" rf") succeeded
Notepad advanced from 30 个字符 to 33 个字符

Takeaway

Focus recovery should be treated as a layered problem, not a single primitive:
- App.switch is useful but not sufficient in every browser-to-desktop state
- a verified taskbar fallback is stronger when the snapshot already exposes the target button