[
Lists Home |
Date Index |
Thread Index
]
On Mon, 10 Feb 2003 08:16:43 -0600, Bullard, Claude L (Len)
<clbullar@ingr.com> wrote:
>
> If SOAP did not work because of the features it does not use, that would
> be a different issue. It does work. What is the problem?
"Unicode with angle brackets" demonstrably works. That doesn't mean that
the XML 1.0 spec, the whole 1.0 spec, and nothing but the XML 1.0 spec
demonstrably works.
Since XML "works" for many things, people are pushing it harder. The
problem is that when the envelope is pushed, the rough edges cause
considerable drag, in performance, interoperability, amount of code needed
to do what one would think would be simple stuff (such as combining
separate components into a single document), and so on. That's what unites
many of the permathreads here: SML, BinaryXML, InfoSet vs syntax, document
vs data ... people are finding that XML works, sorta, but conclude that it
could work better for them with a few lops or tweaks. Even authoring simple
documents -- a core use case if there ever was one -- is a pain in the butt
for non-geeks, hence the SML, WikiML, etc. things that people are
inventing to make their lives easier. Organized communities of users are
writing "interoperability profiles" that warn developers away from some of
bits of the XML 1.0 spec. Isn't that a sign of a "problem"?
Innovation and deprecation are taking place precisely because XML works
well enough to be worth learning, but not well enough to stifle arguments
about changes and alternatives. The question as I see it is whether the
keepers of XML should work to make the core of XML adapt to all this change
and experience, or let nature take its course and let the different
communities fork off. Reasonable people disagree on this, but few assert
that the SML, binary XML, Web services, etc. users of XML aren't motivated
by "problems" applying XML.
|
