RE: (FC-Devel) What is happening with FreeCase ?

[crawford.mathieson@nokia.com]

My initial posting yesterday morning seems to caused a flood of responses. I
have to agree with the individuals who say that the FreeCase project as
described on the web page is far too complex.

We use a CASE tool here called PROSA. It was developed from a very simple
Structured Analysis & Structured Design SASD (Like Shlaer Mellor) tool. This
now supports UML as well in the same executable. The executable is less than
1Mbyte which is tiny for a CASE tools these days. It does not have a
repository as such as all the attributes and other information for a model
are stored in the diagram files. They have a simple and consistent file
format which is human readable (not binary like other tools) and you can
actually read a diagram juts by looking at the file in a text editor. The
only problem with PROSA is it costs £2000 per seat and of course it is not
easy to get additions and fixes added to it as it is not open source.

I would suggest maybe thinking again about the requirements for a UML tool
and definitly would suggest using WxWindows as a cross platform development
tool. I have used this and it is similar in its ease of use as Qt or Gtk.
But to summarise, please keep it simple and then maybe make it more complex
later.

-----Original Message-----
From: EXT Arjan J. Molenaar [mailto:A.J.Molenaar@xirion.nl]
Sent: 09 February 2000 08:09
To: 'freecase-devel@seul.org'
Subject: RE: (FC-Devel) What is happening with FreeCase ?
Sensitivity: Confidential


> 
> >> Even having doubts about UML.  It just seems too
> >> complex. I'm more enamoured with Shlaer Mellor as
> >> time goes by.  Draw pictures, stick the information
> >> in a database, generate code in desired language
> >> (like writting a report).
> > 
> > But UML is complete...
> 
> Can you run UML models?  Action semantics were still being 
> worked out last I
> heard.
> 
> > It's model is not simple, but is can easely be used as
> > a superset of (almost) every other modeling language.
> > That's after all what it's ment for :-)
> 
> You are mixing the model and the meta-model.  UML is an 
> instance of the
> meta-model (MOF).  Other modeling languages would be other 
> instances of the
> meta-model, not subsets of UML.

What I was trying to say is that, since UML is some sort of "Merge" of most
existing modeling languages, it will probably address most features of those
languages in some sort (just gessing :-).
BTW can you give me a link to this Shlaer Mellor thing?

Arjan