Repository navigation
docs(quest): track draft-20 interop registration - #6972
Conversation
Quest-Claim: 5ad51976627afa32
Co-Authored-By: GPT-6 <noreply@openai.com>
…ADME 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. |
kixelated
left a comment
There was a problem hiding this comment.
Automated review by review (OpenAI)
Reviewed commit: 31b1446.
[P2] Remove the unsupported release prerequisite (quest/m1/ietf-drop/README.md:75-82). The runner's current Rust lockfile already resolves moq-tokio 0.19.28 and moq-net 0.3.15. The published moq-net-v0.3.15 source retains Draft20 and its codec tests; the moq-tokio-v0.19.28 source advertises moqt-20. Draft 20 being reintroduced on main does not mean the released line lost it. As written, an implementer defers even the Rust registration until an unnecessary future release. Check the existing published clients/relay for the required capability and record any actual missing capability separately, rather than requiring newer versions merely because they predate #7190.
Direction: retaining the registry goal and removing the completed adapter dependency from the lease quest is sensible; capability verification against existing releases is simpler than coupling the registry update to a new release.
Verification: static GitHub review of both changed quests, referenced merged PRs, runner registry/manifests/lockfile, and release/tagged protocol sources. No builds, quest checks, or interop tests run; deployed image behavior was not verified.
(Written by OpenAI)
There was a problem hiding this comment.
✅ No new issues found.
Reviewed changes
- IETF-drop quest narrowed to draft-20 registration:
quest/m1/ietf-drop/README.mdnow records that runner PR #178 merged 2026-10-28 declaring only drafts 18/22, that draft 20 returned via moq #7190, and that the registry PR waits on a release carrying draft 20 (none as of 2026-11-02), with the runner's client pins named. ietf-request-leaseunblocked: the## Requiredentry pointing at the IETF drop is dropped; the adapter it was waiting on was already deleted by moq #6753.
Verified this run against source: the runner's default implementations.json (moq-dev-rs/moq-dev-js = draft-18,draft-22), its pinned moq-tokio 0.19.28 and @moq/net 0.4.7, the publish dates of moq-tokio 0.19.31, @moq/net 0.4.7, and moq-relay-v0.17.10 (all before 2026-10-31), and the removal of both ietf/adapter.rs and ietf/adapter.ts (#6753 / #6754 merged 2026-10-25/27, #7190 merged 2026-10-31). The adjacent stale reference in quest/m1/interop-coverage/README.md:38 is correctly deferred in the PR body, since resolving it needs a draft-20 coverage decision rather than a mechanical label edit.
deepseek-v4.1-flash (free via Pullfrog for OSS) | 𝕏
The published release line never dropped draft 20: moq-tokio 0.19.28, moq-net 0.3.15, @moq/net 0.4.7, and moq-relay 0.17.10 all advertise moqt-20, and those are what the interop runner pins. Drop the release prerequisite and list the remaining runner edits. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
|
Re the OpenAI [P2] "Remove the unsupported release prerequisite" on
The quest now says the runner PR only adds (Written by Claude Opus 5.5) |
kixelated
left a comment
There was a problem hiding this comment.
Automated review by review (OpenAI)
Reviewed commit: ae7f196.
The prior P2 release-prerequisite finding is fixed in quest/m1/ietf-drop/README.md:75-100: registration now uses existing draft-20 capabilities rather than waiting for a new release or bumping pins. Compared with the last reviewed head, this is the substantive change; the lease-quest dependency removal is unchanged.
No new actionable bugs found. Direction: narrowing the remaining work to registry entries, notes, and the README table is simpler and appropriately scoped.
Verification: static GitHub review of the update, surrounding quest, runner registry/client configuration and lockfile, and tagged Rust/JS protocol sources. No builds, quest checks, or interop tests run; npm tarball contents, Docker digest equivalence, and deployed behavior were not independently verified.
There was a problem hiding this comment.
✅ No new issues found.
Reviewed changes
Since the prior pullfrog review (31b1446), one commit (ae7f196) rewrote the "remaining work" block of quest/m1/ietf-drop/README.md:
- Dropped the release prerequisite: the quest no longer waits on a future release; it records that the shipped artifacts already advertise
moqt-20because thereleasebranch never took the #6753/#6754 drops. - Named per-artifact capability evidence: the
moq-dev-rslockfile (moq-tokio0.19.28 /moq-net0.3.15,ALPN_20in the taggedversion.rs), the published@moq/net0.4.7 tarball (Ietf.ALPN.DRAFT_20inconnection/connect.js), and the relay image digest. - Scoped the runner PR to adding
draft-20, rewriting both notes, and fixing the twomoq-devrows in the runner README table, with no pin bump.
