Horton

Re: pgsql: Revoke PUBLIC CREATE from public schema, now owned by pg_databas

Robert Haas <robertmhaas@gmail.com>

View thread View on postgr.es
From: Robert Haas <robertmhaas@gmail.com>
To: Noah Misch <noah@leadboat.com>
Cc: "pgsql-hackers@postgresql.org" <pgsql-hackers@postgresql.org>
Date: 2022-11-30T15:44:12Z
Lists: pgsql-hackers 
On Wed, Nov 30, 2022 at 10:01 AM Noah Misch <noah@leadboat.com> wrote:
> On Wed, Nov 30, 2022 at 08:39:23AM -0500, Robert Haas wrote:
> > On Wed, Nov 30, 2022 at 2:07 AM Noah Misch <noah@leadboat.com> wrote:
> > > In general, the documentation should prefer simpler decision trees.
> >
> > True, but I found the current text confusing, which is also something
> > to consider.
>
> Could remove the paragraph about v14.  Could have that paragraph say
> explicitly that the REVOKE is a no-op.  Would either of those be an
> improvement?

Well, I thought what I proposed was a nice improvement, but I guess if
you don't like it I'm not inclined to spend a lot of time discussing
other possibilities. If we get some opinions from more people that may
make it clearer which direction to go; if I'm the only one that
doesn't like the way it is now, it's probably not that important.

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

Commits

  1. Doc: word-smith the discussion of secure schema usage patterns.

  2. Fix the public schema's permissions in a separate test script.

  3. Revoke PUBLIC CREATE from public schema, now owned by pg_database_owner.