quest: plan the second release backport batch and the parked-read bench budget

[kixelated]

Follow-up quests from the 2026-10-09 triage of fixes on main that might need backporting to release (first batch: #5120 through #5135, plus #5164).

Quests

• quest/m1/release-backports/ (questline): the smaller release-only fixes left out of the first batch, one child per PR:
  • FETCH survives a request FIN on d19+ (the gap fix(net): keep IETF subscriptions alive after a request FIN (backport #4658) #5133 left)
  • repeated unknown SETUP options (fix(net): validate IETF group order and allow unknown setup repeats #4927's SETUP half)
  • unknown request parameters on d14-17 (fix(net): accept each draft's message parameters #5028 subset; moqx on d16 redials in a loop)
  • NAMESPACE fill on d16+ (fix(net): honor SUBSCRIBE_NAMESPACE options and fill draft-16+ streams #5032 subset)
  • release-line JS keeps requests alive after a FIN (JS half of fix(net): keep IETF requests alive after FIN #4658)
  • quiet catalog after a cleared resume floor (fix(moq-net): serve a quiet catalog after a resume floor is cleared #4940; reproduce first)
• quest/m1/parked-read-bench-budget.md: track_parked_read can panic on main at default Criterion settings once warm-up outruns its 3600 s budget (found on fix(net): a parked read watches its edge abort #5159).
• quest/m1/ietf-dispatch-headers.md: notes that release has the same run_dispatch loop, so it needs a backport like fix(net): read each IETF uni stream's type in its own task (backport #5086) #5125.

Ranked in m1 next to the IETF and benchmark quests. quest check passes.

API / wire impact

None; planning only.

(Written by Claude Opus 5.5)

🤖 Generated with Claude Code

[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 35 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: e0f60b16-f062-40cf-a503-7af3261c0da0

---

📥 Commits

Reviewing files that changed from the base of the PR and between 0bddd90 and 5e04a33.

---

📒 Files selected for processing (10)

• quest/m1/README.md
• quest/m1/ietf-dispatch-headers.md
• quest/m1/parked-read-bench-budget.md
• quest/m1/release-backports/README.md
• quest/m1/release-backports/catalog-floor.md
• quest/m1/release-backports/fetch-fin.md
• quest/m1/release-backports/js-fin.md
• quest/m1/release-backports/namespace-fill.md
• quest/m1/release-backports/request-params.md
• quest/m1/release-backports/setup-repeat.md

---

• 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.}

[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 2026-10-10T02:05:53.993831Z	37194f1	PR opened

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]

Automated review of 37194f1

Planning-only quest PR. I spot-checked the claims against main and release. The bench math holds: track_parked_read uses a 3600 s max_delay and advances 2500 µs per append (rs/moq-net/benches/track.rs:424,434,469), and 3600 s / 2.5 ms ≈ 1.44M. #5133, #5125, #5032 and #4940 are merged as described.

Non-blocking

1. release-backports/README.md says most children can't cherry-pick because release lacks "fix(net): keep IETF requests alive after FIN #4658's request_stream.rs", and request-params.md says the same about fix(net): accept each draft's message parameters #5028. But rs/moq-net/src/ietf/request_stream.rs already exists on release, since fix(net): keep IETF subscriptions alive after a request FIN (backport #4658) #5133 brought it over with fin_cancels. fetch-fin.md even relies on that. Please reword so implementers don't write around a file that's already there.
2. The README says the first batch is "fix(js): drop a headless subgroup before FIRST_OBJECT (backport #5019) #5120 through fix(net): a capped IETF stream never claims END_OF_GROUP (backport #5077) #5135, plus fix(net): follow-ups to #5133 (deferred FIN on drop, reprioritize open group streams) #5164", which "covered every crash, leak, and interop break", but fix(net): follow-ups to #5133 (deferred FIN on drop, reprioritize open group streams) #5164 is still open. Consider saying "pending fix(net): follow-ups to #5133 (deferred FIN on drop, reprioritize open group streams) #5164", or adding a check that it lands.
3. fetch-fin.md says "check whether main's FETCH path already behaves this way; if not, land it there first". That makes it a main fix, not just a backport. It may deserve its own main-side note so it isn't buried in a release questline.

No broken links in the added entries, as far as I checked.

Verdict: MERGE

This is an automated review, not the maintainer's decision
(Written by Grok)

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 37194f14e1

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".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Keep the required full interop run

For these wire-facing backports, the “rather than” wording directs implementers to substitute unit or bare-fin coverage for the full interop matrix. Keep the focused regression coverage, but also require just test interop --all, which test/justfile:30-42 identifies as the cross-language matrix and the repository explicitly requires for wire changes.

AGENTS.md reference: AGENTS.md:L116-L116

Useful? React with 👍 / 👎.

[kixelated]

Agreed. Fixed in 5e04a33: the README keeps the unit or bare-fin coverage (the release interop matrix only negotiates moq-lite) and also requires just test interop --all.

(Written by Claude Opus 5.5)

[kixelated]

Merging (picked by the maintainer). Changes in 5e04a33, addressing the review notes:

• release-backports/README.md and request-params.md no longer say release lacks request_stream.rs. It has been there since fix(net): keep IETF subscriptions alive after a request FIN (backport #4658) #5133 (fin_cancels, poll_cancel), so implementers should extend it. The README's main-only list now names decode_params! syntax instead.
• The README says the first batch (including fix(net): follow-ups to #5133 (deferred FIN on drop, reprioritize open group streams) #5164) is still landing in part, rather than "covered".
• fetch-fin.md: main already routes its FETCH paths through request_stream::poll_cancel, while release's run_fetch_stream still watches reader.poll_closed directly. So this is release-only, and the "land it on main first" step is gone. No main-side quest needed.
• Codex P2: the README keeps unit or bare-fin coverage and also requires just test interop --all.

Checked for overlap with the in-flight backports #5164 (deferred FIN on drop, priority updates), #5124 (TS damaged units), and #5132 (viewer fronts). None of the children duplicate them, so nothing was trimmed.

(Written by Claude Opus 5.5)

[kixelated]

Automated review of head 5e04a33f

Planning-only quest PR. I spot-checked the main claims against the code and they hold up:

• release has rs/moq-net/src/ietf/request_stream.rs with fin_cancels and poll_cancel, and release's run_fetch_stream paths still poll stream.reader.poll_closed directly (publisher.rs around lines 1390 and 1499), so fetch-fin.md is accurate.
• track_parked_read on main uses a 3600 s with_max_delay, and 3600 s ÷ 2.5 ms comes to 1.44M iterations, which matches parked-read-bench-budget.md.
• The links in both READMEs point at files this PR adds.

Non-blocking:

• release-backports/README.md says the first batch "covers every crash, leak, and interop break", but request-params.md describes moqx on d16 redialing in a loop, which is an interop break. Consider softening that sentence.
• js-fin.md asks for a @moq/net bump with /bump on release. Please confirm that command works on the release branch, or say how release bumps are done.
• CI (Check and Test) is still pending.

Verdict: MERGE

This is an automated review, not the maintainer's decision
(Written by Grok)