Message-ID: <081a16a7fcd3ec0bd17560ee5612655c1e04c78c.camel@j-davis.com>
Subject: Re: should check collations when creating partitioned index
From: Jeff Davis <pgsql@j-davis.com>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: Peter Eisentraut <peter@eisentraut.org>, pgsql-hackers
	 <pgsql-hackers@postgresql.org>
Date: Fri, 17 Nov 2023 13:08:09 -0800
In-Reply-To: <758153.1700252337@sss.pgh.pa.us>
References: <3327cb54-f7f1-413b-8fdb-7a9dceebb938@eisentraut.org>
	 <f3d6bcc63fdbd80d98ccbef975df5f0a55f58cd9.camel@j-davis.com>
	 <758153.1700252337@sss.pgh.pa.us>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.44.4-0ubuntu2 
MIME-Version: 1.0
Archived-At: <https://www.postgresql.org/message-id/081a16a7fcd3ec0bd17560ee5612655c1e04c78c.camel%40j-davis.com>
Precedence: bulk

On Fri, 2023-11-17 at 15:18 -0500, Tom Lane wrote:
> You keep harping on this idea that we are only concerned with
> equality,
> but I think you are wrong.=C2=A0 We expect a btree index to provide
> ordering
> not only equality, and this example definitely is a btree index.
>=20
> Possibly, with a great deal more specificity added to the check, we
> could distinguish the cases where ordering can't matter and allow
> collation variance then.=C2=A0 I do not see the value of that, especially
> not when measured against the risk of introducing subtle bugs.

Fair point.

As background, I don't see a complete solution to our collation
problems and on the horizon. You've probably noticed that I'm looking
for various ways to mitigate the problem, and this thread was about
reducing the number of situations in which we rely on collation.

I'll focus on other potential improvements/mitigations and see if I can
make progress somewhere else.

Regards,
	Jeff Davis