Github user holdenk commented on the issue:
https://github.com/apache/spark/pull/15367
@marmbrus @zsxwing I agree its experimental and we should have more
flexibility here with backports. I also very much agree that structured
streaming in its current state on 2.0 isn't usable - but I'm not super sure
that backporting fixes is the best way to do this? Honestly I spend most of my
time focused on Python & ML (and I've only really been looking at structured
streaming with those two hats on).
I'm really cautious about the idea 2k+ line backport which hasn't even been
released otherwise but I don't have any specific objections to the changes its
just making me nervous. The fact the whats being backported seems to still be
under development is also concerning since doing this backport now puts us in a
position of backporting more (not yet merged into mainline) fixes.
Of course - If the people with the most experience in this area all agree
(and most of y'all [ @marmbrus @zsxwing @tdas but maybe missing @koeninger ]
seem to already be on this PR so I'll leave you to it) that this backport
reasonable that is great - it would probably be good to follow up to the
original backport mailing list thread and update the wiki as well.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
