On Thu, 16 Apr 2009, Kevin Grittner wrote:
> Matthew Wakeling <matthew@flymine.org> wrote:
>> I have been doing some queries that are best answered with GiST
>> indexes
>
> For what definition of "best answered"?
>
> Since an index is only a performance tuning feature (unless declared
> UNIQUE), and should never alter the results (beyond possibly affecting
> row order if that is unspecified), how is an index which performs
> worse than an alternative the best answer?
Don't be misled by my example using integers. I'm doing queries on the
bioseg data type, and the only index type for that is GiST. There isn't a
better alternative.
Matthew
--
"Finger to spiritual emptiness underlying everything."
-- How a foreign C manual referred to a "pointer to void."