Tomas Vondra <tomas.vondra@2ndquadrant.com> writes:
> On Fri, Jul 12, 2019 at 02:00:25PM +0200, Pavel Stehule wrote:
>>> Mutable function are allowed in check constraint expressions but
>>> it is not right. The attached is a proposed fix for it including
>>> regression test.
> Yes, and the question is whether this is the right thing to do (I think
> it probably is).
I'm pretty sure this change has been proposed before, and rejected before.
Has anybody excavated in the archives for prior discussions?
> OTOH, even if we prohibit mutable functions in check constraints, people
> can still create triggers doing those checks (and shoot themselves in
> the foot that way).
There are, and always will be, lots of ways to shoot yourself in the foot.
In the case at hand, I fear we might just encourage people to mark
functions as immutable when they really aren't --- which will make their
problems *worse* not better, because now other uses besides check
constraints will also be at risk of misoptimization.
regards, tom lane