Horton

Re: [18] Policy on IMMUTABLE functions and Unicode updates

Robert Haas <robertmhaas@gmail.com>

View thread
From: Robert Haas <robertmhaas@gmail.com>
To: Jeff Davis <pgsql@j-davis.com>
Cc: Peter Eisentraut <peter@eisentraut.org>, Laurenz Albe <laurenz.albe@cybertec.at>, pgsql-hackers@postgresql.org, Daniel Verite <daniel@manitou-mail.org>, Noah Misch <noah@leadboat.com>, Tom Lane <tgl@sss.pgh.pa.us>, Jeremy Schneider <schneider@ardentperf.com>
Date: 2024-07-24T19:19:55Z
Lists: pgsql-hackers 
On Wed, Jul 24, 2024 at 3:12 PM Jeff Davis <pgsql@j-davis.com> wrote:
> In any case, you are correct that Unicode updates could put some
> constraints at risk, including unique indexes, CHECK, and partition
> constraints. But someone has to actually use one of the affected
> functions somewhere, and that's the main distinction that I'm trying to
> draw.
>
> The reason why collation is qualitatively a much bigger problem is
> because there's no obvious indication that you are doing anything
> related to collation at all. A very plain "CREATE TABLE x(t text
> PRIMARY KEY)" is at risk.

Well, I don't know. I agree that collation is a much bigger problem,
but not for that reason. I think a user who is familiar with the
problems in this area will see the danger either way, and one who
isn't, won't. For me, the only real difference is that a unique index
on a text column is a lot more common than one that involves UPPER.

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