[Modeling] UML extensions

Renato Levy rlevy@i-a-i.com
Mon, 24 Mar 2003 10:49:48 -0500

>I do no think that symbols like ?, !, ||, etc. are more understandable than
>alt, brk, par, etc. They are maybe better readable, ... but reader/writer
>needs to know exact semantics anyhow. I think that original intention of UML
>2.0 developers, when they proposed using of textual shortcuts, was to give a
>hint about the semantics to English speaking users.
>But why to use another notation for standardized modeling elements?
>Alternative notations with exactly the same semantics can confuse users. Do
>not you think that it would make (small) problems with compatibility,
>usability and understandability of the language. Imagine joining of
>models/diagrams built by several groups of analysts, and each group used
>different notation.

I agree with radovan, multiple notations for the same semantics are 
counter-productive. This is actually one of the errors in UML that we can 
see over and over again. It also hurts the developement of supporting tools.

> > >page 12, second para: why are Seq, Strict, Critical region etc. not
> >
> > After spending a few hours with the author of the chapter, I understand
> > they are there.  IMO, some aspects of them are not well defined.  But, we
> > can choose a) to fix them up, or b) ignore them.  Electrical engineers
> > to want them.
>Dear all,
>I suggest to support all kinds of CombinedFragment. At least AUML
>specification can say nothing about them - i.e. they will be supported
>implicitly. Why to make restrictions of UML in cases if it is not really
>necessary? Maybe also some agent system can use e.g. critical regions in the
>design model, or weak sequencing that can better describe asynchronous
>At the other hand, I think that strict is defined very unclearly
>(intuitively???). The definition uses a metaclass CombinedInteraction, that
>is not mentioned at any metamodel diagram nor it is described textually.
>Therefore OMG (or we?) should provide better description.

Again I agree, if it ain't broke, DON'T fix it.

Renato Levy
Principal Scientist
Intelligent Automation, Inc.
7519 Standish Place, ste 200
Rockville, MD 20855
phone: (301)294-5241
fax: (301)294-5201
WWW: http://www.i-a-i.com