[botandbang]

What and why

Pressing d in the player listed the audio outputs on the UI thread (deck::list_outputs called outputs.list directly), so the screen didn't redraw until cpal had enumerated every host's devices, which can take a while with ALSA on Linux. Keys pressed meanwhile went to a list that wasn't drawn yet: the scheduled Bombadil run saw d then the next key close the list unseen. Now:

• src/tui/deck.rs: list_outputs opens the list at once, on the system's default, and lists the devices on a worker through app.spawn_or, as read_sets does. listed_outputs fills the list when they come back, selecting the output playing, and drops a listing for a list closed or opened again since (a listing id, as Msg::Sets drops an older load). Outputs gains listing and failed.
• src/tui/app.rs: Msg::Outputs carries the result; animating() counts the outputs being listed, so they're drawn as soon as they're in rather than at the next idle poll.
• src/tui/ui.rs: while the devices are listed, the list shows "listing the outputs…" under the default; when they can't be, it says why in the list ("can't list the outputs: …"), in the style of "No audio output". The devices replace that row, so with one device the popup keeps its size.
• src/tui/audio.rs: Outputs::list returns a Result, and listing the devices passes on cpal's error, where it used to show only the system's default as if there were no others.
• CHANGELOG.md: a line under Unreleased.
• bombadil/tui.sh and CONTRIBUTING.md gave --seed 7 as an example option, which Bombadil 0.7.8 doesn't have; the example is --columns 80.

Opening an output (use_output, and open_preferring when a track starts) still runs on the UI thread and looks the device up by name; moving that means handing a cpal stream across threads, and isn't in this change.

How it was tested

• New tests in src/tui/deck/tests.rs: the list opens with "listing the outputs…" before its devices are in and fills when they arrive; a late listing after the list was closed, or for an older listing after it was reopened, is dropped; a listing that fails (or panics) says why in the list and Enter still plays through the default. Each fails with the old synchronous list_outputs.
• background_results_land_where_they_belong (Hegel) sends Msg::Outputs with current and stale ids, Ok and Err, to open and closed lists; the_screen_moves_only_while_something_does counts the outputs being listed. Both fail with the id check or the animating() line removed.
• CI=true cargo nextest run -p locallibrary-cli: 1,046 passed; cargo clippy --workspace --all-targets --locked -- -D warnings; cargo fmt --all --check.
• TIME_LIMIT=5m ./tui.sh: every d drew the list with "listing the outputs…". The run stopped on digitsSwitchViews in full screen, where digits play hot cues; fix(locallibrary): get the scheduled suites passing again #200 fixes that in the spec.

Checklist

• Tests cover the change
• For the website: Storybook stories cover any new UI states
• For locallibrary: apps/locallibrary/CHANGELOG.md has a line under Unreleased, if people using it will notice
• The READMEs, docs/ and --help still say what the code does (apps/locallibrary/docs/rpc/ too, for a sidecar change)

[github-actions]

Rust coverage

From run 323. Both are held to the thresholds in .cargo/config.toml.

Tests	Lines	Functions	Regions
All	97.38% (77588/79676)	96.45% (12290/12743)	95.66% (170428/178151)
Hegel property tests alone	95.42% (76026/79676)	93.89% (11964/12743)	93.5% (166575/178151)