Re: [SQL] Timestamp indexes

[Tom Lane]

"Mitch Vincent" <[EMAIL PROTECTED]> writes:
> Looks like that index scan is very unattractive...

Yes, though not as bad as the cost estimator thinks (almost a 5:1 ratio
in estimated cost, but hardly any difference in real runtime).  Still
have some work to do in tweaking the estimates, obviously.

                        regards, tom lane