> It's not the case that XML vocabularies "succeed" if and only if they have
> a huge public practice. Reuters Health articles embed an rdf:Description
> element into html:head, and this "succeeds" in the sense that our
> customers successfully figure out what to do with it. We supply a DTD
> (and any other schema type on demand, but no demand yet) specifying how
> to mix the subset of XHTML with the subset of RDF that we use.
I am one of the people who has long argued in favor of namespaces because of
a perceived benefit-- but I have yet to actually need it. I am sure that a
SAX filter that split the events to separate consumers based on NS URI would
be great and I have always expected to need one... I just haven't yet.
With the above example I would say that is a "success" whether or not it has
a huge public following. My litmus test is usability. My only question is,
now that the DTD has been defined the need for namespaces is gone. To put it
another way-- is there anything that Reuters Health consumers are doing with
the data that actually requires namespaces. If the ":" were allowed in a
name (as it originally was) would their code change? If so, could they still
easily accomplish the same thing based on the prefixed name (what is now
prefix + local) instead of basing it on the URI?
Best Regards,
Jeff Rafter
Defined Systems
http://www.defined.net
XML Development and Developer Web Hosting
