Hello all, The principles and ideas behind MDA are not new when regarded in isolation. We have had models for a long time. Some type of code generation is...
Hi ... and to aboid a double definition the dependency should therefore be an implicit corrolary of its cause, not the accidental assertion of a diligent...
... else ... tools ... an implicit ... modeller. ... to avoid ... declaration of the ... Do you mean that any assertion of a dependency by modellers is ...
Hi A significant proportion of dependencies are required to overcome the inadequacies of the modelling tool/language/practitioner. Hopefully these will...
Hi Aron ... Could you rephrase this statement. I think I understand what you are saying, however, its not very clear. I would not have drawn an immediate link ...
Hi Aron I am not sure whether these principles are intended to characterise MDA (i.e., if you adhere to these principles, then you are doing MDA) or are...
... conceptually ... are saying, ... link ... what you ... promote ... The principle of separation of concerns says, in my understanding, that we should...
Hi Aron and all. 1. Separation of concerns I agree that this is a very important principle, but I think I also agree with Anthony that "platform specificity"...
... Answered in reply to Ashley. AVA > - explicit dependencies: If anything is dependent on something AVA > else this should be explicitly noted and...
... Ashley has explain well what I could not. This is what I meant, thanks Ashley. ... I agree it is not MDA related specifically. When I was referring to...
Hello again, after a couple of days of inactivity of my side, ... agree with Anthony that "platform specificity" and "level of abstraction" should not be...
... this ... that any ... not ... OK, I agree that traceability to requirements and / or use cases is not definitely a part of MDA. But do you also mean...
Enter your vote today! A new poll has been created for the mda-discussion group: Generators normally produce most of the infrastructure part and only parts of...
mda-discussion@yahoog...
Oct 20, 2003 7:22 pm
22
Following up on the discussion about platform specificity vs. level of abstraction, there is a paper that is relevant to this issue. It is available from the...
Hi, please note that this poll is about **current** tools. I know tools implementing the different appproaches except for the last one if you keep in mind that...
Aron and all. ... Perhaps so, but it is still IMHO wrong. To say that Artefact A is more abstract than Artefact B means (I think): 1. Everything that Artefact...
... *business specific*. Good, I'll include this if I make a final list. Of course, new issues bring up more questions... Do you, or others, have ideas how...
<vendor disclosure>I work at Compuware on the OptimalJ product team</vendor disclosure> I went to work for Compuware because I really like their OptimalJ tool....
My view is that the reality of UML 1.4 (and I think still UML 2.0, though to a lesser degree) is that there is no way to completely define the behavioral...
Aron ... These are good questions. ... One specific use of more business specific models is to validate the behaviour of the model by execution prior to...
... I am not sure that it is correct to describe OptimalJ as a "full implementation of MDA" if you take into account the statement by Richard Soley (CEO of...
Ashley, ... models ... I totally agree with the idea that models should be testable and simulatable, and that validating them prior to building the application...
Hello Aron, My approach in these matters is a bit different, I'm not in search of underlying principles of MDA but rather at requirements and the way thy can...
Aron ... I am not sure what you mean by "which models". Do you mean "which UML notations"? If so, I don't think there is a clear answer, as the notations can...
Hello again, (changed subject) ... Well there are other aspects which can not be expressed (entirely) in the PIM, like performance and security. These have to...
Hi Aron ... Yes, that's true. And, as you point out, these have to be handled at the PSM or even hardware level. A PIM, even if *completely business specific*...
Hi Aron (Sorry -- I sent an incomplete version of this by mistake. This is the full version!) ... Yes, that's true. And, as you point out, these have to be...
Hello Gentelman ... allowed ... the current ... enough to build completely business specific PIMs. The difficult part is specifying behaviour, and there are...