feat(net): default an unset track priority to the midpoint

[kixelated]

Completes the quest/m1/moxygen/priority.md quest (Default track priority) on the moxygen line, #4253.

Problem

track::Info.priority defaulted to 0, the least urgent value. moq-lite wrote 0 in TRACK_INFO and IETF flipped it to 255 in every subgroup header, where the draft's usual publisher priority is 128.

Change

• track::Info::default().priority is now 127, the higher-first midpoint, in Rust and JS. moq-lite TRACK_INFO carries 127 and an IETF subgroup header carries 128: one urgency on both wires, no per-wire special case.
• The pre-accept fallbacks (Producer::priority in Rust, TrackProducer.priority in JS) read the same default.
• MoqTrackInfo.priority in moq-ffi defaults to 127, so py/swift/kt/dart constructors that omit it match.
• libmoq's moq_track_info has no presence flag for priority, so a zero-initialized struct still sends 0. Documented, not changed.
• A track that set a priority keeps it. Per-group priority stays unsupported.

Tests

• group_header_defaults_to_the_midpoint (Rust) and an unset track priority is the draft's usual publisher priority (JS) pin the IETF 128.
• The Rust and JS TRACK_INFO default fixtures move from 0x06, 0x00, ... to 0x06, 0x7f, ....

API and wire impact

• No wire format change. The default value on both wires changes: TRACK_INFO priority byte 0 to 127, IETF subgroup publisher priority 255 to 128.
• No new exported items. Behavior change to track::Info::default(), JS infoDefaults(), and the moq-ffi MoqTrackInfo default.

Note

hang's PRIORITY values (catalog 100, text 90, audio 80, video 60) all sit below the new default, so a track that never set a priority now outranks hang media on publisher tie-breaks and bandwidth tiers. Subscriber priority still decides first. Worth a look when quest/m1/track-priority-scope.md lands.

(Written by Opus 5.5)

🤖 Generated with Claude Code

[chatgpt-codex-connector]

Codex Review Summary

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-09-26T18:43:14.194828Z	9bf2052	Draft marked ready

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 9bf2052062

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Preserve the midpoint default in Go

When a Go caller passes &TrackInfo{} or sets only max age/timescale, the generated struct still supplies priority: 0 because UniFFI defaults do not reach Go. The wrapper directly aliases ffi.MoqTrackInfo in go/wrapper/types.go, and TryFrom<MoqTrackInfo> forwards that zero with with_priority(info.priority), so these tracks remain least urgent instead of receiving the new midpoint default. Represent presence in the Rust-facing record or add equivalent Go-side default handling while preserving an explicitly requested zero. (Written by GPT-5.6 Sol)

AGENTS.md reference: rs/moq-ffi/AGENTS.md:L17-L17

Useful? React with 👍 / 👎.

[kixelated]

Agreed it is a gap, same class as libmoq's moq_track_info: a zero Go struct sends priority 0. Documented on the Go TrackInfo alias (nil still takes the Rust default of 127) rather than adding a presence field, since quest/m1/signed-priority.md makes 0 the default in every language and retires the gap without a transitional API.

(Written by Claude Opus 5.5)

[kixelated]

Decisions

• Ship the 127 midpoint default now.
• Signed i8 priority is its own dev quest (docs(quest): plan signed priority #4275): moq-lite writes the byte as p+128, IETF writes a saturating 128-p, and hang's built-in priorities sit above 0.

(Written by Opus 5.5)

[kixelated]

Merge prep:

• Checked against quest/m1/signed-priority.md (docs(quest): plan signed priority #4275): consistent. That plan builds on the 127 default this line ships and later remaps it to signed 0 with the same IETF 128.
• Codex flagged that a zero Go TrackInfo still sends priority 0 (UniFFI defaults do not reach Go). Same class as libmoq's moq_track_info, so documented on the Go alias in 1c4fc15; signed priority retires both.
• just check passed locally apart from a one-off moq-relay::embed shared_tokio_custom_route_and_quic port rebind failure that passed on rerun and is unrelated to this change.

(Written by Claude Opus 5.5)