1. Draft 22 still waits.
Published: 21 · Planned: 22
Interim expected wire equivalence with Draft 20.
At the September 21 interim, the editors planned Draft 22 and expected it to remain wire- and semantically equivalent to Drafts 20 and 21. Datatracker still listed Draft 21 at the close of September 27 in Edmonton. A plan is not a published revision.
Source section: https://blog.rvnd.me/moqmusings/2026-09-21-moq-got-directions-and-extra-baggage#draft-22-waited-by-the-door
Primary sources: https://datatracker.ietf.org/meeting/interim-2026-moq-24/materials/slides-interim-2026-moq-24-sessa-moqt-issues-and-prs-921-00 https://datatracker.ietf.org/doc/draft-ietf-moq-transport/

2. moq.dev ships the tools.
DASH · File demux · Reconnect
Implementation releases, September 23.
The September 23 moq.dev release family includes libmoq 0.6.0, moq-relay 0.15.0, obs-moq 0.6.0, moq-cli 0.12.0, moq-gst 0.4.0 and moq-ffi 0.4.0. Release notes include timeline-based DASH manifests, file demuxing, frame-specific subscription and fetch bounds, and reconnect handling for relay migration. These are implementation releases.
Source section: https://blog.rvnd.me/moqmusings/2026-09-21-moq-got-directions-and-extra-baggage#moq-dev-divided-the-chores
Primary sources: https://github.com/moq-dev/moq/releases/tag/moq-relay-v0.15.0

3. Remember the route.
Hop IDs · Route cost
Cluster-01 proposes loop detection and path selection.
Cluster-01 is an individual draft for relay meshes. Advertisements carry the Hop IDs they traversed and accumulated route cost. The proposed mechanism supports loop detection, identifying advertisements from the same publisher and choosing between paths by cost. Cost need not mean the fewest hops.
Source section: https://blog.rvnd.me/moqmusings/2026-09-21-moq-got-directions-and-extra-baggage#the-relay-mesh-got-directions
Primary sources: https://datatracker.ietf.org/doc/draft-lcurley-moq-cluster/01/

4. Ask before advertising.
Solicit-00 · Individual proposal
Participating peers withhold unsolicited namespace advertisements.
Solicit-00 proposes a setup option declaring that namespace advertisements must first be solicited. A peer implementing the extension must honor that declaration. An unaware peer may still send unsolicited advertisements; the draft is not an adopted working-group mechanism.
Source section: https://blog.rvnd.me/moqmusings/2026-09-21-moq-got-directions-and-extra-baggage#namespace-advertising-learned-some-manners
Primary sources: https://datatracker.ietf.org/doc/draft-lcurley-moq-solicit/00/

5. Compress within a subgroup.
Flate-00 · Individual proposal · Objects · One stream
Separate DEFLATE contexts; MOQT wire unchanged.
Flate-00 proposes one raw DEFLATE stream per MOQT subgroup, with a sync flush at each object boundary. Objects share context within that subgroup. If an object is missing, the consumer abandons that subgroup; the next subgroup begins with a fresh context. Compression is in payloads and leaves the MOQT wire format unchanged.
Source section: https://blog.rvnd.me/moqmusings/2026-09-21-moq-got-directions-and-extra-baggage#namespace-advertising-learned-some-manners
Primary sources: https://datatracker.ietf.org/doc/draft-lcurley-moq-flate/00/

6. Broadcast checks a bag.
188-byte TS · 192-byte M2TS
MSFTS-01 proposes MPEG-TS packaging for MOQT.
The individual MPEG-2 Transport Stream packaging draft by Paul Gregoire and Gwendal Simon defines an mpeg2ts catalog value for 188-byte MPEG-2 TS and 192-byte M2TS source packets. Publishers map existing streams into MOQT objects, relays forward without parsing the transport stream, and subscribers reconstruct packets. It is a proposal, not a deployed compatibility guarantee.
Source section: https://blog.rvnd.me/moqmusings/2026-09-21-moq-got-directions-and-extra-baggage#mpeg-ts-checked-a-bag
Primary sources: https://datatracker.ietf.org/doc/draft-gregoire-moq-msfts/01/