From: Tom Lane <tgl@sss.pgh.pa.us>
To: lacanoid@ljudmila.org
cc: pgsql-bugs@lists.postgresql.org
Subject: Re: BUG #18998: No materialized views in
 INFORMATION_SCHEMA.TABLE_PRIVILEGES
In-reply-to: <18998-d0b1ca418d70eed4@postgresql.org>
References: <18998-d0b1ca418d70eed4@postgresql.org>
Comments: In-reply-to PG Bug reporting form <noreply@postgresql.org>
	message dated "Thu, 24 Jul 2025 17:50:36 -0000"
MIME-Version: 1.0
Content-Type: text/plain; charset="UTF-8"
Content-ID: <1746230.1753391930.1@sss.pgh.pa.us>
Content-Transfer-Encoding: 8bit
Date: Thu, 24 Jul 2025 17:18:50 -0400
Message-ID: <1746231.1753391930@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/1746231.1753391930%40sss.pgh.pa.us>
Precedence: bulk

PG Bug reporting form <noreply@postgresql.org> writes:
> View NFORMATION_SCHEMA.TABLE_PRIVILEGES contains information for views and
> tables, but not materialized views. I found this to be somewhat unexpected.
> Since materialized views are conceptually somewhere between views and
> tables, it would perhaps make sense to include them.
> The standard of course doesn’t say anything about this because it does not
> deal with materialized views at all.

Yeah, precisely.  Our take on the information schema is that it should
show what the standard says, and thus objects that are outside the
standard should not appear there at all.  Obviously that's a bit of
a judgment call, but it saves us from having to make a bunch of other
judgment calls about just how much we'd want to warp the information
schema's behavior.

For example, in the case at hand it'd be darn weird for matviews
to be reflected in TABLE_PRIVILEGES but not TABLES, so then we'd
have to allow a nonstandard value of TABLES.TABLE_TYPE, and then we're
already some way down the slippery slope to something that claims to
be the information schema but contains much that a standards-compliant
client wouldn't know what to do with.

			regards, tom lane