Re: add Turtle examples to specs

Danny Ayers wrote:

> I would imagine endorsement of any  'plain' XML syntax for RDF aside
> from RDF/XML would be difficult, given the number of proposed
> alternatives out there. Whether there would be much to be gained from
> this is another question - the horse has already bolted in regards to
> many in the XML community (and that's before considering the issues
> around OWL++ serialisations).

A few years ago in Brisbane there was a panel sort of taking the cover 
off of the then-fairly-new idea of RDF as the key to indexing/annotating 
whe Web via metadata.

The "what" of it was quite overwhelming and the "how", which is now (as 
it was then) being picked apart for its various supposed 
near-impossibility of comprehension, was little debated - more just 

Tim was in the position expressed in a Japanese Senryu poem: "The 
champion takes lightly all kinds of advice."

Just as earlier he had been almost ridiculed for not perfecting 
HTML/HTTP by those who thought it would have been better to wait another 
decade (until some of the wish lists could be included) to actually 
implement hypertext, so it seems we tend to confuse the details of 
implementation of metadata implementation with the problem of teaching 
it rather than of using it.

This pattern is actually a recurrence of some reactions to SGML and now 
a derivative thereof, XML, is seen as unfit for service in regard to 
being the lingua franca of metadata via RDF.

There is an old notion that clearly applies here: insistence on the 
"perfect" delays the "good".

Just as many nags will be raised if some "friendlier" combination of 
acronyms than XML/RDF gets chosen,, so N3 will be too 
"limited/simplistic" or triples need to be quads or...

It's a lot easier for critics to postulate perfection than for the 
creator to satisfy their complaints.


Received on Saturday, 11 March 2006 11:56:57 UTC