RE: [w.formann@netsurf213.neuss.netsurf.de: Re: (FC-Devel) Is the project still alive??? [EXPIRED TRIAL LICENCE]

["Jacques Bergeron" <jacques.bergeron@dogico.com>]

Hi Wolfgang!

> Wolfgang wrote:
>
> Please! Ignore the Amigos! If you want to create an UML tool, then go to
> OMG
> and read their standard! There are too many differences between the
> books of
> those Amigos and the OMG definitions. Otherwise, if you want to write an
> Amigo-Tool, go on and continue :-(

So I created a certain effect citing the Amigos, good!

I will try to make my point using the UML 1.3 specs (june 1999). In section
2-14 - Model Management (where they talk about Packages, Models and
Subsystems) we have interesting phrases like:
"Because this is a logical model of the UML, distribution or sharing of
models between tools is not described"
or
"It is expected that tools will manage presentation elements, in particular
diagrams, that are attached to model elements"

In section 6.3 we have the UML-XMI-DTD which add identifiers to elements (no
reference to that before) but not a word on how to use that for model
distribution or sharing between tools.

So maybe I'm missing something or the specs (the OMG ones) are not specific
about how to structure a decentrlized application like Chris is trying to
do.

I was just asking him what was his strategy about that problem. Pointing
that different approaches could be taken, all of them satisfying the OMG
specs. Based on that basic architecture we can evaluate how we can
contribute.

> Sorry to say so, but UML is (part of) my job :-(

Well UML is not part of everyone's job and sometimes poeple are just
searching for a way to spend time on projects that they consider important
:-)

> > "Packages are primarely intended as access and and configuration control
> > mechanism to permit developers, in large work groups, to organize large
> > models and evolve them without getting in each other's way".

> Not the package itself is used to organize models, it is the package
> with
> the stereotype <<subsystem>>. Small difference in notation, big
> difference
> in usability!

Interesting, I was thinking that Subsystems (as defined by OMG specs) served
as functional grouping mechanism. "A subsystem is a grouping of model
elements that represents a behavioral unit in a physical system". Can you
point me to the sections taking about the role of Subsystems as access and
configuration control mechanism.

Finally, since UML is a big part of your job maybe you could suggest me a
way to have an interesting contribution in a free software project on
modeling / CASE tool.

Regards,

Jacques