Re: GRANT ON ALL IN schema

From: Brendan Jurd <direvus(at)gmail(dot)com>
To: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
Cc: Petr Jelinek <pjmodos(at)pjmodos(dot)net>, Jaime Casanova <jcasanov(at)systemguards(dot)com(dot)ec>, Robert Haas <robertmhaas(at)gmail(dot)com>, Abhijit Menon-Sen <ams(at)toroid(dot)org>, pgsql-hackers(at)postgresql(dot)org
Subject: Re: GRANT ON ALL IN schema
Date: 2009-10-13 01:35:32
Message-ID: 37ed240d0910121835y12107b58tc2e8b127e45b7e14@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

2009/10/13 Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>:
> I started looking at this, and the first thing I noticed was that it
> adds TABLES, FUNCTIONS, and SEQUENCES as unreserved keywords.  Now
> I'm not a fan of bloating the parser that way, but I have to admit
> that "GRANT ON ALL TABLE IN SCHEMA" wouldn't read well.  What I am
> wondering is whether we should not go back and adjust the syntax
> for the default-ACLs patch to use the same keywords, ie not
>
> ALTER DEFAULT PRIVILEGES ... GRANT ... ON TABLE TO ...
>
> but
>
> ALTER DEFAULT PRIVILEGES ... GRANT ... ON TABLES TO ...
>
> Comments?

My personal feeling is that the syntax of ALTER DEFAULT PRIVILEGES
works fine as it stands. When you specify a default priv of "GRANT
SELECT ON TABLE TO dave" on a schema, it means that whenever you
create a table it implicitly does a "GRANT SELECT ON <new table> TO
dave".

I think the symmetry between the default priv and the related GRANT
outweighs the consideration of whether the command parses more like a
valid English sentence.

Cheers,
BJ

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message David Fetter 2009-10-13 01:38:28 CTE bug?
Previous Message KaiGai Kohei 2009-10-13 01:25:39 Re: [PATCH] Largeobject access controls