On 03/03/2020 00:02, Don Seiler wrote: > On Mon, Mar 2, 2020, 12:30 Vik Fearing <vik@xxxxxxxxxxxxxxxxxxx> wrote: > >> On 02/03/2020 18:09, Don Seiler wrote: >>> The REAL reason for this is that I'm wondering if I created a NOT NULL >>> check constraint with "NOT VALID" would that then NOT be considered in >> such >>> a "short-circuit" case until I ran the VALIDATE CONSTRAINT on it? >> Perhaps I >>> should have just asked this in its own thread but I started diving into >> the >>> query plan thing. >> >> You cannot do this because NOT NULL isn't a real constraint (meaning it >> does not appear in pg_constraint). There have been several attempts to >> make it a real constraint over the years but so far nothing has come of >> them > > > Using the check constraint method seemed to allow for the "not valid" step. > I'm curious what the difference is between a NOT NULL check constraint > versus setting the column to NOT NULL (assuming both are validated). You kind of answered your own question there. -- Vik Fearing
