I thought that this was clear to everybody but it seems that I was wrong. In yesterday’s panel at the MoDELS’09 conference, Gary (sorry I don’t remember his LAST NAME) commented that when he asks to a company/group about the development method they use, still half of them reply they use UML!!.
It is no surprise that then these companies get disappointed with UML and claim that modeling is not useful at all.
You may like UML or not but in either case UML is not going to fix the problems of your development process, that’s your job, don’t try to use UML as the scapegoat.
FNR Pearl Chair. Head of the Software Engineering RDI Unit at LIST. Affiliate Professor at University of Luxembourg. Â More about me.
Jordi, true, UML is not a method. But it is not a notation either. UML is really a full-blown language. It has a common notation (the graphical one everyone knows), but admits others.
http://abstratt.com/blog/2008/02/07/textual-notations-and-uml-compliance/
Cheers,
Rafael
IN Mexico IS the same: many organizations believe that UML IS a method NOT a notation … I think that the problem IS that many OF them have learned UML without focusing IN the importance OF the way TO use it, so many forget the importance OF the development method …
I think that they refer to some development method where UML is very used. I found that in companies is very common to use different (or in this case, incorrect) terminology.
But I’m almost sure that they don’t think that a language is a methodology. There is a problem of communication between academia and companies.
and I completely agree with the communication problems. We need to work harder on this.
I’d rephrase a little BIT your sentence. The UML standard defines BOTH the abstract syntax AND the concrete syntax FOR the LANGUAGE. OF course, this does NOT mean that you cannot define your own concrete syntax (e.g. a textual one). AS you mention, AS long AS your concrete syntax IS still conformant WITH the UML metamodel you’ll be able to interchange your models with other UML tools.
I hope we are in agreement here.
Notation = concrete syntax
My point is that the UML specification is at least 80% semantics (abstract syntax + rules defined on them), and 20% notation (which is even optional). So calling UML a notation is really not making it justice.
Motion accepted. I’ve changed the title OF the post TO “make justice” TO the UML 😉 (though this does NOT affect the point OF the post)
Haha, was not expecting that…
Yes, that definitely did not affect the point of your post. Sorry for sidetracking the discussion, ‘Pedantic’ is my middle name… 🙂
Semantics = abstract syntax + rules defined on them ??
Sorry, but I do not understand this equivalence :-S
As far as I know…
… the set of words/terms/concepts/elements included in a language, and the rules that define how these words are combined to build correct “sentences”, define (both together) the abstract syntax of that language.
… the concrete syntax defines the graphical or textual representation/notation asociatiated to all the words/terms/concepts/elements included in the language.
… semantics deals with the “meaning” of the words/terms/concepts/elements and sentences… i.e., how these words/sentences are interpreted in another (natural, modeling, programming, execution) language.
Oxford English Dictionary:
SEMANTICS: 1. the branch of linguistics and logic concerned with meaning. 2 the meaning of a word, phrase, sentence, or text.
I’m afraid that the only semantics explicitly embedded in UML are the well-formedness rules that complement the definition of its abstract syntax. To have more precise semantics we should map the UML to another language with well-defined semantics. I’d recommend you to read this blog post by Steve Cook on this same topic