[
Lists Home |
Date Index |
Thread Index
]
Mike Champion wrote:
> I was not at all persuaded by the "Sam the XML guy and Kerry the RDF
> guy" fable. This basic approach can work if it is abstracting out the
> essence of a lot of real-world cases, but I doubt if those exist for
> this "RDF kicks raw XML's butt" scenario. That gets to the heart of many
> people's skepticism about RDF -- it *sounds* like a good idea, but one
> doesn't see very many people actually reaping all those benefits.
It's a fable, so you can moralize about anything. But even so, why
is Kerry using XML at all? All that overhead when n-triples would
interchange just as well. And it seems Sam could do with an
introduction to XPath :)
I guess people don't reap the benefits because they're in a hurry
and don't always want tommorrow's complexities dealt with today's
overhead. That's why RSS1.0 was resisted, it was too much to chew
off at the time (RDF isn't entirely suitable for drip feed or
evoutionary development). On the other hand, getting from XLink to
RDF isn't so hard, so using XLinks or something like them in an
RSS2.0++ could be a cleaner approach than forcing namespaces into
RSS infrastructure.
[aside: It's interesting to watch RDF gradually align itself with
the relational model. Maybe RDF will wind being some kind of RDB
interlingua; if so I'd like to see some articles explaining how to
make your relational model RDF friendly :)]
Bill de hÓra
--
Propylon
www.propylon.com
|
