HeartSaVioR commented on issue #28026: [SPARK-31257][SQL] Unify create table syntax (WIP) URL: https://github.com/apache/spark/pull/28026#issuecomment-607532712 I agree turning on the legacy config is just a workaround which I will never recommend end users to turn off. I expect I would fail to explain to end users why this query creates parquet provider but other query creates Hive provider, and I guess that's not only me. (You may be able to understand although such complication but that's different story if you have to make others understand.) If I remember correctly, the Spark community practice for similar case has been "revert". There's no point of keep the commit be live as it's known to be bad and we will have better patch in near future. It only makes sense if someone disagrees the commit is bad; is that a case? Or do we want to let some individuals/teams/vendors to manage it with risks? ps. Honestly, I feel really odd we allow turning on legacy config by default - not for this issue but for other issues - end users may never indicate the fix, although there should be the valid reason the fix is landed.
---------------------------------------------------------------- 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]
