Repository navigation
test: give the control-error status cases room for a stalled runner - #82
Merged
Merged
Conversation
"does not treat an HTTP <code> control error as a stopped browser" gave getChatGptBrowserStatus a 300 ms budget against a server that answers at once. On macOS Intel CI the runner stalled past that three times (2026-10-28 and 10-29; the tests took 336-1622 ms), the fetch timed out, and the status read browser_slow instead of browser_control_unavailable, failing main and PR runs. The budget is not what these cases test, so it is now 10 s. A scratch test that blocks the event loop for 500 ms right after the call gives browser_slow at 300 ms and browser_control_unavailable at 10 s. Verified: tsc; the file passes (8 tests). Not reproduced by CPU load alone locally (12 busy processes, or pinning to one CPU). Claude-Session: https://claude.ai/code/session_01EMqNwnnhHXznd2P2PDsaDV
What this touches
1 file(s), 5 line(s) changed. Tests changed. This is an automated first pass, not a review. CI is the gate; a person decides. |
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. |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Problem
tests/browser-slow-vs-dead.test.ts> "does not treat an HTTPcontrol error as a stopped browser" gavegetChatGptBrowserStatusa 300 ms budget against a local server that answers at once. On macOS Intel CI the runner stalled past that, the fetch timed out, and the status readbrowser_slowinstead ofbrowser_control_unavailable. This failed three times:Change
Test only. These cases now use a 10 s budget, because the budget is not what they test. The slow-port case still uses 300 ms against a server that waits 2 s, and the tiny-timeout case still uses 10 ms. No CHANGELOG entry: nothing user-visible changes.
Verification
browser_slow, the CI failure;browser_control_unavailable.tsc. The file passes (8 tests).