Repository navigation
quest: record pre-cut follow-ups from the quest-spawn run - #6790
Conversation
Twelve quests for breaks on main and gaps found while landing the 2026-10-24 quest-spawn PRs: the loom build, `just check` without user namespaces, the red TS workflow, OBS dropping groups, a slow moq-cli test, range-count cost, the superlinear fill path, a possible js/net close race, relay demand gaps, fMP4 boundary sync, moq-c decode start, and one moq-ffi build definition. bench-ci gains the malloc-pin question. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
Warning Review limit reachedYou'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 17 minutes. View limit details
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. Comment |
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
kixelated
left a comment
kixelated
left a comment
There was a problem hiding this comment.
Automated review by review (OpenAI)
Reviewed commit: 551585c
No actionable bugs found in this documentation-only change. The quests make pre-cut follow-ups discoverable, distinguish suspected defects from reproduced failures, and preserve reproduction-first investigation instead of treating proposed fixes as established conclusions. Spot checks confirm the stale Loom call, FFI Data-kind construction and C decoder’s Latest start setting.
Direction: linked, independently scoped follow-ups are appropriate. Keeping skipped checks distinct from passed checks and preserving congestion assertions are useful acceptance constraints.
Verification: static inspection of all changed quest files and selected implementation paths. No quest/Markdown checks, tests, benchmarks or CI reproductions run; reported workflow failures and performance measurements were not independently verified.
Adds quests for reported-drop parity between Rust and JS, the flaky OBS audio-candidate test, and upstream bug reports (faad, retina, mr-boxington). The release cut gains the moq-cli/moq-gst MIGRATE entries and the wrapper moq-ffi floors; the interop and dead-code audits gain the imquic announce timeout, a dead RecvBandwidth branch, and leftover `dynamic` names. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
kixelated
left a comment
kixelated
left a comment
There was a problem hiding this comment.
Automated review by review (OpenAI)
Reviewed commit: b014ddd
No new actionable bugs found. The additions preserve the unresolved Rust/JS drop-finality decision rather than silently choosing semantics, and capture the OBS flake, upstream reports, migration entries and wrapper version floors as explicit follow-ups. Linked source PRs support the drop-finality and release-compatibility concerns.
Direction: extending existing audits and the cut checklist avoids competing tracking paths; upstream reporting retains an explicit approval boundary.
Verification: static comparison against 551585c, added planning text and referenced PR context. No quest/Markdown checks, tests or reproductions run; reported upstream failures were not independently reproduced.
…s it Per #6722's decision that fetch-range waits on a linear fill path. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
kixelated
left a comment
kixelated
left a comment
There was a problem hiding this comment.
Automated review by review (OpenAI)
Reviewed commit: 24a7ee9
No new actionable bugs found. Linear filling is now an explicit prerequisite of FETCH-as-range and first in the parent’s required list. The reciprocal prose link is not a dependency, so no cycle is introduced. This prevents expanding use of the known superlinear path before its performance work lands.
Verification: static comparison against b014ddd, separating incoming main changes from the three-file plan adjustment. No quest validation, tests or benchmarks run.
|
Maintainer decisions (2026-10-25), recorded in moq-dev/moq#6798:
(Written by Claude Opus 5.5) |
# Conflicts: # quest/m0/ci/README.md # quest/m1/broadcast-lifecycle/README.md # quest/m1/test-flakes-2/README.md
# Conflicts: # quest/m1/c/README.md
# Conflicts: # quest/m1/test-flakes-2/README.md
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…anning-700add Record the 2026-10-25 answers in the quests #6707 and #6790 landed: datagram size, description-less AAC, catalog parity, MKV/FLV codecs, partial groups, drop finality, and resume. decode-start stays under the hand-written m1/c line, and api-audit/requests renames moq-c too. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Records the breaks on
mainand the gaps found while landing the 2026-10-24 quest-spawn PRs, so they get fixed before the cut instead of living in PR bodies.m0 (CI)
ci/loom-build:--cfg loomfails to build on main (tests/loom.rs:621,with_groupsremoved by refactor(net)!: one with_range cursor setter #6138). Reproduced locally.ci/check-without-userns:just check's flake module fails whereunshareis refused and cancels other lanes.m1
ts-reliability/validation-green: the TS workflow has been red since 2026-10-21 (congested and overflowing paths), plus five lanes newly losing video frames on 2026-10-25. Last green and first-bad commits are in the quest.obs-data-decode: moq-ffi decodes every media track asKind::Data, so OBS likely drops every other group (from fix(obs): stamp decoded pictures with their own timestamp and drain at EOF #6711, not yet reproduced).test-flakes-2/export-echo-time: the 35 s moq-cli echo report test.subscribe-ranges/range-cost: per-group cost grows with a peer-chosen range count (feat(moq-net): carry ranges on lite-07 SUBSCRIBE #6715).subscribe-ranges/fill-linear: the superlinear fill path measured in test(net): benchmark range filling; scope fetch as a range #6722.broadcast-lifecycle/js-close-fin: check js/net for the close race fix(net): read SUBSCRIBE FINs that raced a peer's close #6741 fixed in Rust.relay-demand-gaps: three gaps from fix(net): a widened max age re-offers skipped cached groups #6717.fmp4-boundary-sync: a declared non-sync boundary opens a passthrough group (fix(mux): start fMP4 passthrough video at a sync sample #6743).c/decode-start: moq-c decoders hardcodeStart::Latest(feat(bindings)!: decoders read gaps and ends #6745).ffi-build-flags: one definition of moq-ffi's binding features (ci: check the bindings in separate lanes #6716).bench-cigains the question of pinning glibc malloc for every bench (fix(mux): pin glibc's heap in the ts_schedule bench #6727).resume-drop-final: Rust and JS disagree on whether a reported drop is final (fix(net): report drops through a resumed JS subscriber #6758). The maintainer decides.test-flakes-2/obs-audio-candidate: the OBS "failed audio candidate preserves playback" test fails under load (fix(obs): stamp decoded pictures with their own timestamp and drain at EOF #6711).m2
upstream-reports: report the faad, retina, and mr-boxington bugs (each post needs approval).release/cutgains the moq-cli/moq-gst MIGRATE entries from refactor(cli,gst,publish)!: align flags and drop the publisher linger opt-ins #6522 and the wrapper moq-ffi floors (feat(bindings): every binding names itself in SETUP #6775).audit-interopgains the imquicannounce-subscribetimeout.audit-dead-codegainsRecvBandwidth::polland leftoverdynamicnames.Left out on purpose: the red Nightly (owned by
m0/nightly), the moq-rtmpBackoff::validatedead code (them0/nightly/feature-shapesclass, noted on #6703), and the HLS quests already written in open PRs.Public API: none. Wire: none.
Checks:
quest check(555 documents ok),sh/markdown.sh check.🤖 Generated with Claude Code
Decisions
fill-linearranks abovefetch-rangein the subscribe-ranges epic, andfetch-range.mdlists it under Required.quest/m1/env-unknown.md.(Written by Claude Opus 5.5)