moq-lite-05: replace SUBSCRIBE_OK/FETCH_OK with on-demand Track Stream

[kixelated]

What

Replace the per-response publisher metadata in draft-lcurley-moq-lite.md with a dedicated, on-demand Track Stream, and remove SUBSCRIBE_OK and FETCH_OK.

Why

A track's publisher-chosen properties — Priority, Ordered, Cache, Timescale, Compression — were being echoed on every SUBSCRIBE_OK (and, as recently drafted, every FETCH_OK). These values are immutable for the life of a track, so repeating them on every response is redundant — especially when fetching a track group-by-group.

Making them immutable also avoids a relay fan-out problem: a publisher-side change to one upstream subscription would have to propagate out to every downstream subscriber (and invalidate cached groups). Subscriber properties are the opposite — they fan in at a relay, which already merges them, so those remain mutable via SUBSCRIBE_UPDATE.

What changed

• New Track Stream (0x6): TRACK request (broadcast path + track name) → single TRACK_INFO reply → FIN (or reset on error / missing track). TRACK_INFO carries the immutable publisher properties. Fetched once and cached; reused across every SUBSCRIBE and FETCH.
• Removed SUBSCRIBE_OK: a subscription is now accepted implicitly (rejection = stream reset). The resolved group range moved into dedicated SUBSCRIBE_START / SUBSCRIBE_END messages. SUBSCRIBE_START resolves the absolute start (>= requested; a larger value implicitly drops the leading range); SUBSCRIBE_END signals no group follows a given sequence (stragglers may still be dropped before FIN). Subtypes renumbered: START=0x0, END=0x1, DROP=0x2.
• Removed FETCH_OK: a FETCH returns bare FRAME messages directly on the stream.
• Prioritization example reworked: since publisher priority can no longer be bumped mid-subscription, active-speaker adaptation is now driven by subscriber priority.
• Repointed cross-references, updated the Subscribe/Fetch stream overviews, and refreshed the moq-lite-05 changelog.

Reviewer notes / open questions

• No explicit subscription ack. Removing SUBSCRIBE_OK means acceptance is inferred from the absence of a reset (and from SUBSCRIBE_START, which may be deferred for "latest" subscriptions). If a positive "subscribed" signal is desired, it would need a minimal ack — which partly reintroduces what this removes.
• TRACK_INFO is a parse prerequisite. Frames are unparseable without it (timescale gates the per-frame delta fields). The subscriber MUST obtain it first; it MAY open the Track and Subscribe/Fetch streams concurrently and buffer frames until it arrives. This trades a potential first-access round-trip for eliminating per-response repetition.
• Cache invalidation is currently tied to broadcast re-announcement; if there's a stronger notion of track identity, the cache should key off that instead.

Validation

• Cross-reference checker reports zero dangling anchors. Fixed a real #track anchor collision (data-model ## Track, overview ### Track, message ## TRACK) by giving the overview an explicit {#track-stream} anchor.
• Not built: kramdown-rfc / xml2rfc are unavailable in this environment and the lib/ submodule is unpopulated, so make was not run. Structure is sound but the rendered build is unverified.

🤖 Generated with Claude Code

[coderabbitai]

Warning

Review limit reached

@kixelated, we couldn't start this review because you've reached your PR review rate limit.

More reviews will be available in 49 minutes and 28 seconds. Learn how PR review limits work.

Your organization has run out of usage credits. Purchase more in the billing tab.

⌛ How to resolve this issue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

We recommend that you space out your commits to avoid hitting the rate limit.

🚦 How do rate limits work?

CodeRabbit enforces hourly rate limits for each developer per organization.

Our paid plans include higher PR review limits than trial, open-source, and free plans. In all cases, reviews become available again over time. During sustained high-volume PR review activity, CodeRabbit may temporarily slow when the next review becomes available.

Please see our Fair Usage Limits Policy for further information.

ℹ️ Review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

Run ID: 1d304313-6ad5-4294-9dc0-0bf12ab6d61e

📥 Commits

Reviewing files that changed from the base of the PR and between 1f0225a and a31c367.

📒 Files selected for processing (1)

• draft-lcurley-moq-lite.md

Walkthrough

This pull request updates the moq-lite specification to introduce a new Track Stream (0x6) with TRACK/TRACK_INFO message exchange, shifting per-track media configuration (timescale, compression, priority, cache settings) from SUBSCRIBE_OK to immutable TRACK_INFO fields. The Subscribe Stream behavior is rewritten to model implicit acceptance with explicit SUBSCRIBE_START/SUBSCRIBE_END/SUBSCRIBE_DROP response messages. Frame and datagram field presence conditions are updated to reference the new TRACK_INFO source. Priority semantics are clarified, expiration and cache references are updated, FETCH response semantics are simplified (no response header, stream reset on error), and the changelog reflects these moq-lite-05 wire-level and semantic changes.

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title accurately summarizes the primary change: replacing SUBSCRIBE_OK/FETCH_OK with an on-demand Track Stream in moq-lite-05.
Description check	✅ Passed	The description comprehensively relates to the changeset, detailing what was changed, why, and how, with validation notes.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

✨ Simplify code

• Create PR with simplified code
• Commit simplified code in branch claude/gallant-hertz-102168

---

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 and usage tips.}

[coderabbitai]

Actionable comments posted: 4

🧹 Nitpick comments (1)

draft-lcurley-moq-lite.md (1)

371-384: 💤 Low value

Consider adding language identifiers to code blocks in the Priority example.

