rdblue commented on issue #28026: [SPARK-31257][SQL] Unify create table syntax URL: https://github.com/apache/spark/pull/28026#issuecomment-611035487 @gatorsmile, this isn't a new feature. This is fixing a long-standing problem that there are two different parser rules for create table that generate different plans. Now, there is one rule and additional metadata passed to conversions that already take place. Updating the existing paths for those properties is a reasonable part of the fix, instead of throwing exceptions. What you're asking for is to purposely disable passing only some of that metadata through one code path, but not the other. I've heard no rationale for that or discussion of risks introduced by passing it. What is the risk that justifies doing more work to remove passing a few metadata properties? If the only justification is that this is a "new feature" then should we also detect when Hive's `PARTITION BY` syntax is used for DataSource tables and throw an exception there? I think we all agree that would be silly. So what is the distinction between that and what you're asking?
---------------------------------------------------------------- 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. For queries about this service, please contact Infrastructure at: [email protected] With regards, Apache Git Services --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
