Re: apply_scanjoin_target_to_paths and partitionwise join

Robert Haas <robertmhaas@gmail.com>

From: Robert Haas <robertmhaas@gmail.com>

To: Ashutosh Bapat <ashutosh.bapat.oss@gmail.com>

Cc: Richard Guo <guofenglinux@gmail.com>, Tom Lane <tgl@sss.pgh.pa.us>, Andrei Lepikhov <lepihov@gmail.com>, Jakub Wartak <jakub.wartak@enterprisedb.com>, arne.roland@malkut.net, PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>, Julien Rouhaud <rjuju123@gmail.com>, Etsuro Fujita <etsuro.fujita@gmail.com>

Date: 2025-12-24T13:28:23Z

Lists: pgsql-hackers

Thanks for writing back. I only just noticed this thread again.

On Wed, Dec 17, 2025 at 5:53 AM Ashutosh Bapat
<ashutosh.bapat.oss@gmail.com> wrote:
> My patch removed a redundant SET enable_partitionwise_join = on. That
> change is not included in your commit, I believe, because it's
> irrelevant to the fix. However, it's better to avoid the redundancy to
> avoid confusion. PFA patch for the same.

Yeah, I couldn't really understand your test case changes, so I
basically redid those from scratch. I personally think that the
regression tests are kind of horrible about how they use SET enable_*
commands. It's often extremely difficult to figure out what the
current values are at any given point in what may be a very long test
case file. I don't have an opinion at the present time on whether
changing this makes it more or less confusing.

> [3] https://www.postgresql.org/message-id/786.1565541557%40sss.pgh.pa.us

This is a really interesting link. Your email contained no reference
to this footnote (unless I missed something) but it seems quite
relevant to this discussion.

-- 
Robert Haas
EDB: http://www.enterprisedb.com