On Mon, Jan 12, 2009 at 8:08 PM, Rickard Öberg <[email protected]> wrote:
> From a both practical and principal point of view, if Entity A has a > Value B with a reference to Entity C, then if A is aggregated then B > *may or may not* be a valid value, if you take aggregate logic into > account (i.e. are A and C part of the same Aggregate or not, or is C an > aggregate root). To have a value be ok or not depening on who is > referencing it is wrong. For this simple reason I would say that Values > cannot have references to Entities. Ahhh... I now remember the very clever discussion we had a few weeks ago about Aggregates and how to have Aggregate Support built into Qi4j Runtime, which invalidate the whole argument I was giving before. Thanks for the reminder. Now, back to the simple Address case; In a 'simple system', the Address is a Value and it has values for both City and Country, which could be returned from some 'repository'-type service. And if such service is backed by entities, is irrelevant to the Address, as long as the Entity instances are not returned directly. So, does it make sense to create a library to support such pattern, incl the Entity-to-Value conversion?? In reality, that would create an Enumeration system, but flexible enough to allow for dynamic update of what are relevant values available. Cheers Niclas _______________________________________________ qi4j-dev mailing list [email protected] http://lists.ops4j.org/mailman/listinfo/qi4j-dev
