Re: Re: pg_dump scripts are no longer ordinary-user friendly

On Mon, 5 Mar 2001, Tom Lane wrote:

> Philip Warner <pjw(at)rhyme(dot)com(dot)au> writes:
> > At 22:26 5/03/01 -0500, Tom Lane wrote:
> >> Now that you mention it, is it a feature at all? Or a bug? ISTM poor
> >> form for a data-only restore to assume it may turn off all pre-existing
> >> triggers.
>
> > Do you recall any of the history - why was it added in the first place?
>
> No, I don't recall. It might be worth digging in the archives.

Foreign key constraints with data following the full constraint definition
if the data was in the wrong order.

Unfortunately it does allow invalid data to be loaded, but for circular
cases I'm not sure how you can do this safely. I guess for fk, if all the
data loading was in a single transaction and you did something to override
the normal deferrable-ness of the constraint and forced the constraints to
be deferred, it would check at the end of the full load. This still
breaks for multiple dump files per table and for other random user
triggers that are unsafe on restore though.