Peter Eisentraut <peter_e@gmx.net> writes:
> The patch might make sense anyway, but I think it might not be such an
> overwhelming winner in practice.
As always with patches that are meant to improve performance,
some experimental evidence would be a good thing.
regards, tom lane