> Stonebraker was wrong - and must have been bacause today we
> want to get SQL92 compliant - and that spec didn't existed
> when he designed our rule sytem. The rule system is
> something we got from the good old v4.2 Postgres. That
> wasn't an SQL database, the querylanguage was POSTQUEL. So it
> isn't surprising that the original rule system spec's don't
> meet today's SQL needs.
>
> For thing's like aggregates, distinct/grouping and the like,
> we need to take a step backward and really do some kind of
> view materialization (create a real execution path for the
> view's definition). But don't force that to be done whenever
> a view is used - that doesn't make things better.
Thanks. Now I understand why aggregates cause problems with rules.
-- Bruce Momjian | http://www.op.net/~candle maillist@candle.pha.pa.us | (610)
853-3000+ If your life is a hard drive, | 830 Blythe Avenue + Christ can be your backup. | Drexel Hill,
Pennsylvania19026