The spec is much improved. I notice the "Key concepts and components"
section is still missing a few pieces. Most notably there's no notion
of an XML Name. I wonder if we also want to include comments,
processing instructions or the DOCTYPE declaration here.
I'm also not happy with the demotion of DTDs to just another schema
language. In particular valid is well-defined in the XML 1.0 spec in
reference to DTDs only and exclusively. The notion of validity given
in this article appears to contradict the language of the spec. A
document can be valid against a RELAX NG schema but still be strictly
invalid according to the XML 1.0 spec. Which is not to say that I
prefer DTDs to other schema languages--I don't.--but I do think we
need to be careful to follow the spec and the spec's definitions here.
I suggest we add a section on DTDs and validity rather than merging it
in with other schema languages. For better or worse--probably worse,
the XML specs do specially preference DTDs over other schema
languages.
--
Elliotte Rusty Harold
