Github user icexelloss commented on the issue: https://github.com/apache/spark/pull/21427 I do think the current default behavior might be confusing to users and hard to debug. I have also received similar complaints. I think at the very least, we should make sure when column names of the schema and return value matches but orders are different, we should match by column name as it is extremely unlikely user want any other behavior in this case. This will mostly keep the current behavior unchanged, with the exception that "same column name, different order" which the new behavior is strictly better.
--- --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org