aglinxinyuan commented on PR #7785: URL: https://github.com/apache/texera/pull/7785#issuecomment-5479554235
Both advisories taken — head is now `a36189aa`. **The dead `implicit val timeout`.** Good catch, and the reasoning holds all the way down: with `ask` gone nothing in the class or in any of its three subclasses takes an implicit `Timeout`, and both surviving ask-pattern sites declare their own — `WorkflowActor.scala:90` for the `?` on `:93`, and `AmberClient.scala:61`. Removed the val, the now-orphaned `pekko.util.Timeout` import, and `DurationInt` from the duration import; `FiniteDuration` stays, it types five parameters. `implicit val self` and `implicit def ec` both stay, for the record: `self` is read explicitly at `PekkoActorRefMappingService.scala:40`, `:48` and `:111`, and `ec` supplies the executor for the `scheduleOnce` / `scheduleWithFixedDelay` overloads at `:48` and `:58` that take one. The compiler is the real check on this one — anything still resolving that implicit would now fail to build, which is the thing a grep cannot promise. `compile`, `Test/compile`, both `scalafmtCheck`s and `scalafixAll --check` pass. Eleven suites, `WorkflowActorSpec` included since it owns the production ask site: **169 succeeded, 18 failed** — and plain `57a4230` gives the identical 169/18, test for test. All 18 are `ExecutionResultServiceSpec` and `ResultExportServiceSpec` failing to open a relative local-filesystem Iceberg path on this machine. **The override count.** Right — three, not two; `parent` in `FailingParentActorService` was the one missing. The paragraph now names all three and adds that no subclass body needs an implicit `Timeout` either, which is what let the val go. Also folded the `timeout` removal into the evidence table, corrected the counts to 49 deleted / 3 inserted, and set the commit subject to the retitled PR so the two no longer disagree. Two notes, neither needing anything from you: round 2 posted twice 76 seconds apart (`5067255859` and `5067267715`, same `ai-review-diff` hash) — looks like a double fire. And GitHub now calls the branch BEHIND rather than conflicting; main moved four days while this sat, with no conflict. Say the word if you'd rather I rebase again than let the queue handle it. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