Lines 371-384 contain code blocks demonstrating priority values, but they lack language identifiers (triggering markdownlint MD040). While these are illustrative pseudo-code rather than executable code, adding a language hint (e.g., ```text) would satisfy linters and improve rendering consistency.

✨ Optional improvement for code block formatting

-```
+```text
 ali/audio + bob/audio: subscriber_priority=2 publisher_priority=2
 ali/video + bob/video: subscriber_priority=1 publisher_priority=1

- +text
bob/audio: subscriber_priority=4 publisher_priority=2
bob/video: subscriber_priority=3 publisher_priority=1

</details>

<details>
<summary>🤖 Prompt for AI Agents</summary>

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In @draft-lcurley-moq-lite.md around lines 371 - 384, Update the two fenced code
blocks showing priority examples so they include a language hint (e.g., change
the triple backticks to text) to satisfy markdownlint MD040 and ensure consistent rendering; specifically modify the blocks containing "ali/audio + bob/audio: subscriber_priority=2 publisher_priority=2" and the subsequent block starting with "bob/audio: subscriber_priority=4 publisher_priority=2" to use text fences instead of plain ``` fences.

</details>

</blockquote></details>

</blockquote></details>

<details>
<summary>🤖 Prompt for all review comments with AI agents</summary>

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In @draft-lcurley-moq-lite.md:

• Around line 288-298: Augment the Subscribe Stream section to remove ambiguity
  by specifying timing rules for SUBSCRIBE_START and stream resets: state that
  when a subscription cannot be fulfilled because the track does not exist or is
  invalid the publisher MUST promptly reset the Subscribe stream (ideally within
  one RTT or a configurable fixed timeout) rather than leaving it pending; when
  the track exists but currently has no groups, allow the publisher to delay
  sending SUBSCRIBE_START until the first group is produced for live tracks, but
  REQUIRE the publisher to send SUBSCRIBE_END immediately for an ended track that
  has no matching groups (or reset if it cannot accept the subscription);
  reference SUBSCRIBE_START, SUBSCRIBE_END, SUBSCRIBE_DROP and the Subscribe
  Stream in the new guidance so implementers know where to apply these timing
  expectations.
• Line 357: Update the sentence describing "Publisher Priority" so it refers to
  the lifetime of the Track rather than the subscription: find the line that reads
  "Publisher Priority is fixed for the lifetime of the subscription (see
  TRACK_INFO)" and change its wording to state that Publisher Priority is fixed
  for the lifetime of the Track (or: "fixed for the lifetime of the Track (see
  TRACK_INFO)"); ensure the phrasing matches TRACK_INFO's guarantee that fields
  are immutable for the Track and keep the parenthetical TRACK_INFO reference
  intact.
• Line 754: Clarify the ambiguous phrase "any such stream" by specifying that
  when a subscriber receives TRACK_INFO with an unrecognized compression value it
  MUST NOT open Subscribe (0x3) or Fetch (0x4) streams for that track, and if
  there are already open Subscribe/Fetch streams for that track they MUST be reset
  with a protocol violation; note that the Track Stream (0x6) from the publisher
  will already be FIN'd and does not need resetting. Update the sentence to
  reference TRACK_INFO, Subscribe (0x3), Fetch (0x4), and Track Stream (0x6)
  explicitly so readers know which streams to avoid opening and which existing
  streams to reset.
• Around line 311-318: The text currently says subscribers "SHOULD" discard
  cached TRACK_INFO after a duplicate active ANNOUNCE, which is unsafe; change
  this to a mandatory requirement and specify failure behavior: require that upon
  receipt of a new active ANNOUNCE the subscriber MUST discard any cached
  TRACK_INFO for that track and MUST re-request TRACK_INFO before parsing further
  FRAMEs, and if FRAMEs arrive that cannot be decoded due to stale Publisher
  Timescale or Publisher Compression, the subscriber MUST reset the stream (send
  RST) and re-fetch TRACK_INFO; alternatively (or additionally) introduce a stable
  version/epoch field in TRACK_INFO tied to the active ANNOUNCE so subscribers can
  detect staleness on-wire (reference: TRACK_INFO, ANNOUNCE, FRAME, Publisher
  Timescale, Publisher Compression).

---

Nitpick comments:
In @draft-lcurley-moq-lite.md:

• Around line 371-384: Update the two fenced code blocks showing priority
  examples so they include a language hint (e.g., change the triple backticks to

specifically modify the blocks containing "ali/audio + bob/audio:
subscriber_priority=2 publisher_priority=2" and the subsequent block starting
with "bob/audio: subscriber_priority=4 publisher_priority=2" to use ```text
fences instead of plain ``` fences.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

Run ID: a9437e4d-719b-4377-8d17-125e69458796

📥 Commits

Reviewing files that changed from the base of the PR and between 746f3ce and 1f0225a.

📒 Files selected for processing (1)

• draft-lcurley-moq-lite.md