xml-dev - RE: RDF or UML (Was RE: XML and Healthcare,RDF spec bug was: Re:Default

["Bullard, Claude L (Len)" <clbullar@ingr.com>]

Thanks guys.  Ok, another MMTT that resolves in a choice 
of the top level systemic definitions.  UML gets an edge 
for having a widely accepted graphic symbol set and is 
well-supported by existing tools.  RDF is favored by 
some web heros wanting to define their own way to do 
the kinds of things the PDES folks defined.  Tough 
sell in both cases.  Politics is local.  Appeals to 
mathematics are just a choice of systemic definitions.

This comment:

"With respect to this, I tend to take the pragmatic side: you cannot
solve the interoperability problem by generating a wealth of
incompatible DTDs (syntaxes)."

strikes me as funny.  XML does not solve any problems of interoperability, 
just data portability.  Incompatible DTDs are a symptom of a lack 
of requirements for defining the characteristics of interoperation as
predictable behavior over portable data (say, shared).   

It has been said again and again so this time, shouted:

DATA IS PORTABLE.  SYSTEMS INTEROPERATE.  That is why 
XML structures are data objects, not classes.

Failing to "get it" on that one is behind a lot of the 
monkey motions of the web.  The moon in the water is 
not the Moon.  XML is water.

Len Bullard
Intergraph Public Safety
clbullar@ingr.com
http://fly.hiwaay.net/~cbullard/lensongs.ram

Ekam sat.h, Vipraah bahudhaa vadanti.
Daamyata. Datta. Dayadhvam.h


-----Original Message-----
From: johns@syscore.com [mailto:johns@syscore.com]
Sent: Friday, July 21, 2000 8:10 AM
To: Bullard, Claude L (Len); 'Jonathan Borden'
Cc: xml-dev@lists.xml.org
Subject: RE: RDF or UML (Was RE: XML and Healthcare,RDF spec bug was:
Re: Default Namespaces - why don't they apply toattributes?)


Len asked:

"How does RDF stack up next to UML for modeling semantics?"

It seems to me that RDF is at least 2 meta levels higher than UML (the
modeling language). Here is what UML 1.3 says about meta levels (excuse the
formatting):

"Layer		Description					Example
meta-metamodel 	The infrastructure for a		MetaClass,
MetaAttribute,
			metamodeling architecture.
MetaOperation
			Defines the language for
			specifying metamodels.
metamodel 		An instance of a meta-metamodel.	Class,
Attribute, Operation,
			Defines the
Component
			language for specifying a
			model.
model 		An instance of a metamodel.		StockShare,
askPrice,
			Defines a language to
sellLimitOrder,
			describe an information
StockQuoteServer
			domain.
user objects (user data) An instance of a model.
<Acme_SW_Share_98789>,
			Defines a specific			654.56,
sell_limit_order,
			information domain.
<Stock_Quote_Svr_32123>"

So, I think that RDF is at the meta-metamodel layer. UML is at the model
layer (with a subset based on UML Class diagrams at the metamodel layer).

In other words, RDF is a language that is so generic you can model any
meta-model in it. By the time you get to the model layer, languages are
specific to the thing being modeled. So UML is specific to object models (it
knows about visibility for example, a concept that doesn't exist in data
models). IDEF1X, on the other hand, is specific to data models (it knows
about identifying relationships, a concept that doesn't exist in object
models). Other modeling languages exist for modeling integrated circuits
(VHDL) and so on.

Yours,
John F Schlesinger
SysCore Solutions
212 619 5200 x 219
917 886 5895 Mobile

-----Original Message-----
From: Bullard, Claude L (Len) [mailto:clbullar@ingr.com]
Sent: Friday, July 21, 2000 9:25 AM
To: Jonathan Borden
Cc: xml-dev@lists.xml.org
Subject: RDF or UML (Was RE: XML and Healthcare,RDF spec bug was: Re:
Def ault Namespaces - why don't they apply toattributes?)


How does RDF stack up next to UML for modeling semantics?

Len Bullard
Intergraph Public Safety
clbullar@ingr.com
http://fly.hiwaay.net/~cbullard/lensongs.ram

Ekam sat.h, Vipraah bahudhaa vadanti.
Daamyata. Datta. Dayadhvam.h

-----Original Message-----
From: Jonathan Borden [mailto:jborden@mediaone.net]

1) RDF was designed for modelling the rich types of semantic relationships