I want to start building a prototype using XML-Data, and probably
Microsoft's XML Java parser. Am I wasting my time building something to
this spec, or is the current RDF completely different?
Assuming that it's useful to proceed... I've created a small schema
according to xml-data and successfully parsed it using the DTD from Appendix
A. My question (finally) is how can I use this schema to validate an XML
file that conforms to it? After parsing the schema, should I convert it
within the XML processor to DTD objects, then proceed _as_if_ the schema
really originated from a DTD file?
It seems wrong to duplicate the existing validation logic that works for
DTD's and create another for schemas. This is probably part of the argument
in support of those who say that the xml-data schema is a bad idea, and we
should write all schemas directly in DTD syntax. However, coming from an
artificial intelligence background, the idea of a metadata representation
language appeals to me.
Thanx for any thoughts or advice!
Dave Carlson
Ontogenics Corp.
Boulder, CO