FANNG1 commented on PR #10721:
URL: https://github.com/apache/gravitino/pull/10721#issuecomment-4281104094

   I wonder if we may be putting a bit too much Paimon-specific logic into the 
Flink connector here.
   
   Would it make sense to keep the Flink connector thinner and let it do only 
the minimum adaptation work, e.g. extracting  /  from table properties and 
converting them into a , while leaving provider-specific semantics such as PK 
fallback,  handling, and bucket-number validation to the Gravitino server side 
()?
   
   My concern is that once this logic lives in the connector, we may end up 
duplicating the same Paimon semantics again for Spark / Trino later, and it 
also increases the chance of semantic drift between connector-side and 
server-side implementations. Curious what you think about centralizing that 
part on the server side.


-- 
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