Re: Updating a unique constrant - Mailing list pgsql-general

From Peter of the Norse
Subject Re: Updating a unique constrant
Date
Msg-id Pine.LNX.4.44.0406071329040.19430-100000@Radio1190.Colorado.EDU
Whole thread Raw
In response to Re: Updating a unique constrant  (Bruno Wolff III <bruno@wolff.to>)
List pgsql-general
On Sun, 6 Jun 2004, Bruno Wolff III wrote:

> Uniqueness constraints can't be deferred. One common strategy is to change
> all of the values to values in a set disjoint from what they are now union
> with what they will be. (Commonly people make the negative of their
> current value.) And then change them to the correct new values.

Thanks for the advice. I already screwed up my test database when tring to
make it work without the unique index.
--
Peter of the Norse


pgsql-general by date:

Previous
From: "Henrik Steffen"
Date:
Subject: index with LIKE
Next
From: Chris Ryan
Date:
Subject: Re: [HACKERS] The pgreplication project