[Home] [By Thread] [By Date] [Recent Entries]

  • To: 'Peter Rodgers' <pjr@1...>, xml-dev@l...
  • Subject: RE: Universal Schema to Validate an Arbitrary Well-Formed Document
  • From: "Bullard, Claude L (Len)" <len.bullard@i...>
  • Date: Wed, 19 May 2004 10:48:06 -0500

I don't understand that.  You are saying that the decision 
to name the top level element is too constraining?

len


From: Peter Rodgers [mailto:pjr@1...]

On Wed, 2004-05-19 at 16:25, Dare Obasanjo wrote:
> It can't be done. You need to at least provide a name for the top level
element. What you can do is specify that the content model of an element can
be any content.   

That's what I'd feared.

This seems a considerable flaw in XML Schema and by association WSDL.

To elaborate on why we wanted such a schema...

It's our practical observation that a loosely-coupled document based
web-service ought only apply constraints to enforce tight/tighter
coupling as necessary (typically production) but when possible the
constraints are best kept loose - this is particularly the case during
development, when data model and process may be in flux.

In other words the development philosophy is - create the
service/process and constrain after the fact as necessary.

It seems like XSD and WSDL approach the world with the opposite
philosophy mandating that a schema be developed before the service.

Site Map | Privacy Policy | Terms of Use | Trademarks
Free Stylus Studio XML Training:
W3C Member