Re: Re: psql : \dn+ to show default schema privileges

From: Laurenz Albe <laurenz(dot)albe(at)cybertec(dot)at>
To: 노명석 <stan(dot)num(at)kakaocorp(dot)com>, pgsql-hackers(at)lists(dot)postgresql(dot)org
Subject: Re: Re: psql : \dn+ to show default schema privileges
Date: 2025-05-21 12:52:13
Message-ID: ac305e39929cc72b57fe52aa0c250ca848779ddc.camel@cybertec.at
Views: Whole Thread | Raw Message | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Wed, 2025-05-21 at 17:26 +0900, 노명석 wrote:
> You've raised valid concerns about hiding information (the actual NULL
> status of nspacl) and the potential for a (small) compatibility break
> if psql were to display default ACLs when nspacl is NULL. This leads
> me to think: if there's an information discrepancy when the actual
> value is NULL but the display shows the default ACL, then wouldn't
> explicitly adding the default ACL to nspacl during CREATE SCHEMA
> itself (instead of leaving it NULL by default) pose an even greater
> risk of more significant compatibility issues? I agree that this isn't
> something that can be changed easily.

The behavior is the same if there is a NULL or the explicit default
value in "nspacl". So the information that you are missing if you
don't get to see the default value is marginal - essentially that
someone has granted or revoked privileges on that object.

The change in the "psql" output (that might surprise experienced
users) is the bigger concern in my opinion. But it is not a very big
concern either.

> Separately, regarding my initial point in the first email about schema
> owner privileges not being included in the output of pg_dump – do you
> think it would be better to send a separate email to suggest adding an
> explanation for this to the documentation?

I think that is unnecessary. "pg_dump" doesn't need to preserve
everything literally, as long as the behavior is not changed.

But that's just my opinion.

Yours,
Laurenz Albe

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message David Steele 2025-05-21 13:24:12 Re: Regression in statement locations
Previous Message Fujii Masao 2025-05-21 12:34:20 Re: Make wal_receiver_timeout configurable per subscription