From: | Thomas Kellerer <spam_eater(at)gmx(dot)net> |
---|---|
To: | pgsql-hackers(at)postgresql(dot)org |
Subject: | Re: CTE inlining |
Date: | 2017-05-03 15:51:46 |
Message-ID: | 1493826706018-5959509.post@n3.nabble.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-hackers |
> I could tolerate telling people to use OFFSET 0 (and documenting it!)
> as a workaround if we can't get something more friendly in.
I agree with that.
> If we go with WITH INLINE then we're likely not solving anything, because
> most people will simply use WITH just like now, and will be subject to the
> fencing without realizing it.
I agree - the default behaviour should be change to match what everybody
expects. The current behaviour should be the exception.
> Yes, and we're missing the opportunity to confirm with what other
> systems do, and the spirit of the SQL language's declare what I want,
> not how to do it, model.
Essentially *all* other systems optimize CTEs the same way they optimize
derived tables. I think even MySQL does it like that in the upcoming 8.0
release.
I have never heard anyone saying that the Postgres implementation is an
advantage and that they would hate to see this disappear. I usually hear
"Why is Postgres doing that? Can't they change that?"
Maybe I have a limited view on this, but from where I stand, simply changing
it would help everybody I know and would not break anything. I don't even
think a replacement for the old behaviour would be necessary.
--
View this message in context: http://www.postgresql-archive.org/CTE-inlining-tp5958992p5959509.html
Sent from the PostgreSQL - hackers mailing list archive at Nabble.com.
From | Date | Subject | |
---|---|---|---|
Next Message | Robert Haas | 2017-05-03 15:52:47 | Re: password_encryption, default and 'plain' support |
Previous Message | Robert Haas | 2017-05-03 15:47:04 | Re: transition table behavior with inheritance appears broken (was: Declarative partitioning - another take) |