Message-ID: <3a2d42ac40a5f7aac315227aab374c693b701781.camel@cybertec.at>
Subject: Re: Documentation enancement regarding NULL and operators
From: Laurenz Albe <laurenz.albe@cybertec.at>
To: Luca Dametto <postgresqlmailing@damettoluca.com>, 
	pgsql-general@postgresql.org
Date: Wed, 18 Dec 2024 15:32:34 +0100
In-Reply-To: <emc054a8a5-7ede-476f-8bea-0558d655bbd1@6050b6ba.com>
References: <emc054a8a5-7ede-476f-8bea-0558d655bbd1@6050b6ba.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
User-Agent: Evolution 3.54.2 (3.54.2-1.fc41) 
MIME-Version: 1.0
Archived-At: <https://www.postgresql.org/message-id/3a2d42ac40a5f7aac315227aab374c693b701781.camel%40cybertec.at>
Precedence: bulk

On Sun, 2024-12-15 at 17:35 +0000, Luca Dametto wrote:
> Most programming languages return "true" when two null values are compare=
d,
> and false when, being the two values nullable, one of them is null and th=
e
> other one isn't.=20
>
> [not SQL, however]
>=20
> Whilst I'd love to discuss the reasons of this, I understand that it woul=
d
> be a waste of time for everyone, as we cannot change the status-quo even
> if it made sense, as it would break many thousands of projects.=20
>=20
> For that reason, I'd just like to improve the documentation to add at lea=
st
> a note about "hey, this won't work as you might expect, because it works
> in a different way than 99% of programming languages out there.".

There are already efforts to improve the documentation concerning NULL:
https://commitfest.postgresql.org/51/5086/
If that doesn't cover your case, perhaps you can join the discussion
and suggest improvements.

Yours,
Laurenz Albe