Bill de hOra wrote:
Daniel Aleksandersen wrote:
What I wanted was to question the standard. To me it appears broken.
Flawed, at least.
If you want to insist on this line, the flawed standard is XMLNS, and
its idea of a default namespace.
For the record: I agree that some people have problems understanding how
XMLNS works. That's certainly a problem with that spec, but there are
countless FAQs that explain it, and the XML parsers *do* get it right.
The problem is with people not using XML parsers to consume XML, or not
reading the documentation of the API (local-name vs name in DOM, for
instance).
What we should have done in the spec is make sure that we have multiple
examples showing different serializations for the same infoset.
What we still can do is publish a document (IETF Informational, web
site, whatever) giving and explaining examples.
BR, Julian
