Re: GRANT ON ALL IN schema - Mailing list pgsql-hackers

From Tom Lane
Subject Re: GRANT ON ALL IN schema
Date
Msg-id 21246.1249491592@sss.pgh.pa.us
Whole thread Raw
In response to Re: GRANT ON ALL IN schema  (Robert Haas <robertmhaas@gmail.com>)
Responses Re: GRANT ON ALL IN schema  (decibel <decibel@decibel.org>)
Re: GRANT ON ALL IN schema  (Peter Eisentraut <peter_e@gmx.net>)
List pgsql-hackers
Robert Haas <robertmhaas@gmail.com> writes:
> ... bulk-grant could be based on object type,
> object name (with wildcard or regexp pattern), schema membership, or
> maybe other things, and I think that would be quite useful if we can
> figure out how to make it clean and elegant.

Yeah.  In the end you can always write a plpgsql function that filters
on anything at all.  The trick is to pick some useful subset of
functionality that can be exposed in a less messy way.

Or maybe we are going at this the wrong way?  Would it be better to try
harder to support the write-a-plpgsql-function approach?  I don't think
the documentation even mentions that approach, let alone provides any
concrete examples.  It might be interesting to document it and see if
there are any simple things we could do to file off rough edges in doing
grants that way, rather than implementing what must ultimately be a
limited solution directly in GRANT.
        regards, tom lane


pgsql-hackers by date:

Previous
From: Robert Haas
Date:
Subject: Re: GRANT ON ALL IN schema
Next
From: Sam Mason
Date:
Subject: Re: the case for machine-readable error fields