diff --git a/Cargo.lock b/Cargo.lock index a0070ca43ee..f46c261dac2 100644 --- a/Cargo.lock +++ b/Cargo.lock @@ -11131,9 +11131,9 @@ dependencies = [ [[package]] name = "web-transport-proto" -version = "0.6.2" +version = "0.6.3" source = "registry+https://github.com/rust-lang/crates.io-index" -checksum = "1a81bcc79bf48ad0b819bfa51ec28524ee7c356056d43c316727d6b505dcc71e" +checksum = "d2baf87f7a4701d964fd23529b440db3da535b1d7539289f7a463c3481dafcff" dependencies = [ "bytes", "http", diff --git a/quest/m0/README.md b/quest/m0/README.md index b4807b18949..9a2a5c415a9 100644 --- a/quest/m0/README.md +++ b/quest/m0/README.md @@ -41,7 +41,6 @@ Each lands its fixes in its own PR and plans only what needs a decision. ## Required - [Broadcast epochs](/quest/m0/broadcast-epoch/README.md) - every first-party publisher that can restart mints a fresh route epoch, the newest wins a path, and only routes with the same epoch resume a subscription -- [web-transport-proto 0.6.3](/quest/m0/web-transport-proto-0.6.3.md) - `main` and `release` lock the upstream release that bounds decoded CONNECT headers, so a peer can't make a relay hold megabytes per session before auth - [64-bit Hop IDs](/quest/m0/hop-id-64.md) - a moq-transport peer with any non-zero 64-bit Hop ID keeps its session, and Hop IDs are `u64` in Rust and JS - [MALFORMED_AUTH_TOKEN session code](/quest/m0/malformed-auth-token-code.md) - a 0x16 close reports a malformed token instead of `unknown code=22`, and moq-net sends it for an unparseable token - [qmux 0.5.3](/quest/m0/qmux-0.5.3.md) - crates.io publishes qmux 0.5.3 with the builder session config the release backport of #5609 needs diff --git a/quest/m0/web-transport-proto-0.6.3.md b/quest/m0/web-transport-proto-0.6.3.md deleted file mode 100644 index 08fea551ed2..00000000000 --- a/quest/m0/web-transport-proto-0.6.3.md +++ /dev/null @@ -1,35 +0,0 @@ -# [XS] Take web-transport-proto 0.6.3 - -## Goal - -The `main` and `release` lockfiles resolve web-transport-proto 0.6.3, which -bounds decoded CONNECT headers to a 16 KiB field section, counting each field -as name + value + 32 bytes. Until then, a peer can make moq-tokio, moq-uring, -or the iroh backend hold about 2.7 MB per session from one 64 KiB CONNECT, -before auth: unique short literal names amplify the frame about 42x (the -[QPACK header accordion](https://foxio.io/blog/header-accordions-qpack-amplification-attacks-against-http-3-servers) -class, without its dynamic-table variant, which the decoder refuses). - -Waiting on moq-dev/web-transport's -[Bound decoded CONNECT headers](https://github.com/moq-dev/web-transport/blob/main/quest/a0/qpack-field-section.md) -quest. Check: `cargo info web-transport-proto@0.6.3` resolves. - -## Plan - -Decided 2026-10-21 (maintainer): - -- **Fix upstream, not in-tree.** The iroh backend decodes CONNECT inside - web-transport-iroh with the same crate, so only an upstream release covers - every backend, and web-transport-proto stays upstream - ([Drop web-transport-trait](/quest/m2/drop-web-transport-trait.md) records - why). This is the one exception to taking no work in moq-dev/web-transport. -- **Lock-only bumps on both lines.** `cargo update -p web-transport-proto` in - a PR against `main` and a backport PR against `release`, whose - `web-transport-proto = "0.6"` also takes 0.6.3, so moq.dev's relay gets the - bound at its next patch release. No `Cargo.toml` change. -- **No moq-side test.** The upstream unit test covers the decoder. Confirm - both servers' SETTINGS now carry `MAX_FIELD_SECTION_SIZE`; they build them - with `Settings::enable_webtransport`, which 0.6.3 extends. -- moq-uring's 64 KiB `MESSAGE_LIMIT` stays: it also bounds SETTINGS, and - 0.6.3's sync decoder refuses an oversized HEADERS frame at its declared - length, before the payload is buffered. diff --git a/quest/m1/release-audit/README.md b/quest/m1/release-audit/README.md index 7fa6e8165ff..cc352ed7572 100644 --- a/quest/m1/release-audit/README.md +++ b/quest/m1/release-audit/README.md @@ -101,7 +101,6 @@ child records whether its behavior is new on main or inherited from release. - [JS subscription ranges](/quest/m1/subscribe-ranges/js.md) - range lists and frame offsets build on the bounds fixed here - [OBS video source](/quest/m2/obs-moq-video/source.md) - reproduced inherited timing and EOF defects remain with the decoder replacement owner -- [QPACK dependency fix](/quest/m0/web-transport-proto-0.6.3.md) - decoded header expansion is already tracked - [Compressed tracks](/quest/m1/compress/README.md) - mid-group widening and application migration have existing owners - [Archive consumer](/quest/m1/archive/consumer.md) - current producer and CLI reader still need the new-format roundtrip - [Measured release performance](/quest/m1/release-perf/README.md) - existing paired benchmark regressions and their confirmation runs diff --git a/quest/m2/drop-web-transport-trait.md b/quest/m2/drop-web-transport-trait.md index bada1cf0ebd..58becb0a6c2 100644 --- a/quest/m2/drop-web-transport-trait.md +++ b/quest/m2/drop-web-transport-trait.md @@ -50,7 +50,7 @@ from that repo as accepted exceptions, and web-transport-proto may stay. Decided 2026-10-21 (maintainer): web-transport-proto stays upstream, with no migration quest. web-transport-iroh decodes CONNECT with it, so an in-tree copy would duplicate the crate whenever iroh is on. Its QPACK bound ships as -an upstream patch ([web-transport-proto 0.6.3](/quest/m0/web-transport-proto-0.6.3.md)). +the upstream web-transport-proto 0.6.3 patch. ## Required