Re: [RFC] Removing "magic" oids

On Sun, Oct 28, 2018 at 12:26 AM Andres Freund <andres(at)anarazel(dot)de> wrote:
>
> Hi,
>
> On 2018-10-28 00:21:23 +0530, Amit Kapila wrote:
> > On Sun, Sep 30, 2018 at 9:18 AM Andres Freund <andres(at)anarazel(dot)de> wrote:
> > > I think we should drop WITH OIDs support. pg_dump should convert WITH
> > > OIDs tables into tables that have an explicit oid column (with an
> > > appropriate default function), pg_upgrade should refuse to upgrade them.
> > >
> >
> > Is there any technical reason why you think pg_upgrade should refuse
> > to upgrade them? I think there is an argument to break backward
> > compatibility here and many people on the thread seem to be okay with
> > that, but refusing to upgrade sounds more restrictive.
>
> They'd not be on-disk compatible, because the column isn't stored as a
> normal column but in the t_hoff space.
>

Yeah, and which means we need to re-write all such tables which is not
an attractive option and probably quite some work. So, users have to
dump and restore their databases which can be time-consuming for large
databases, but I think we don't have any better option to offer.

--
With Regards,
Amit Kapila.
EnterpriseDB: http://www.enterprisedb.com