Niklas Uhrberg wrote:
> Hi, I discussed this briefly with Rickard last week before it entered
> the mailing list.
>
> It started with the possibility of implementing the association class in
> UML which has no counterpart in "plain java".
>
> So, it's just a matter of putting data on an association between
> entities, an example is startDate on things like employment (just an
> example, maybe not a good one).
>
> I'm a but concerned about the term "role" in the discussion because
> roles are used for something else in modelling.
> Drawing on the example with the spouse relation, a person has an
> associated person which is his or her spouse , the other person entity
> plays the role of being the spouse.
So the "role" is really the association name in Qi4j, i.e.
interface Person
{
Association<Person> spouse(); // "spouse" is the role
}
> Possibly one would like to put start- and end dates on this association
> and this would motivate an association class between them as opposed to
> just a regular association.
>
> I write this just to see to that we mean the same things with "role" and
> "association class" and to motivate that "AssociationRole" is IMHO a bit
> misleading.
> Note that the data we put on an association *can* be another entity in
> the strict meaning that it's not just a way of implementing it in Qi4J.
>
> Example: A person har the spouse association to another person. This
> association is qualified be the "introduced by" association which is yet
> another Person entity.
Exactly. In RDF this is called "reification", but I'm not sure I want to
use that term here. Any other terms which are better than "role"? The
concept is valid (and extremely useful I think), but the term needs to
be refined, just as we changed EntitySession to UnitOfWork, and I'm
about to replace LoadingPolicy with Usecase.
/Rickard
_______________________________________________
qi4j-dev mailing list
[email protected]
http://lists.ops4j.org/mailman/listinfo/qi4j-dev
