Responding to McNeile... ... One rule of thumb: if the behavior is so complicated that Harel starts to look attractive, it is too big. B-)) [Mellor actually...
Responding to Nedunuri... ... I agree that in a well formed OO application it is somewhat unusual for events to have a data packet (within a subsystem;...
Responding to Nedunuri... ... Hopefully the application abstracts invariants and defines things like seating capacity in external configuration data. B-) In...
... When dealing with systems that are distributed and/or modular the designer has no choice but to deal with events that arrive in the "wrong" state. ...
Responding to H.S. ... ... If *all* the computer program is doing is tracking the position of the pawn (and this is what I am assuming), it does not need to...
Moderator's Note: Before approving the following message, I wrote to the poster, telling him that normally we do not allow commercial advertising on the group....
... capacity of ... Not on the class diagram. Otherwise you are looking at a class for each instance of bay, gate, waiting area, and airplane. Capacity for...
I have a couple of questions concerning instances. At the top of page 99, the Executable UML Book says the following: "We are careful to use "instance" to mean...
For more background on Ashley's question, check out messages 155 and earlier. Under the threads "Reclassification Problem" and "Compound Generalization" we...
It's been a few days since this question and no one has responded, not even H.S. ;-). So I'll give it a shot. ... this question ... book.copyright ... "No...
Note from Marc: Peter provides an excellent clarification of my original points. I'm in complete agreement with everything he writes below. -- Marc Balcer ... ...
... SYNTAX. A ... simply ... compiler can ... then they ... [SNIP] ... while ... syntax, ... "problem", ... I would have to disagree on this. I proposed in an...
... While I agree with Peter's point that defining common semantics is a valuable and necesary first step, I more firmly agree with Lee's. I cannot imagine...
Responding to Bestler... ... Where did I say that one should assume that such things don't happen? Of course the developer must deal with them! The point here...
Responding to McNeile... ... The computer should not know why the move was selected. In fact, it should know nothing about the context of the decision. All...
Marc, Peter Thanks for your very detailed answers the question. Might I add that it really was quite an innocent question (one that popped into my head as I...
... I'm afraid I'm not following you. OK perhaps the example isn't great but lets run with it. Boeing747 is a subtype of PassengerJetAircraft. Its modeled in...
I am interested to note that the U2 proposal for UML Version 2 makes a high-level distinction between "Behavioural" and "Protocol" state machines. This is...
Responding to H.S.... ... Then, perhaps, an air-traffic control system should not track the movement of an aircraft if it was caused by the plane's cat jumping...
... Specifically, I think you are asking why we used OCL to represent constraints in Chapter 8, but did not use OQL for queries in Chapter 7. OCL was part of...
Responding to McNeile... ... My (secondhand) knowledge of the U2-Partners submission is that it is an attempt by a small group of vendors to insert their...
... examples I ... to the ... view to ... domain ... Boeng747 ... trigger some ... you get ... What you describe is a model that can only be used for...
Responding to McNeile... Since the authors haven't responded I'll try reading their minds... ... I think they are only trying to make the point that the OO...
... A complete split as described here strikes me as typical of the problems in full UML. There simply are no real-world projects that can afford to fully...
Responding to Bestler... ... eUML employs less than half of the UML artifacts. In addition eUML does not require one to model the implementation (more...
... It appears to me that you are looking at Executable UML as a means of modeling the implementation. This tends to make the Moore vs. Mealy debate more...
... The core concept of both Recursive Design and eUML that I am in most violent agreement with is that systems can be expressed as a collection of objects...
... Again, I am at somewhat of a loss as to what you are trying to express. What I am inferring from the above is that the solution model contains platform...
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.
