xuanyuanking commented on pull request #31638:
URL: https://github.com/apache/spark/pull/31638#issuecomment-788974534


   > The behavioral change is bound to file data source, right? I prefer adding 
the source option instead of adding config, because 1) Spark has a bunch of 
configurations already 2) I'd prefer having smaller range of impact, per source 
instead of session-wide.
   
   Agree. Especially for the 1), fully agree we should carefully to add new 
configs only when it's really needed. It should be my fault that didn't provide 
more context as mentioned in 
https://github.com/apache/spark/pull/31638#discussion_r585636401. Actually the 
same user code can pass in version 2.4 but fail now. If we add a source option, 
code changes is needed for them on controlling the behavior. It might not be a 
user friendly fix IMO. If this is a new behavior, I totally will follow your 
suggestion to add an option instead of config.
   
   > I'd love to see the proper notice on such case as well since we are here.
   
   Yes, thanks for the reminding. I'll add warning log for the cases you 
mentioned in proper places (like the `hasMetadata` return false) as well as SS 
user guides (do you think user guides is a also a reasonable place?). Do you 
prefer to do this in the current PR or a separated one?


----------------------------------------------------------------
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]



---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to