"Michel Walsh" vanderghast@VirusAreFunnierThanSpam wrote in
:

You are also missing that GROUP BY WITH AGGREGATE can use the
similar query plan than the one used by DISTINCT and GROUP BY
without aggregate (as you can check my 'claim' using MS SLQ
Server), where the aggregates are done for a small fraction (less
than half of one percent of the whole query execution). Jet is
inefficient in both cases, I mean, with or without aggregate.

Have you tested that Jet's GROUP BY for aggregates has a performance
problem?

I NEVER said that result that are identical implies the execution
plan should be the same. I said LOGICALLY EQIVALENT description
SHOULD be optimized the same way. And DISTINCT and GROUP BY
without aggregate ARE logically equivalent descriptions.

But they *aren't* logically equivalent.

I'm done here.

--
David W. Fenton http://www.dfenton.com/
usenet at dfenton dot com http://www.dfenton.com/DFA/