[kixelated]

Problem

ExportSource sampled private consumer counters after reading frames, so gaps and declared ends were observed after the fact. Exporters need the read stream's ordering while retaining fMP4 group and duration boundaries.

Approach

Share a private stream of Read items and exporter boundaries. ExportSource handles gaps and ends from that stream, and the consumer counters are removed. Final packets from a stopped audio encoder keep the old timeline until their group ends.

Impact

• Public API: none; all new plumbing is crate-private.
• Wire: none.

Alternatives

Keeping separate consumer counters leaves two observations to reconcile and loses the ordering between a gap and an end.

Validation

• The gap/end regression fails on unmodified main.
• The audio-drain regression caught an early generation advance before the final fix.
• just rs test -p moq-mux --no-fail-fast: 1,328 passed, including unchanged exporter tests and the exact fMP4 endpoint regression.
• just check: passed, including 3,046 default tests, 460 capture tests, 178 play tests, wasm checks and the media feature matrix.
• quest check: passed.

Follow-ups

None.

(Written by GPT-6)

---

^{Need help on this PR? Tag @codesmith-bot with what you need.}

[kixelated]

Completed the quest and removed its tree entry. Local just check, all 1,328 moq-mux tests, and quest check pass. The new regressions cover a gap immediately followed by an end and a stopped encoder's final packets keeping their original timeline. Public API and wire format are unchanged.

No implementation decisions or follow-up quests remain. Left as a draft as requested by the spawning workflow.

(Written by GPT-6)

[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 2026-10-14T23:41:46.876336Z	ad4ec5c	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.

[coderabbitai]

Warning

Review limit reached

You've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository.

Next included review available in 6 minutes.

Check out review usage here.

View limit details

Limit details: You’ve used all 4 included reviews currently available.

Learn how review limits work.

Review configuration:

⚙️ Run configuration

• Configuration used: Organization UI
• Review profile: CHILL
• Plan: Advanced
• Run ID: 23b01275-25d7-404c-8058-ed427db73b02

📥 Commits

Reviewing files that changed from the base of the PR and between dcdda2f and ad4ec5c.

📒 Files selected for processing: 5

• quest/m1/README.md
• quest/m1/export-source-read.md
• rs/moq-mux/src/container/consumer.rs
• rs/moq-mux/src/container/source.rs
• rs/moq-mux/src/container/ts/import.rs

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[kixelated]

Merge summary for ad4ec5c3f716b0d43dc3fbe2051448dfcc3119f1:

• Rebased onto main (fix(moq-mux): a source the output cannot follow fails the export promptly #5490, fix(ts): carry a full-rate CBR feed at its own multiplex rate #5516, fix(net): hand each group sequence out once per reader #5446 and others) with no conflicts. just rs test -p moq-mux passes all 1,352 tests, and moq-mux clippy is clean.
• Review: Codex reviewed this head and found nothing (👍). CodeRabbit was rate-limited.
• Behavior note: a marker group that a latency skip sheds now counts as a gap (skips only). It used to also bump markers (restart). The jitter buffer still re-anchors when the jump outruns its bound, and a contiguous shed marker stays on the current clock, which is the right outcome. This is the intended Read semantics from feat(mux): read gaps and declared ends as items in delivery order #5288.
• The synthetic GroupEnd that ExportSource emits on a gap makes fMP4 cut the fragment when a group is evicted mid-way, instead of waiting for the next group's frame.

Public API: none. Wire: none. No follow-ups.

(Written by Claude Opus 5.5)