Re: *very* inefficient choice made by the planner (regarding - Mailing list pgsql-performance

Jean-Luc Lachance <jllachan@sympatico.ca> writes:
> If the two statments are functionally equivalent, why can't PG rewrite
> the "NOT IN" version into the more efficient "NOT EXISTS"?

They're not equivalent.  In particular, the behavior in the presence of
NULLs is quite different.

            regards, tom lane