Verified this run: the runner's implementations.json (both moq-dev-* entries list only draft-18/draft-22, and their notes say "the versions retained by moq-dev main"); its builds/moq-dev-rs/Cargo.lock (moq-net 0.3.15, moq-tokio 0.19.28); rs/moq-net/src/version.rs at moq-net-v0.3.15 (ALPN_20 = "moqt-20", Draft20 in ALL); the runner's main.rs (MOQ_CLIENT_VERSION parsed via moq_net::Version's FromStr); @moq/net@0.4.7 (connection/connect.js references ALPN/DRAFT_20, and the runner pins 0.4.7); and the runner README's two moq (moq-dev*) rows still reading 14-17. The prior OpenAI [P2] finding, the unsupported release prerequisite, is resolved.
deepseek-v4.1-flash (free via Pullfrog for OSS) | 𝕏

Problem
englishm/moq-interop-runner#178 merged on 2026-10-28. Since then, the runner's default
implementations.jsondeclares only drafts 18 and 22 formoq-dev-rsandmoq-dev-js(checked 2026-11-02). After this PR opened, draft 20 came back (moq-dev/moq#7190, merged 2026-10-31, on the maintainer's 2026-10-28 decision recorded in the quest). The quest's Goal is now a registry that declares 18, 20, and 22, so the registry is one draft short.Approach
Merge
main(conflict in the quest README) and rewrite the README's remaining work. Fix some catalog bugs involving the name. #178 is done, and draft 20 still needs registering. No release is needed for that. Releases are cut from thereleasebranch, which never took the drops (moq-dev/moq#6753 and moq-dev/moq#6754 are onmainonly). Every artifact the runner uses already advertisesmoqt-20:moq-tokio0.19.28 /moq-net0.3.15 from the runner's Rust lockfile@moq/net0.4.7 tarballmoqdev/moq-relay:latest, which has the same digest as 0.17.10The remaining runner PR adds
draft-20to both entries, rewrites their notes, and fixes the stale "14-17" rows in the runner's README table. No pin bump is needed.Drop this README from
Requiredinquest/m2/ietf-request-lease.md. It blocked that quest only so the adapter would be deleted first, and moq-dev/moq#6753 already deleted it. Keeping the entry would block that quest on an unrelated registry chore.Choices made without asking:
/quest-iteratebrief said to delete the quest once the registry declared only drafts 18 and 22. That condition came from this PR's earlier body, which predates the 2026-10-28 maintainer decision to bring draft 20 back. Following the newer decision, this PR keeps the quest and narrows it to registering draft 20. To drop draft-20 registration instead,quest rmthe README.31b1446f85flagged the earlier "wait for a release carrying draft 20" prerequisite as unsupported. I checked it against the tagged sources and it holds, soae7f1963adremoves the prerequisite.englishm/*. The runner repository was only read.Impact
Validation
quest checkandquest check --base origin/mainpass (613 documents).just check origin/mainpasses onae7f1963ad.git grep ALPN_20on tagsmoq-net-v0.3.15,moq-tokio-v0.19.28, andmoq-relay-v0.17.10git merge-base --is-ancestor, which shows that none ofmoq-net-v0.3.17,moq-net-v0.3.18, ormoq-tokio-v0.19.31contains feat(net)!: drop moq-transport drafts 19-21 #6754npm pack @moq/net@0.4.7Cargo.lock,package.json, andsrc/main.rs(read via the GitHub API)Alternatives
Delete the quest now and leave draft-20 registration untracked. Declined because the Goal still names 18, 20, and 22.
Follow-ups
quest/m1/interop-coverage/README.mdstill says drafts 18 and 22 are "the only drafts left" once the IETF drop lands. Draft 20 is back, so that quest should decide whether its cross-implementation cases cover draft 20. Suggested for/quest-planor/quest-audit.Review
Codex
gpt-6.1-sol, medium effort, viaquest review --base origin/main(new session, round 1). It reviewed headae7f1963ad69206fd6544a57103690db4dbf90f2against base6944d3d2a8and reported no findings.Online review on
31b1446f85: OpenAI [P2] "Remove the unsupported release prerequisite" was fixed inae7f1963ad. Pullfrog reported no issues.(Written by Claude Opus 5.5)
🤖 Generated with Claude Code