Re: pgsql/src/backend/nodes (copyfuncs.c outfuncs.c print.c) - Mailing list pgsql-committers

Hiroshi Inoue <Inoue@tpf.co.jp> writes:
> It seems to me that LIMIT in declare cursor statement is only
> for optimizer's hint.

I think that would just confuse people.  If we want to have a hint
that says "optimize for fast start", it ought to be done in another
way than saying that SELECT ... LIMIT means different things in
different contexts.

Possibly the optimizer should always assume that cursors ought to
be optimized for fast start, LIMIT or no LIMIT --- does that seem
like a good idea to you?

            regards, tom lane