Re: CREATE TABLE/AS does not allow WITH OIDS?

From: Thomas Lockhart <lockhart(at)fourpalms(dot)org>
To: Neil Conway <neilc(at)samurai(dot)com>
Cc: pgsql-hackers(at)postgresql(dot)org
Subject: Re: CREATE TABLE/AS does not allow WITH OIDS?
Date: 2002-11-07 19:33:30
Message-ID: 3DCAC00A.2060709@fourpalms.org
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

> How would we make use of 'hasoids' in the case of multiple source
> tables, or a source table defined by an SRF?

Choose a convention. At the moment, the first source table provides more
characteristics to the target than do the other sources, and that could
be true for OIDs also. I would be inclined to have the logic be that if
*any* source table has OIDs, then the target gets them. Unless it was
explicitly specified using the WITH/WITHOUT syntax of course.

> Unless we can define some consistent semantics for when we use the
> hasoids of the source table(s), I'd say we should just add the
> WITH/WITHOUT OIDS syntax.

Right. There is no problem coming up with consistant semantics, though
there may be a problem getting a consensus on what those should be.

In the meantime we have broken compatibility with the previous release
(since targets are created without OIDs) and should think about fixing that.

- Thomas

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message joel w. reed 2002-11-07 19:42:29 Re: PL/Perl and Perl 5.8
Previous Message Bruce Momjian 2002-11-07 19:15:29 Re: RC1 on Friday?