docs(voice): clarify client transport and proxy ownership - #3969
Conversation
|
Important Review skippedAuto reviews are disabled on base/target branches other than the default branch. 🗂️ Base branches to auto review (2)
Please check the settings in the CodeRabbit UI or the ⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: ASSERTIVE Plan: Team Run ID: You can disable this status message by setting the Use the checkbox below for a quick retry:
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
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. |
리뷰 · 우선순위 48 / 80이 PR은 Codex 음성 클라이언트와 OpenCodex 프록시의 소유 경계를 무엇을 말하나. Codex가 마이크·스피커·WebRTC·캡션·mute·스레드 전환 cleanup을 갖고, OpenCodex는 call-create와 sideband 릴레이·위임된 텍스트의 일반 Responses 라우팅만 맡는다는 점. 텍스트 provider를 고른다고 realtime speech 모델이 바뀌거나, voice 미지원 클라이언트에 음성이 켜지지 않는다는 점. upstream Codex 커밋 두 개(WebRTC helper / TUI voice)는 클라이언트 책임을 설명하는 참고이고, merge 날짜가 데스크톱 앱 출시 시점을 증명하지 않는다는 점. 선택 진단 왜 점수가 구현 PR보다 낮은가. 코드 경로·동작은 #3968이 바꾸고, 이 PR은 가이드 한 파일(
메인테이너의 판단이 필요한 지점
너의 추천 이 댓글은 grok-bot이 작성했습니다 |
|
✅ Deterministic PR hygiene checks passed. |
|
Independent documentation-build evidence for the exact cumulative head 3468a2d: the Astro build completed successfully with 425 pages in an isolated checkout. I reused a dependency installation with the identical frozen docs lockfile and used fresh HOME/OPENCODEX_HOME/CODEX_HOME, reduced CPU priority, and an eight-core cap. No install, deployment, microphone test, service restart, or production configuration write was performed. Existing chunk-size and missing 404-content warnings remain. This fills the non-deploying build gap described in the PR, not the runtime or live-audio validation gap. Full cumulative CI run 34172698360 is still the separate exact-head gate; cancelled per-layer jobs are not passing evidence. Keep the parent-first landing order for #3968 and this child. This comment is build evidence only, not final approval of the entire voice stack. |
Summary
Document Codex's ownership of local audio, WebRTC, captions and spoken handoffs, and OpenCodex's ownership of call creation, sideband relay and delegated text routing. Explain metadata-only diagnostics and why service-tier speed descriptions do not establish voice latency or a desktop rollout date.
Manual stack: depends on #3968. This cumulative head includes the runtime fix; retarget to
devafter the parent lands and merge bottom-up.Verification
16ddf43c92f7cfa164a6bcb0d80101abc9450b81, based on dev900567af3.git diff --check: PASS. No unresolved review threads.Maintainer integration decision: acting as
lidge-jununderMAINTAINERS.md, integrate intodevwithout a second maintainer approval after the parent lands. Revalidate current head/base/actor and prospective merge-tree equality with the verified cumulative tree. This is maintainer integration, not self-approval.Checklist
Landing verified: #3968 merged as
ac84ab2c5cf767d1ca7cd5612cdc80348b1e6139; #3969 merged as9e1468d4b7a41b498ed2aca98507ada2c741afea. Fetched dev contains the tested cumulative head, and both trees are exactly68b6eb5a5644b19e321be916de8b4d6429174648. The only merge-generated duplicate product CI (parent push run 34176133959) was cancelled; it is not verification evidence.