Re: XML Enriched N-Triples (XENT)

(going off topic - I couldn't think of a subject line other than "in defence 
of Mark Butler" - I guess he doesn't need any really)

Sean:
> yet another person who so
> openly slams RDF/XML [2] without, as far as I know, following the old
> "don't criticize if you can't do any better" maxim.
...
> [2] http://www-uk.hpl.hp.com/people/marbut/isTheSemanticWebHype.pdf

I have read Mark Butler's slide set above and, while it expresses a relatively 
extreme position, he backs it up and makes a good case.

I think it is a mistake to ignore such criticism merely on the grounds of it 
not being constructive: particularly since those grounds are false. Mark 
constructively suggests a plan of action: page 36, (including)

> Produce a standard XML serialisation of RDF which is
> - easier to understand
> - more compatible with existing XML tools
> Demonstatre:
> - how to use RDF/XML with existing XML tools

Having said that I think I believe that any further serialization work for RDF 
should either:
- be very limited in scope (XENT is a little too complicated for my liking)
- or be at a higher level than a syntax, but provide a means to annotate an 
arbitrary XML schema/relaxng/... to provide a mapping to RDF of a conforming 
document.

If you want easy to read and write then XML (of your choice) works.
If you want fully general then RDF/XML exists, and a subset of it might do; or 
XENT or some variant might be better. (RDF/XML suffers from the property 
problem).

I am a lot less convinced by Mark's arguments against the semantics of the 
semantic web - mainly because I am a thorough going non-realist. I note in 
human discourse we never can escape words and symbols to arrive at some 
platonic ideal, and I don't expect machines to do any better.

I also hope that we are getting close to the point where the bashing of M&S 
will stop ... I believe that the current working drafts are a significant 
improvement.

Jeremy
(I am not speaking on behalf of HP either)






  

Received on Monday, 16 June 2003 04:39:32 UTC