RE: Linkbases, Topic Maps,and RDF Knowledge Bases -- help me understand, please

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

As the man's sig says, "real programmers can program 
assembly in any language".

But we do have to consider cost.  What one can do with 
the choice of representation and how much it costs to 
use that representation must be considered.  Unless 
we are clear about the overlap and the difference 
we find ourselves doing too much, too little, 
reinventing yet another engine, or waiting for a 
specification to settle down so we can invent an engine 
we may not need. 

In all cases, understanding what works now is important. 
Simplicity is advantageous in terms of adequacy, not 
completeness.

So when is the high level adequate and when should 
one bear the costs of creating and maintaining low 
level descriptions? 

Len 
http://www.mp3.com/LenBullard

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

-----Original Message-----
From: Michel Biezunski [mailto:mb@infoloom.com]

RDF and Topic maps address basically the same problem, but the layer of
represention is different. Topic maps are better for "high level"
description (chemistry),
RDF is closer to what a computer actually does in terms of connecting
granules of information
(physics). I think what needs to be done is to articulate these two levels
in the clearest possible
way, and then take the features of topic maps and rdf and assign them to
where they belong,
while resolving the overlap. It's not an obvious task, but I think it's
worth trying.