I�d like to ask you a question about modelling composit component in UML 2.0: If we have component C which continent some sub-components, for instance C1,C2, C3. As I know, in order to model such a component, we write C with (c1: C1, c2:C2, c3:C3) that�s means: the component C with instances of C1, C2 and C3 and not C1, C2, C3 them-self. My question, if (for instance) C1 have, in its turn, sub-components (C11, C12), in this case, how we can model C, C1 and their sub-components C11, C12 in UML 2.0
You just wind up with a tree. C has lines to C1, C2, C3. C1 has lines to C11 C12, C13, etc. A complex tree can quickly get too large to take in on a single diagram. That's where it's nice to distinguish between the model which contains everything and a diagram which shows some small part of the model for some specific purpose. For example, a diagram with C, C1, C2, C3 is perfectly valid for a particular level of abstraction. Another diagram with C1, C11, C12, C13 is good at another level.
Bookmark Scott Ambler's Agile Modeling site. I refer to it fairly often.
Scott pops in here now and then, too. He's pretty big on that "purpose" I mentioned above. I like to fill in "_______ reads this to learn ________ so they can ________" That will help you know what level of abstraction to put in, or whether to skip the diagram entirely.
A good question is never answered. It is not a bolt to be tightened into place but a seed to be planted and to bear more seed toward the hope of greening the landscape of the idea. John Ciardi