On Mar 29, 2008, at 9:25 PM, Kevin Menard wrote:
> Likewise, there may be a reason a user is mapping multiple
> relationships,
> and in that case, I'd call the term "redundant" pejorative.
Ok, let's call them "overlapping" or something else. The important
thing is what that means: there are two or more overlapping
collections based on the same join condition. And this is a curse for
object graph consistency. Users should create their own filter on top
of a single most inclusive collection and stop mapping the overlapping
ones ... or risk messed up object graph right away. I'd say this
should be a warning in the modeler.
> If Cayenne does not need the relationships that it is creating,
> however, those would be
> redundant. As near as I can tell, that is the case.
Ok, maybe to reduce the number of cases we need to analyze, why don't
we stop this discussion, and work on reducing the number of runtime
relationships created (hmm... I only see a single case: a to-many part
of a 1..N), then see what harm is caused by the remaining ones.
> If we want to go down the path of allowing multiple reverse
> relationships, I
> can lead the work up. I don't want you to think I'm trying to shell
> this
> off on you. I just don't want to be making large architectural
> changes
> without someone else keeping me in check.
Yes, in order for us to keep consistent architecture, I feel like I'd
have to be involved anyways. I don't think it is good for anybody if
we have a set of diverging architectural visions in one product.
Sounding like a control freak, but I don't see any other way around.
Andrus
This archive was generated by hypermail 2.0.0 : Sat Mar 29 2008 - 16:23:54 EDT