From: Tom Lane <tgl@sss.pgh.pa.us>
To: patrick.van.dijk@gmail.com
cc: pgsql-bugs@lists.postgresql.org
Subject: Re: BUG #18523: String compare not consistent
In-reply-to: <18523-d7b3cacb64d91cda@postgresql.org>
References: <18523-d7b3cacb64d91cda@postgresql.org>
Comments: In-reply-to PG Bug reporting form <noreply@postgresql.org>
	message dated "Wed, 26 Jun 2024 09:43:42 -0000"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <1515721.1719417844.1@sss.pgh.pa.us>
Date: Wed, 26 Jun 2024 12:04:04 -0400
Message-ID: <1515722.1719417844@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/1515722.1719417844%40sss.pgh.pa.us>
Precedence: bulk

PG Bug reporting form <noreply@postgresql.org> writes:
> When I try the following on a Azure/Linux version of PostgreSQL i see some
> strange results that are not correct.
> On Windows the behavior is correct.

"Correct" is in the eye of the beholder, and non-C collations
tend to act in ways far more complicated than you seem to expect.
Check the locale settings on the two machines; unless they are
both "C", there's not really any guarantee of consistency
between Windows and anybody else.

As a demonstration that this isn't some Postgres-specific
weirdness, you can try asking sort(1) what it thinks.
On a RHEL8 machine I get

$ echo -e '|\n0' | LANG=C sort
0
|
$ echo -e '|1\n01' | LANG=C sort
01
|1
$ echo -e '|\n0' | LANG=en_US.utf8 sort
|
0
$ echo -e '|1\n01' | LANG=en_US.utf8 sort
01
|1

and I'd expect the same from any Linux machine with a similar
vintage of glibc.

			regards, tom lane