xml-dev - Semantic Web: The Bluster Continues

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

For those interested, take note of Ed Dumbhill's semantic 
web primer on XML.COM:

"The Semantic Web has already been the subject of much bluster 
among the XML developer community and will doubtless continue to 
be so. Arguments rage over the usefulness of the technology, the 
difficulty of using RDF, and so on. However, the Semantic Web vision
of a machine-readable web has possibilities for application in most 
web technology -- while some complain about its lack of definition, 
its broad scope properly reflects the quietly radical effect it will 
have on the Web."

A radical effect on the Saturn Corporation upon the fifth 
visit to fix a problem which they consistently failed to 
diagnose then said, "We just hook it up to the computer 
and it tells us what to replace": the SL1 and 
Saturn will be replaced.  New contract negotiation, new winner 
based on poor service performance by previous vendor.  The 
service model scales outward and is effective for 
determining action if an internal service fails.

The web is as effective as the quality of services 
it provides.  Negotiated contracts where names have temporal 
and contract scopes are the basis for reliable services.  
Ed says industry needs a vehicle object-relationship model. 
They've had those for years and so does the military (MIL-STD-1808 
for example).  What web technology contributes 
is the ability to scale services using such down into local 
shops that could not afford them before.   The problem will 
be the reliability of the services based on sharable semantics.

The reliability of components sensitive to both a draconian 
parse and a URI may be at issue.  We see failures of working 
components such as XSLT when the URI changes.  Pushing 
XSLT scripts to a very large user base given a single 
change to a single string seems contrary to cost.  I have 
doubts about sharable definitions that rely heavily on URIs. 
We must be careful not to mistake the results of toy systems 
developed experimentally for industrial quality services.
 
My guess is that web services concepts will continue to 
dominate and provide superior focus for creating 
requirements than the vision of a semantic web based 
on sharable semantic networks.  Savvy vendors will 
concentrate development resources on applying WSDL, UDDI, etc.

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

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