RE: [RSS-DEV] Re: [syndication] Simple Semantics Resolution - RSS 2.0 Module
- Thanks Doug.
> I like the direction, but:You are absolutely right - there are at least three variables: the target
> - its based on XSLT. Maybe it would be prefereable to allow for a link
> to another file, like an python program, or a WSDL document to indicate
> a web services, etc.
> I would suggest you use instead an existing element from the RDDL
> namespace, where the nature URI indicates whether its xslt,WSDL
> document, etc, and the purpose would be "conversion to RDF".
format (RDF in this case); the type of processor (XSLT); and the instance
instructions (URI of the XSLT).
Although I hadn't considered either WSDL or RDDL specifically, I had looked
along these lines (DAML-S ;-) and decided not to aim for any kind of
generalization of the service description.
RSS 2.0 and its predecessors have gained popularity in large part due to the
perception of them as being 'simple'. Their simplicity relative to RSS 1.0
may be questionable once you start including namespaces in RSS 2.0, but I
don't think there's any doubt about the perception.
So I felt it was necessary to follow the principle of 'do the simplest thing
that can work'. Whether or not it is appropriate to apply this principle to
specs is another matter; likewise there is no other defence for accusations
of 'dumbing down'. But to have any chance of catching the RSS 2.0 flow,
anything that looked like it might appear difficult was pulled.
So although I very much appreciate your suggestions, and may well
incorporate them in a later version, right now the focus is on *simple* (and
I need to demo it working with XSLT ;-)