UseCase Relation in use case diagrams

[qwerty]

I wonder where from the UseCase Relation stems which is defined in the context of use case diagrams.

We have some kind of philosophic discussion about it and it is argued that a "use" in classical sense is connectd with Dependency. But this UseCase Relation is obviously a meta class of association (due to the fact that it is a solid line). Now I browsed Superstructures 2.3 and 1.5 but non seem to mention such a UseCase Relation.

Any light out there?

q.

[Geert Bellekens]

q.

I don't think there's something like a "use case" relation in UML. It seems like that is an invention of EA.
See p613 (actual page 630) in the UML 2.3 superstructure:

[highlight]Use cases may have associated actor[/highlight]s, which describes how an instance of the classifier realizing the use case and a user
playing one of the roles of the actor interact. Two use cases specifying the same subject cannot be associated since each
of them individually describes a complete usage of the subject. It is not possible to state anything about the internal
behavior of the actor apart from its communications with the subject.
When a [highlight]use case has an association to an actor[/highlight] with a multiplicity that is greater than one at the actor end, it means that
more than one actor instance is involved in initiating the use case. The manner in which multiple actors participate in the
use case depends on the specific situation on hand and is not defined in this specification. For instance, a particular use
case might require simultaneous (concurrent) action by two separate actors (e.g., in launching a nuclear missile) or it
might require complementary and successive actions by the actors (e.g., one actor starting something and the other one
stopping it).

Geert

[qwerty]

Thanks Geert, I also found that. But as you said, it's probably an invention by Sparx. I was mislead by this just thinking that anything inside EA is based on some UML spec. Obviously this is not the case. Well, I'm warned now.

Can anyone from Sparx comment?

q.

[KP]

I always assumed the UseCase relationship was a hangover from UML 1, but looking at the UML 1.3 spec it doesn't get a mention:

The relationships are associations between the actors and the use cases, generalizations between the actors, and generalizations, extends, and includes among the use cases.

So, sorry I don't know the answer except that it's there for historic reasons. I can't think of any reason why anyone would use it, but I'm sure we'd get complaints if we "retired" it.

[qwerty]

Yes, you would :-) I found it useful as the Use relation is always directed UC->Actor without an arrow and thus neatly indicates secondary (involved) actors. I added a Directed Assoctiation which always goes from Actor->UC and is additionally directed thus indicating the primary (triggering) actor (in contrast to the currently offered Associatipon which is of direction Unspecified and not visibly to dfferentiate to the Use relation). Of course a simple association would have done in both cases but the quick link table allows this mechanism only with different meta classes (or stereotypes which in this case wouldn't look nice).

So thanks, the history is clarified and we know why we use what and can justify it.

q.

[Oliver F.]

For a while now I came to the conclusion that the use relation can be handy to denote an interaction between use cases which is helpful in use case normalisation in later analysis phases.
Eg. by desribing a generic functionality (like eg. "manage notifications") and relating all use case depending on this functionality with a "use" relation gives an idea how this can be modularized. The special use of the dedicated relation is helpful also in searches (as well as in the compatibility matrix) as it distinguishes itself from the generic dependency relations. In all the profiles I have created since I took care that a use relation is included.

Oliver

[qwerty]

Seems that Neil was pretty right in his assumption.

q.