H.S. ... If you remember, it was the Open event that we were discussing in this thread (in discussing what domain object has the protocol: Open, Deposit,...
H.S. Reviewing our extensive recent exchanges, I think it is clear that we have a deep conceptual disconnect on the concept of protocols. This is evidenced by...
Responding to McNeile... ... Note that I said, "... very similar to...". ... Why not? It depends upon the level of abstraction. Typically use cases are...
Responding to McNeile... ... Where do the Born and Die events come from? Something has to happen in the domain to trigger them. That 'something' is not...
Responding to McNeile... ... I has to be (a) in order to map it into a state machine action in the protocol state machine. However, what is atomic for a domain...
Responding to McNeile... ... Then explain to me why your Account protocol {Open, Deposit, Withdraw, Close} does not include a test that the Withdraw is valid...
Responding to McNeile... ... If you really mean that, then I think you have a serious problem with your naming conventions. B-) I don't see Open as an event...
Responding to McNeile... ... For my part, what I find frustrating here is that I have repeatedly described in detail what my interpretation of what you do _in...
H.S. ... Let say that the Open event sets up the Account overdraft limit (an attribute of the Account). How is this attribute given its value in your solution?...
H.S. ... I don't understand the role of the "activity" concept here. I think it is redundant. ... No. I would state this as: (D) The protocol state machine...
H.S. Perhaps we should free up this list for others to use? I don't think we should really monopolise it any longer. Should we continue our debate off the...
Responding to McNeile... ... I agree. We are repeating ourselves and progress is glacial. ... OK. ************* There is nothing wrong with me that could not...
Maybe while the gun smoke is drifting away and the ring of battle sounds slowly leafs the ear, we can start looking at the remains left on the field and try to...
Hi ... The terminology and pictures in the MDA guide encourage the view that there is one PIM and one PSM. For real systems this unrealistic. There is a steady...
Its interesting that you say that the model is adapted. Many approaches instead take the view of creating new models at each "level". Also, I follow DSL and...
... At some levels you create, e.g. when transforming a state machine to reifying classes. But once you start using fine grained transformations for re-use,...
... Would you equate CIM to a declarative model with pre and post conditions while a PIM adds actions to achieve the post conditions? Are you using CIM inside...
Responding to Ed, ... I have a slightly different view on this matter I'm afraid... Just some of my favorite selections form the 'Basic Concepts' chapter of...
Hi Neil ... No equate, what you describe could well be a CIM, but I don't preclude other forms of CIM. ... A declarative model may be very difficult to execute...
Hi Jelle ... I don't see the difference. ... There may be many systems under construction. Some with different Platform Models, others with different Mark...
Jelle ... let ... At least in my view, H.S. and I were debating (is that the right word?) different approaches to building a PIM. IOW, the difference between...
Ed ... I think I would shift the abstraction level. In my view: PIM would specify "sort" PSM would specify "quick sort" This is because (at least in my view)...
... To try and answer my own question: This is what the MDA Guide (v1.0.1) says about the CIM: "A computation independent model is a view of a system from the ...
According to the 'MDA Guide 1.0.1', which I've just found: CIM describes 'the situation in which the system will be used'. PIM 'describes the system, but does...
Responding to Jdgerbrandy... ... It depends upon what one means by 'design'. B-) The basic breakdown I find useful is: CIM: a specification model. It is a...
Responding to Willink... ... I am uncomfortable with this analogy since sorting itself is an implementation solution (e.g., one can construct an index...
Responding to Earnshaw... ... FWIW, I see declarative vs. whatever as a matter of convenience rather than an inherent characteristic. As long as the semantic...
Responding to Jdgerbrandy... ... Actually, the difference will lie in the PIM itself. I argue that we would create very different PIMs from the same set of...
Having problems with message search?
Fill out this form to ensure your group is one of the first to be migrated to the new message search system.
