kevin(dot)perais(at)trivia-marketing(dot)com writes:
> I've been noticing several times that clauses IN and NOT IN are often buggy.
None of your examples demonstrate any such thing. What's much more likely
is that you've forgotten about NOT IN's weird (but spec-mandated) behavior
with NULLs, and/or misspelled a field name so that the output of the
sub-SELECT is actually an outer reference.
> I'll show you 2 concrete cases to illustrate that.
My idea of a "concrete case" would be something that someone else could
reproduce from the given information, which would certainly require test
data.
regards, tom lane