Sign in
All relays

MOQtail SyncPlay

MOQtail

syncplay.moqtail.dev:4448

WT draft 14 IPv4

A deployment of MOQtail. Everything below was measured on this deployment, at whatever version and configuration it was running — it is not a measurement of that codebase. About this endpoint

At a glance

Identifies itself as
did not say
Protocols accepted
h3
Media delivery
Namespace accepted, track never asked for
Violations recorded
1

Certificate

Subject
CN=*.moqtail.dev
Issuer
C=US, O=Let's Encrypt, CN=YR1
Valid until
2026-12-09 34 days left
Trust
Publicly trusted
Verification
Completed with the certificate verified

Control messages

Which control messages this endpoint answered, per draft. An error reply is still an answer — ok and err are both support; what is recorded as missing is a message that got no answer at all. A dash is neither: it means the exchange ended on our side, so nothing about this endpoint was measured there.

Draft SUBSCRIBEFETCHTRACK_STATUSSUBSCRIBE_NAMESPACEPUBLISH_NAMESPACESUBSCRIBE_TRACKSPUBLISH
14 WT ✓ err ✓ err ✗ ✗ ✓ ok · ✗

Media delivery

One publisher and two subscribers, to see whether an object published through this endpoint reaches a client that asked for it.

Namespace accepted, track never asked for draft 14 · WebTransport

The subscription routed to the publisher could not be read. The endpoint broke a rule of the draft: each new request must carry the next Request ID in the sender's sequence.

Two publishers of one track

Two publishers announce the same namespace and publish the same tracks, one with identical payloads and one with payloads that differ by publisher. Partway through, the publisher serving the subscribers disconnects without notice. A fresh subscription, sent afterwards, shows whether the remaining publisher could still be reached.

The relay closed a serving publisher’s session early draft 14 · WebTransport
Publisher Role Announcement Subscribed to Ended by the relay
first publisher closed The relay closed its session while the other publisher was present. accepted, PUBLISH_NAMESPACE_OK in 26 ms — Session ended: an unreadable reply
second publisher standby The relay accepted it and never subscribed to it while a subscriber was waiting. accepted, PUBLISH_NAMESPACE_OK in 26 ms — —
Track Objects From When the publisher left Ended by
identical payloads nothing arrived — not measured —
differing payloads nothing arrived — not measured —

Specification violations

Each new request must carry the next Request ID in the sender's sequence.

Draft 14 over WebTransport, during the delivery run, routing the subscription

Specification
draft-ietf-moq-transport-14, Section 9.1
Required response
close the session with INVALID_REQUEST_ID (0x4)

Setup parameters

What this endpoint sent back in its setup response, per draft. A parameter with no name is a code point the negotiated draft does not define.

draft 14 · WT

  • 0x02 max_request_id = 2305843009213694000

History

9 sweeps, newest first. A step where the probe build or the measuring location changed is marked, and is not read as a change of the relay.

  • 27 Oct 2026 → 29 Oct 2026 · measured differently

    • Offering a track: not measured → not attempted.
SweepReachableDraftsDeliveryOffers a trackViolationsTwo publishers
5 Nov 2026yes14Namespace accepted, track never asked fornot attempted1no defects
4 Nov 2026yes14Namespace accepted, track never asked fornot attempted1no defects
3 Nov 2026yes14Namespace accepted, track never asked fornot attempted1no defects
2 Nov 2026yes14Namespace accepted, track never asked fornot attempted1no defects
1 Nov 2026yes14Namespace accepted, track never asked fornot attempted1no defects
31 Oct 2026yes14Namespace accepted, track never asked fornot attempted1no defects
30 Oct 2026yes14Namespace accepted, track never asked fornot attempted1no defects
29 Oct 2026yes14Namespace accepted, track never asked fornot attempted1no defects
27 Oct 2026yes14Namespace accepted, track never asked for—1—

Notes

  • — The endpoint is up and refused the offered protocols on at least one transport.

Is this your relay?

If something here is wrong, tell us — we keep what each result rests on, so a disputed one gets re-read rather than re-argued. We can also run this against endpoints that are not public.