On Wednesday 26 Mar 2003 9:59 am, Chris Hutchinson wrote:
> I'm curious why postgres should be blocking like this, as the foreign key
> constraint is presumably read-based.
Actually, I believe it's a write lock (which is the cause of the problem).
What it actually needs to be is a "prevent writes" lock but there is no such
thing at the moment.
--
Richard Huxton