Curt Sampson <cjs@cynic.net> writes:
> On Fri, 9 Aug 2002, Jean-Luc Lachance wrote:
> > Joe Conway wrote:
> > We could have one sequence of OID (4 bytes) per table and a prefix
> > (4 bytes) for a specific table in the system table. So we could
> > have an effective OID of 8 bytes and still keep the benefit of
> > system wide unique OID.
IMHO, it would be a better idea to eventually change to not creating
OIDs by default on user tables (for backwards compatibility, we should
probably wait a little while to make this change -- but perhaps toggle
it with a GUC option, disabled by default, in the short-term?). Since
WITHOUT OIDS is a space optimization in development sources, we're
most of the way there already...
> Or we could just create an explicit "object ID" column in those
> system tables that need it, and drop the whole object IDs thing
> entirely.
As far as I can tell, OIDs on system tables are exactly that: an
explicit "object ID" column that uniquely identifies entries in system
catalogs.
Cheers,
Neil
--
Neil Conway <neilconway@rogers.com>
PGP Key ID: DB3C29FC
