zpi1223 commented on pull request #3644:
URL: https://github.com/apache/iceberg/pull/3644#issuecomment-986367673


   > > > Hi @zpi1223! Thanks for the PR.
   > > > For your own information, If this is what I think you're talking about 
(that `MERGE INTO` is resolved by position), this has actually been fixed in 
Spark 3.2.
   > > > I have some unit tests here demonstrating that `MERGE INTO` will do by 
name resolution starting with Spark 3.2: #3440
   > > > Of course, Spark 3.2 support won't be generally available until 
Iceberg 0.13.0 is released, but wanted to let you know.
   > > 
   > > 
   > > Hi @kbendick , Thank you very much for your prompt reply. I personally 
think that these two are not conflicting.
   > > First,You are talking about the realization of functions, and what I 
want to express in this PR is to add certain warning information at the 
document level, so that the description information on the official website can 
be more complete, so as not to cause misunderstandings and affect the use in 
actual production.
   > > Secondly,Because we misunderstood the first time we used it, it took a 
certain amount of time to use it correctly at the end.
   > 
   > I don’t disagree. We should probably document it as well as it’s not 
clearly documented in the Spark docs (and people likely come to our docs for 
MERGE INTO information given that vanilla spark doesn’t implement merge into, 
just provide the primitives for us to do it).
   > 
   > But maybe the documentation can state that it’s a Spark bug and not an 
issue in Spark 3.2 and later?
   
   I still don't understand.
   First of all, the original intention of this PR is to improve the document 
information to avoid unnecessary understanding of the related Merge Into 
function when users use it, so that users can get a better experience.
   
   Secondly, I did not understand the Spark 3.2 bug that you explained. This PR 
is just a document-level information supplement. What does it have to do with 
Spark bugs?


-- 
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]



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

Reply via email to