> Tim's point was that you couldn't have a native XML programming
> language because there is no data model for XML. Fact is, there are
> many. Pick one.
I picked mine Jonathan. The XQuery and XSLT Data Model.
It's the only one among all possibilities that satisfies ALL my
requirements:
1) be an abstract data model, not an API that implies a certain data
access pattern
2) cover the entire Infoset
3) cover schema validated as well as non schema validated data
4) be usable
5) have at least one data processing language based on it
(there are actually three: XPath 2.0, XQuery and XSLT 2.0)
6) show evidence that it can be implemented
7) co-exist peacefully with the other XML data models
That's my reasoning.
Best regards,
Dana
