SEZ9 commented on PR #11885: URL: https://github.com/apache/seatunnel/pull/11885#issuecomment-5611764700
@DanielLeens thanks for re-fetching the review body and restating the F1–F8 mapping against `befad345c` — that clears up the rendering confusion, and the explanations make sense to me. One thing before I mark items closed: your comment describes what the diff contains, but I still need to verify each point against the actual changes at `befad345c` rather than the restatement alone. Could you point me at the concrete locations so I can check them quickly? - **F1/F6** — the test method in `IncrementalPhaseStateTest` that deserializes bytes produced *without* `stopOffset` (the pre-upgrade checkpoint case), and the round-trip test that goes through real serialization. - **F2** — the `createIncrementalSplit()` hunk with the `resolvedStopOffset == null` guard, and where `HybridSplitAssigner` guarantees incremental splits are only created after snapshot completion. - **F5** — the `resolveLatestStopOffsetWithRetry()` hunk (attempt count / backoff). - **F3** — the `runLatestStopStartupMode` assertion that id=2000 is not yet present after id=1 arrives. - **F4/F7/F8** — the doc hunks for the `latest` snapshot-completion semantics, and please make sure the docs checkbox in the PR description is ticked so the checklist matches the diff. Once I can see those hunks I'm happy to close everything out — nothing in your explanation looks wrong to me, I just want the closure to rest on the diff rather than the summary. <!-- streview-comment:931 --> -- 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]
