Re: Table constraint ordering disrupted by pg_dump

Philip Warner <pjw@rhyme.com.au> writes:
> While it't not a bug, it would be nice if pg_dump reproduced definitions as
> faithfully as possible. To that end, would it be worth selecting the
> constraints in OID order (using oid from pg_relcheck)?

If it's just another clause in a query, you might as well.  I wouldn't
take any risks for it though...

			regards, tom lane