Re: ALTER TABLE .. SET WITH / WITHOUT OIDS

> I can see some value in the SET WITHOUT OIDS half of the patch, as this
> gives a way of reclaiming space in a table that was unthinkingly created
> with OIDs. But I'm not sure we should expend the code space to provide
> just a one-way transformation

This is the part I'm after -- for those older tables created before OIDs
could be removed. Editing dump files won't work. I can just muck with
the system tables, but others may not be able to (for whatever reason).

> Or we could modify heap_update to assign an OID if it reads a zero OID
> from the old tuple in a relation that is marked as having OIDs. Not
> sure whether I like that behavior either, but at least it would provide
> a mechanism for causing the old rows to acquire OIDs.

Thought about that, but didn't like it as it makes oid more special than
it already is. Adding a default to a column works basically the same
way.

Another, and my original option, was to touch all tuples to assign new
OIDS at addition time -- but that requires a lot of resources.

--
Rod Taylor <rbt(at)rbt(dot)ca>

PGP Key: http://www.rbt.ca/rbtpub.asc