Re: Notices for redundant operations

Peter Eisentraut <peter_e(at)gmx(dot)net> writes:
> I found a few notices and warnings that inform you that the command you
> are executing has no effect because the object is already in the state you
> want it. I think these are useless, and there is also some inconsistency.
> Does someone want to defend keeping them?

Do you think the commands should just silently do nothing? That seems
strange to me. I like giving a NOTICE (not WARNING though, those guys
should be changed).

I guess the best argument for putting out a notice is that there's some
chance the user is making a mistake, eg, telling you SET NOT NULL when
what he really wanted to do was DROP NOT NULL. If that argument doesn't
impress people then I suppose being silent is the thing to do.

regards, tom lane