On Mon, 2 Dec 2002, Joel Burton wrote:
> Looks right to me: index scan for the less-common option, seqscan for
> the most common. Why don't you think this, as a btree, will work for
> you?
No, I'm sure a btree will work. However, won't the index be
inefficient (i.e., very flat) if there are many entries with the same
value? Or is this not a problem?
--
David.