Merlin Moncure <mmoncure@gmail.com> writes:
> On Thu, Nov 18, 2010 at 12:47 AM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
>> Yes, which begs the question of why bother at all.
> Pavel's performance argument is imnsho valid.
Well, that argument is unsupported by any evidence, so far as I've seen.
More to the point, if there is indeed an interesting performance win
here, we could get the same win by internally optimizing the existing
syntax. That would provide the benefit to existing code not just
new code; and it would avoid foreclosing our future options for
extending FOR in not-so-redundant ways.
regards, tom lane