Alvaro Herrera <alvherre(at)2ndquadrant(dot)com> writes:
> On 2019-Feb-09, Tom Lane wrote:
>> Oh ... then why don't we go ahead and get rid of the constraint entry,
>> too?
> Because each partition has its own pg_constraint entry. (Otherwise
> there's no place to put the column numbers into -- they can differ from
> partition to partition, remember.) The only thing we do is mark it as
> child of the parent's one.
Uh-huh. And what happens after DETACH PARTITION ... are you going to run
around and recreate these triggers?
regards, tom lane