Re: Sorting out acl fixes - Mailing list pgsql-hackers

From Christopher Kings-Lynne
Subject Re: Sorting out acl fixes
Date
Msg-id 40FF2059.3020903@familyhealth.com.au
Whole thread Raw
In response to Sorting out acl fixes  (Christopher Kings-Lynne <chriskl@familyhealth.com.au>)
List pgsql-hackers
> 1. Should we make the owner aclitem NEVER appear in the acl list?  ie. 
> when we do the first grant on an object, we don't put in a default acl 
> for the owner.  Instead we special case the aclcheck to always allow the 
> owner full privilieges?  Also, if the first grant was 'select' for the 
> 'other' user, and then we changed the owner to the 'other' user, should 
> we erase the 'other' user's aclitem?

I forgot to mention - under this schema grants/revokes to the owner 
become no-ops.

Chris



pgsql-hackers by date:

Previous
From: Christopher Kings-Lynne
Date:
Subject: Re: Why we really need timelines *now* in PITR
Next
From: Tom Lane
Date:
Subject: Re: check point segments leakage ?