jayzhan211 commented on PR #14268: URL: https://github.com/apache/datafusion/pull/14268#issuecomment-2628583804
> > I don't have a strong opinion about how exactly the coercion rules should be changed. > > > > > > > > It does seem to me that we keep churning / thrashing on coercion (and introducing regressions like the one this PR fixes along with https://github.com/apache/datafusion/issues/14154 and https://github.com/apache/datafusion/issues/14383 > > > > > > > > To me this is a sign we need more work as @jayzhan211 says and I am sure we would appreciate help figuring out what to do / making the code better > > > > > > > > That being said, I would like to make the 45.0.0 release candidate soon (like later today or tomorrow) and I worry that making non trivial changes to the coercion logic so close may lead to other unintended consequences > > > > > > > > Is there any way we can fix the regression bug with some sort of smaller patch and then make the changes to coerceable behavior as a follow on PR? > > > > > > @alamb I'm not near a computer until later tonight, but just lemme know what you'd like me to do and i'll take care of it when I'm back home! So what is the actual regressions to be resolved? I don't remember signature coericible is consistent with the docs before so it is not a regression to be fixed now. If supporting binary for string is something used to work, then we can add such coercion logic to singature:: sitting -- 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: github-unsubscr...@datafusion.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: github-unsubscr...@datafusion.apache.org For additional commands, e-mail: github-h...@datafusion.apache.org
