chore(lane): daily maintenance 2026-10-16

[kirillDevPro]

Docs

docs/MOON_PATCH_QUEUE.md lists the Moon GPUI patches a Zed re-sync must keep. Its tooltip entry covered only the 800 ms show delay, the five-second expiry and how hoverable tooltips persist. It did not mention the Moon-only press suppression that #119 (0c4a7c4) added, so a re-sync would silently drop it.

What the code does now: a mouse press on a regular tooltip's trigger replaces its state with SuppressedUntilMouseLeaves, which hides the tooltip or cancels a pending show. TooltipTrigger::check_can_show decides whether a tooltip may show, and check_is_hovered alone clears the suppression once the pointer leaves, even with the button still held. Upstream only cleared the tooltip on press, and the next pointer move after the release scheduled it again.

How I verified it: I read the entry against crates/moon-gpui/src/elements/div.rs (register_tooltip_mouse_handlers, handle_tooltip_mouse_move) and against the pre-change code in 0c4a7c4^. cargo xtask component-mirror --check-baseline and cargo xtask component-audit --check-baseline both pass.

Reviewer: Codex. It disputed the first wording of the upstream sentence: the release itself does not reschedule the tooltip, the next pointer move does. I corrected that sentence.

Also read and found current: README.md, docs/PALETTE_SPEC.md, docs/COMPONENT_AUDIT.md.

Hunt

Ground: moon-util (rel_path.rs).

• Input: RelPath::new(Path::new("C:\\"), PathStyle::Windows), also "C:/" and "d:\\\\".
• Wrong output: Ok("C:"). Trailing separators were trimmed before the absolute-path check, so the drive root lost the separator that paths::is_absolute needs (letter, colon, separator). PathStyle::strip_prefix with an empty parent therefore returned Some("C:").
• Expected output: an error. The function's doc says it returns an error for an absolute path, paths::is_absolute("C:\\", Windows) is true, and C:\foo and C:\. were already rejected.
• Fix: check is_absolute on the input before trimming. Only a drive letter and colon followed by nothing but separators gives a different result.
• Test: rel_path::rel_path_tests::new_rejects_a_windows_drive_root failed before the fix and passes now. Re-run with cargo test -p moon-util --lib new_rejects_a_windows_drive_root.
• Reviewer: Codex. It confirmed the defect and that no other input changes result. Only the error text of already-rejected inputs now quotes the untrimmed path.

Another bug I proved along the way is filed as #123: RelPath::file_stem depends on the host OS for POSIX names that contain a backslash.

Quiet jobs

• Leftovers: quiet, no open lane pull requests.
• Fix: quiet. The only open auto-found issue, Legacy PowerShell drops empty native task arguments after 06aa89a #113, carries needs-human.