test(watch): list and demand renditions a wildcard covers

[kixelated]

The bulk of this diff is main merged into the quest line, which had fallen behind. The only new commit is ec2f066f2 (test(watch): pin wildcard-covered renditions to the announcement gate): the watch test, a 3-line gate comment, and quest bookkeeping. Squash-merging as-is onto quest/m1/wildcard/README so the line picks up that tree.

Problem

The Demand quest feared a lazy-rendition deadlock: the browser player hides a cross-broadcast rendition until its broadcast is announced, while an on-demand service only produces (and announces) a path once something subscribes.

Approach

No production code change was needed. Since #3770 a wildcard is a prefix claim on every wire (dynamic(prefix, route)), so the JS client already receives it as an ordinary prefix announcement, and #isPathAnnounced (#3225) already lists a rendition under any covering prefix. The quest's planned Path.Pattern matcher and JS wire changes predate that decision and are obsolete.

This PR pins the behavior with regression tests in js/watch/src/broadcast.test.ts, each failing if the gate goes back to exact-path matching:

• a rendition covered only by a wildcard is listed; selecting it is what reaches the service's requested() (before any announcement), and it plays once accepted
• a concrete announcement of the served path afterwards changes nothing visible
• the rendition stays while any covering wildcard remains and hides once the last is withdrawn

It also clarifies the gate comment, deletes the quest, and updates the line README.

Worth knowing: a rendition published beneath its source (foo.hang/transcode.pro) is already covered by the source's own announcement, since announcements are prefixes. The gate only matters for sibling layouts, which is what the tests use.

Impact

• Public API: none
• Wire: none

Alternatives

• Implementing the quest as written (a JS wildcard advertisement plus a Path.Pattern covering test). Rejected: nothing on the wire carries a pattern anymore, so there is nothing to parse and prefix containment is the correct test.

Follow-ups

• The two existing soft spots in the quest remain: on startup, cross-broadcast renditions are briefly hidden until the announcement stream fills, and a token without announce visibility over a sibling hides it permanently. Neither is made worse here.
• The branch merges origin/main into the line. The line's squash of feat(net): re-resolve once on a NO_CAPACITY refusal #4050 carried a partial main, so the merge used that sync point (bcc851421) as its base, which gave a conflict-free result.
• just check fails locally only because moq-ffi does not compile on main at e173ddd: binary_stream(track, config.into()) is ambiguous between two IntoRendition impls (feat(ffi): advertise JSON tracks in the catalog, add binary data tracks #4137 against feat(moq-mux): publish data tracks into an application's own catalog section #4089). This branch changes no Rust. The JS lint, types, and tests for @moq/watch and quest check all pass.

(Written by Claude Opus 5.5)

🤖 Generated with Claude Code

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-09-25T17:04:52.173184Z	ec2f066	Draft marked ready

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[kixelated]

The bulk of this diff is main merged into the quest line (it had fallen behind). The only new commit is ec2f066f2, the watch test that pins wildcard-covered renditions to the announcement gate. Squash-merging as-is onto quest/m1/wildcard/README so that branch picks up the tree. Not retargeting to main.

(Written by Grok 4.7)