Hello all, I'm a newbie to xtUML having been introduced to the topic through Shayne Flint's excellent course at the ANU. I'm keen to learn more. While browsing...
Hi, I am currently leading the work in designing a model compiler for Ada. This has raised a question how instance handles and instance handle set variables...
... That all depends on how you write your model compiler. ;-) The correct way to handle this is for A_Instances to refer to 4 instances after the deletion....
... <lwriemen@i...> ... of ... borrowed ... after ... the ... 4? ... Actually it depends upon how you define the semantics of the operations on the instance...
Responding to Mangogna... ... I agree with Reimenschneider that it depends upon how the model compiler interprets it. That, unfortunately, makes it an...
... reasons. ... sure ... when ... By developer, are we referring to the to the developer of the Architecture domain or the developer of the non-Architecture...
As I understand the discussion so far, there are two views: 1.-A_Instances is always updated with each deletion and reflects the currently existing instances....
... If the meta model has: * is contained in contains 1..* Instance Set---------------------------------------Instance then the architect has no choice...
Responding to Riemenschneider... ... Yes; the developer of the PIM. ... You are correct that adding a generic GC is not necessary because the translation...
Responding to Wedin... ... As the proponent of this view I would clarify that the developer needs to ensure that the relationship instantiation is up to date...
OK, some of you are going to cheer. And some of you are not. After some consideration and many requests, I'm announcing a new group, ExecutableUMLTools...
Dear Colleague, FYI--I have attached a call for submissions for 2004 (below). Best Regards Roberto Zicari CALL FOR SUBMISSIONS: LogOn Briefings 2004 in Europe....
Roberto Zicari
zicari@...
Jan 9, 2004 7:49 pm
637
... Actually, if the set is composed of references to the actual instances, then this case is much easier for the architecture than the deletion case. In the...
... On the flip side, there exists the possibility of the analyst deleting the instance in one place and accessing it through the set in another. The result of...
Responding to Riemenschneider... ... My problem isn't with the mechanism nor am I concerned with wormholes. But looking at what I actually wrote I wasn't very...
Blocking is *one possible mechanism.* There are others....
Stephen J. Mellor
steve@...
Jan 10, 2004 11:52 pm
642
H.S. wrote ... I agree. This is an issue of ACID transactional behaviour and not one that should be of any concern to the modeller. The normal ACID transaction...
Responding to Mellor... ... OK, but the issue on the table is: are there /any/ practical mechanisms the engine could deploy to guarantee data integrity that...
... More fundamentally, the execution environment should ensure that the action routines of an object can be coded without regard to the mechanisms for making...
... one that should be of any concern to the modeller. ... behaves as though it is the only one executing -- and thus guarantees that nobody else can fool...
Responding to Cait... ... Right. Providing the developer with basic guarantees of data and referential integrity is the part of the translation engine's job....
Hello, everybody, I would really appreciate your help with the following issue. I want to model a set of interacting classes that represent the server-side of...
Thank you for your answers. These are my further comments. ... here. Middleware? That's right. By "server-side" I mean a set of middleware components with a...
... It is for this reason that I find I can't use the term PIM in conversing with most software developers. If I use PIM, they assume a model that isn't...
Responding to Kovse... ... If I understand this correctly, after getImages executes we then have have a third relationship instantiated, R3, whose members are...
Is there a document mapping Executable UML terminology (as used in the book) to UML 2.0 terminology (per adopted draft std Aug 2003)? If not, I think it would...