I've replaced this kind of unique indexes with just normal (physmessage_id) if there is no other index can index physmessage_id. This is to make sure physmessage_id can use index. But unique index on (physmessage_id, id) is unnecessary here.
Kevin On 11/6/05, Aaron Stone <[EMAIL PROTECTED]> wrote: > > On Sun, 2005-11-06 at 23:01 +0100, [EMAIL PROTECTED] wrote: [kc wrote] > > btw, found a lot of unique index based on (physmessage_id, id). Since id > > is primary, it already ensured uniqueness. > > Like this: UNIQUE(id, physmessage_id)? That would make sense in the case > that either id or physmessage_id is not UNIQUE. physmessage_id is a > foreign key into the physmessage table, and id is the unique id of the > current table. Right? Yeah, sounds useless. > > Aaron > > _______________________________________________ > Dbmail-dev mailing list > [email protected] > http://twister.fastxs.net/mailman/listinfo/dbmail-dev >
