Jim Nasby <jim@nasby.net> writes:
> [ add some error ranges to cost estimates ]
> I believe that would fix this specific case because even though to plans might come out with a nearly identical cost
itis unlikely that they would also have a nearly identical error range.
Actually, I think that *wouldn't* fix this specific case --- did you
look at the details? The two formulations of the plan are really pretty
nearly equivalent; you can do the two nestloops in either order and it's
not clear it'll make much difference. I'm suspicious that the addition
of parameterized planning might open up more scope for this type of
thing, even though in principle it was always possible.
regards, tom lane