Repository navigation
feat(cpp)!: export the C++ package as moq::cpp - #4297
Conversation
One moq:: namespace for every CMake package: find_package(moq-cpp) links moq::cpp (was moq-cpp::moq), matching moq::c from find_package(moq-c). The internal imported staticlib is moq::ffi. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
… reattach The reattached canvas still shows the last frame painted before the detach, and the relay may replay its cached group from then, so a baseline read before the detach let the stale canvas satisfy the resume check and the replayed group then read as the picture going backwards. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 0c094b8caf
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".
| // The reattached canvas still holds the last frame painted before the detach, and the relay | ||
| // may replay its cached group from before it too. Only a frame the fixture emits after this | ||
| // point proves the new session is playing. | ||
| const detached = (await readFixtureState(publisher)).frameId; |
There was a problem hiding this comment.
This hunk changes the browser detach/reattach test independently of the C++ CMake target rename, so applying, reverting, or backporting this commit also changes unrelated interop-test behavior. Keep this change focused by moving the media test fix to its own commit or PR. (Written by GPT-5.6 Sol)
AGENTS.md reference: AGENTS.md:L19-L19
Useful? React with 👍 / 👎.
…/cpp-moq-namespace
|
Summary: the diff is unchanged, still just the The Interop failure ( (Written by Claude Opus 5.5) |
Moves the C++ package's CMake target into a single
moq::namespace, which the maintainer chose on #4288:find_package(moq-cpp)now linksmoq::cpp(wasmoq-cpp::moq), both from the installed package and fromadd_subdirectory(cpp/moq).moq::ffi(wasmoq-cpp::ffi).find_package(moq-cpp)here, andfind_package(moq-c)withmoq::cin feat(c)!: publish the C bindings as moq-c #4288 ondev.Consumers updated: the
cpp/moqprobe test, the C++ interop client,cpp/moq/README.md, anddoc/lib/cpp.Public API and wire
cpp-v*release yet). Package, pkg-config file, headers, and C++ API are unchanged.Open PRs on this line that link
moq-cpp::moq(for example #4281, OBS on the generated C++) need to switch tomoq::cpp.Interop
This PR touches
test/interop, so Interop runs on it. Earlier runs failed on two issues, and neither came from this change:after reattach: presented frame went backwards). test(interop): reattach clears the stale canvas frame #4308 onmainand feat(cpp): moq::Error prints Rust's message #4292 on this line fixed it.python -> js/go -> jstiming out. The cause was the browser's track timeline scans: each arrival cost guards × retained groups, and at the FFI clients' 400 tone groups/s the page never caught up. perf(net): search the track timeline instead of scanning it #4321 fixed it onmain, and it reaches this branch through amainmerge into the line.Verified
just cpp check: it installs the package and builds the probe withfind_package(moq-cpp)andmoq::cppat C++17 and C++23.just checkpasses.just test interop --all: every C++ pair passes. The C client failure noted earlier (target/include/moq.hvsOUT_DIR) is gone after merging the line, which picked up the fix frommain.🤖 Generated with Claude Code
(Written by Claude Opus 5.5)