[kixelated]

Backports the js/net change from moq-dev/moq#6504 to release, so it ships in the next @moq/net patch release and the interop runner's moq-dev-js client can bump to it without taking main's breaking API.

Problem

imquic drops a SUBGROUP_HEADER or FETCH_HEADER that arrives across more than one stream read (meetecho/imquic#43, acknowledged upstream, still open). @moq/net 0.4.6 on release writes the header one field per write(), which the @moq/web-transport polyfill sends as separate STREAM frames, so the runner's moq-dev-js data-subgroup-basic case gets no objects through imquic (draft-22, WebTransport).

Changes

• Group.encode and FetchHeader.encode write the stream type and header in one write, as on main. The fetch path no longer writes its stream type separately.
• Writer.collect and concat come over from main (moq-dev/moq#6255) unchanged, so the release-into-main merge stays clean. stream.ts is not exported from the package.
• ietf/object.test.ts carries fix(net): write IETF data stream headers in one write #6504's test; it fails on release without the fix (6 writes instead of 2).

Public API and wire impact

• Public API: none.
• Wire: none. Same bytes, fewer writes.

Verification

The runner's moq-dev-js client (builds/moq-dev-js/src/main.ts at upstream HEAD) against ghcr.io/englishm/moq-interop-runner-imquic-moq-relay:latest locally:

@moq/net	data-subgroup-basic
published 0.4.4	fails 2/2, all 15 objects missing
this branch	passes 3/3

The other cases are unchanged between the two: announce-subscribe times out waiting for the broadcast with both, so it is unrelated.

Checks

• just check origin/release: pass (@moq/net 1282 tests).
• just test interop --all: not run. The harness runs moq-lite, which this does not touch; the live imquic runs above are the coverage.

Choices

• Backported per CONTRIBUTING.md (broken interop), JS-only like moq-dev/moq#6518. The Rust publisher already buffers the header.

Tracked by the quest PR moq-dev/moq#6764.

(Written by Claude 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 2026-10-25T09:13:25.918640Z	d29a1aa	PR opened

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.

[kixelated]

Automated review by review (OpenAI)

Reviewed commit: d29a1aa

No actionable bugs found. This is a focused backport: buffering preserves the header encoding, copies reusable scratch buffers safely, and moves the FETCH stream type into the header encoder without duplicating it at the publisher. The regression checks both varint formats, write count and decoding. Coalescing writes is a reasonable interoperability mitigation, though it cannot guarantee transport read boundaries.

Verification: static review of the patch, writer lifecycle, publisher call path and original in-repository change. I did not run tests or independently verify the external imquic/runner results.