Repository navigation
fix(ios): disclose whether a canceled read-only runner send was written - #3409
Conversation
Follow-up to #3385. The read-only route (waitForRunner) threw a bare canceled error, so a cancellation that landed before any connect attempt was charged and drained with one status poll instead of withdrawn. It now discloses `dispatched: no` when no attempt could have written the command, and `unknown` once an attempt was interrupted mid-flight or an earlier attempt may have written it. Pins two exchange behaviors with tests: a refused connection withdraws its charge whether or not the request was canceled (the runner never received the command), and a drain whose status read fails stops at once with the charge left abandoned. Refs #3383 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
![cubic-dev-ai[bot]](https://avatars.githubusercontent.com/in/1082092?s=60&v=4){kind=small}
There was a problem hiding this comment.
All reported issues were addressed across 5 files
Reply with feedback, questions, or to request a fix.
View guided diff | Turn on auto-fix | Re-trigger cubic
…ing was written Each route of the read-only connect loop replaced its failure with a bare canceled error, so a cancellation whose own failure proved nothing was written (a usbmux connect still in progress, a refused fetch, a simulator curl that could not connect) reported `dispatched: unknown`. Routes now keep `no` when that failure is the proof, and the final simulator post combines the loop's write evidence with its own failure the same way. An abort that interrupts a POST, or the request's shared abort reason, still reads as possibly written. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Size Report
Startup median (7 runs, lower is better):
|
|
The code is ready at f1a6cfd. The change only touches the cancel branches of waitForRunner, so it reports whether a canceled read-only send was written without adding any device request. No conflicts. Smoke Tests is still running and has no captured failure, and a failure there would be unlikely to come from this diff, because an uncanceled run never reaches the changed branches. Smoke Tests must finish green on f1a6cfd before merge. Not blocking, and you can take or leave these: the "drain whose status read fails" and "refused connection withdraws its charge" tests cover code already on main (the break at runner-exchange.ts:362 and the withdrawUnsent branch at :318), so one executeRunnerExchange test could cancel a read-only command during the retry delay and assert the charge is withdrawn with no status poll; the PR body names only the read-only route, but waitForRunner also serves the readiness preflight (runner-exchange.ts:425) and graceful shutdown (runner-disposal.ts:159), and a canceled preflight now surfaces "no" or "unknown" instead of undefined; and the "possibly written unless provably unwritten" rule lives in both setLastError and the new .catch in runner-startup-transport.ts (:103-108), so one local helper could own it. The two cubic-dev-ai threads, #3409 (comment) (final simulator post) and #3409 (comment) (usbmux connect-phase "no"), no longer apply at this head, so please resolve them. I read the code and the pre-change behavior at 6e80737. I did not run the tests and did not do a live run. I did not exercise the reachable window where a cancel lands in the retry delay after ECONNREFUSED while the runner starts. |
thymikee
added
the
ready-for-human
|
1 participant
Summary
Follow-up to #3385 (#3383) with the reviewer's three notes:
runner-startup-transport.ts):waitForRunnerused to throw a bare canceled error. A cancellation before any connect attempt was therefore charged and drained with one status poll. It now disclosesdispatched: 'no', and the exchange withdraws the charge. An interrupted attempt now counts as possibly written, so a send that was already in flight still drains. The reviewer's suggested one-liner alone would have missed that.status_unavailableexit: a test where status answersstartedand then the runner exits shows the drain stops at once, leaves the charge abandoned, and reportsstatus_unavailable.5 files touched, all in
packages/platform-apple/src/runner.Validation
Commit
f1a6cfd70.pnpm check:affected --runpassed (3182 tests). Each connect route keepsdispatched: noonly when its own failure proves nothing was written.eager-closure-budgets.test.tspassed. An earlier run timed out in two interaction-contract tests while host load was about 220. Both pass alone and passed on the rerun.Each new test failed with its fix reverted: the dispatch disclosure, the canceled-only withdraw, and the missing
break.No live device run: these are unit-level changes to error disclosure and charge accounting, with no new device path.
🤖 Generated with Claude Code