RE: Picking the Tools

[Uche Ogbuji <uche.ogbuji@fourthought.com>]

> Ok.  Both you and Uche make the same point: UML
> is an object modeling methodology.  I agree that
> XML documents are not object-oriented programs.  I know we've
> been round this thread before, but I wanted to
> check the expert opinions.  Berners-Lee comments
> in the SciAm article that a universal system
> is needed.
>
> So what do we have for concept modeling above the
> level of RDF or Topic Maps?   Or is it the case
> one has to commit to one of these?

Unfortunately, I think we're stuck in a crease here.  UML is not
expressive enough to model whatever we'd want to in XML/RDF/TM, etc., but
it does have mature tools and strong adoption.

XML Schemas, RDF and TM are, at the low-level, what we want, but I
quite understand that not too many people would want to do all their
modeling directly in these tools the way things now stand.  (People such
as Jonathan and I are exceptions, and probably weird).

So understandably you ask whether one should just use UML and therefrom
generate the appropriate low-level formats.  I point out that UML is not
expressive enough, that mechanical conversion from UML in any case would
tend to poor results, and that hands-on conversion would be as much effort
as just doing the modeling in the low-level formats to begin with.

I don't think there's any good answer to this quandry right now.  The
tools still need evolving.

Luckily, I think that despite the excessive hype, the Web services
movement might be exactly what is needed to push the evolution of tools
along.


-- 
Uche Ogbuji                               Principal Consultant
uche.ogbuji@fourthought.com               +1 303 583 9900 x 101
Fourthought, Inc.                         http://Fourthought.com
4735 East Walnut St, Ste. C, Boulder, CO 80301-2537, USA
Software-engineering, knowledge-management, XML, CORBA, Linux, Python