On Wed, Jun 16, 2004 at 09:37:41AM -0700,
Jim Lyon <jimlyon(_at_)exchange(_dot_)microsoft(_dot_)com> wrote
a message of 260 lines which said:
XML is rich enough to handle all of the above. The spec is well
debugged. XML Schemas include a mechanism for defining what's legal
in a document, even if part of what's legal is tags that haven't
been defined yet. There are dozens of extant parsers
You certainly know everything I'm going to write but many people on
the group are probably not XML experts, so I would like to emphasize
one point: XML is modular and not all parts have the same status,
speaking from an engineering point of view.
XML, the core, is indeed well debugged and understood, has efficient
(time and space) parsers for every operating system, every licencing
preference and every programming language.
But you say more: you talk about the rest of XML's very rich
world. You mention W3C's XML Schemas, someone else mentioned
Infoset. And these specifications are different standards (the W3C
says "recommandations") which are clearly not so well debugged and not
so well implemented.
To talk specifically about W3C's XML Schemas, they are not consensual,
there are at least two serious competing standards for XML validation
(three, with the old DTDs): W3C's XML Schemas and Relax NG (See RFC
3470 for a discussion). I personally prefer Relax NG. But both are
quite recent, have very few implementations and are far from being
widely mastered.
You cannot have your cake and eat it. Either you stick with XML, the
core, and most of the pro-XML arguments hold or you want to benefit
from the whole XML bestiary and many arguments (well tested, easily
available) vanish.