From: Dan Brickley (Daniel.Brickley@bristol.ac.uk)
Date: Tue Sep 05 2000 - 09:57:32 PDT
On Tue, 5 Sep 2000, Gerald Oskoboiny wrote:
> I haven't really followed the RSS evolution arguments closely,
> but I don't understand why namespaces are thought to suddenly
> make everything so much more complex.
There's something of a conceptual leap involved: pre-namespaces, there
was more of a 1:1 relationship between applications and XML-based data
formats. You might imagine a photograph metadata DTD, and MP3 metadata
DTD, an educationsal Web resources DTD, a Web sitemaps DTD, ditto for
channels etc.
With namespaces, each of these applications can draw upon the same
constructs more easily. Dublin Core ish concepts
(title/description/subject/creator etc) can be recycled rather than
re-invented. So this does simplify things, but there is of course a
tradeoff, since using a whole bunch of stuff defined elsewhere isn't
without social + technical baggage.
Talking of simplicity/complexity, Brian McBride recently circulated this
note:
http://www-uk.hpl.hp.com/people/bwm/rdf/jena/rssinjena.htm
...that shows how one might process the namespaces+RDF flavour of RSS
using his proposed Java RDF API. While some familiarity with the
nodes-and-arcs data model would help, this stuff (to my biased
eyes) doesn't look like rocket science. He gives code examples and the
whole thing's downloadable. The example shows how an RSS-centric app
might use a generic RDF API to generate HTML view of some RSS data.
Feedback from those only dimly-familar with the RDF model would be
interesting...
Dan
This archive was generated by hypermail 2b29 : Tue Sep 05 2000 - 10:00:55 PDT