Bobby B writes:
It wasn't clear from box's original post that his tables were in what you call a "master-detail" relationship. It was Norm who first assumed that.
Well, we actually know (now) that they *aren't* in a "master-detail" relationship, if by that you mean they'd have to have referential integrity constraints defined on them...

But on the other hand, they *are* in a "master-detail" relationship in the sense that that's how they're being used in BOx's query (if I understood it correctly); they are *being* "related" by the JOIN in the query.

At least the latter of my two paragraphs/sentences above is no assumption, on Norm's or anyone else's part.

And that the latter *should* (as in, "ought to") imply the former is also more of a reasoned opinion than an assumption. :-)