Andres Freund <andres@2ndquadrant.com> writes:
> On 2013-11-10 18:16:16 -0500, Tom Lane wrote:
>> Then I take it you also think we should undo the changes that made
>> "DROP TABLE IF EXISTS foo.bar" not fail if schema foo doesn't exist?
>> Because after all, the schema is not the object being dropped.
> No, not the same thing imo, although I find that change debatable.
> Anyway, if we're going to change DROP TRIGGER at the very least ALTER
> TABLE ... DROP CONSTRAINT also needs to be changed, otherwise we'll gain
> nothing.
That would be a plausible next step, but I don't have a problem with
this patch not solving every case like this at once.
regards, tom lane