Re: Updates not atomic with respect to indexes - Mailing list pgsql-general

On Wednesday 28 April 2004 13:37, Tom Lane wrote:
>
> I believe this is meant to be covered by this TODO item:
>
> * Allow DEFERRABLE UNIQUE constraints
>
> because the real issue is that the uniqueness check occurs immediately
> rather than being deferred till end of statement or transaction.
>
> Dunno when it will rise to the top of anyone's priority list...
>
>             regards, tom lane


Thanks for everyones responses. I guess for now I will just need to find a way
around it. Looking forward to deferrable unique contraints though!

Ben