On Thu, Sep 19, 2002 at 06:36:56PM +0900, Jean-Christian Imbeault wrote:
>
> Maybe EXISTS is not always faster than IN ?
I haven't been reading this thread very carefully, but that one is
_certainly_ true. In testing, we have had a wide variety of results
with IN versus EXISTS. It seems to have to do with the relative size
of the two sides. I haven't come up with a perfect rule for deciding
yet.
NOT IN/NOT EXISTS is more often a clear-cut case: NOT EXISTS is
_almost_ (but only almost) always faster.
A
--
----
Andrew Sullivan 204-4141 Yonge Street
Liberty RMS Toronto, Ontario Canada
<andrew@libertyrms.info> M2P 2A8
+1 416 646 3304 x110
