DanielLeens commented on PR #11932:
URL: https://github.com/apache/seatunnel/pull/11932#issuecomment-5645083558

   @SEZ9 Good news on the one condition you attached to the "known Zeta-only 
limitation" acceptance — it already holds on the current head, no new commit 
needed.
   
   I re-pulled `Collector.java` directly from `b6c96b97356d` just now rather 
than relying on my 2026-09-08 summary. The Javadoc is already on the 
`restoreSchema` default method, 
`seatunnel-api/src/main/java/org/apache/seatunnel/api/source/Collector.java:38-47`:
   
   ```java
   /**
    * Restores the collector's local schema before records are emitted from 
restored source state.
    *
    * <p>This method only refreshes local runtime state. It must not emit 
schema change events to
    * downstream operators because they restore their own checkpoint state 
independently.
    *
    * <p>Only the Zeta engine collector restores this state today. Flink and 
Spark retain the
    * default no-op implementation until their translation-layer collectors 
support schema
    * restoration.
    */
   default void restoreSchema(List<CatalogTable> catalogTables) {}
   ```
   
   That last sentence is exactly the note you're asking for — it says outright 
that only Zeta honours this today, so the connector-level restore in 
`IncrementalSourceReader` isn't going to be misread as working on every engine. 
This has been in place since before your 2026-09-11 ask; it's the same text I 
pointed to on 2026-09-08 (issuecomment-5578502768), so nothing regressed or got 
removed in between — I just hadn't re-confirmed it against this specific 
phrasing of the condition until now. My earlier reply flagging it to 
@davidzollo as an open ask was a miss on my part; it's already satisfied.
   
   So: condition holds, that thread can close once you've had a look yourself. 
Still open and unaffected by this: your three numbered asks (F1/F2 on 
`IncrementalSourceReader`'s empty-list/resolver-gating, F3/F6/F8 on 
`JdbcSink`'s restore behavior, F4/F5 on the docs/upgrade note and 
Serializable-payload question) and the Flink/Spark follow-up issue you're 
filing — those are unchanged by this and still need @davidzollo's response, 
since no new commit has landed on this PR since my last pass.
   


-- 
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]

Reply via email to