RE: What good is Restriction?

Understood. But the basic idea remains the same: identify the affected
element definition by context, apply the restrictions to that element. 

The trouble with using existing Schema syntax/structure and a
restriction/cancellation semantic is that by *not* mentioning a child
element, it's not clear whether it should stay the same or be eliminated.
Then you get into developing a secondary language (more operational in
nature) to describe how the referenced element differs from the original.

Not at all easy.

 -----Original Message-----
From: 	Biron,Paul V [mailto:Paul.V.Biron@kp.org] 
Sent:	Wednesday, July 10, 2002 1:39 PM
To:	'Eric van der Vlist'; Jeff Lowery
Cc:	xmlschema-dev@w3.org
Subject:	RE: What good is Restriction?


> -----Original Message-----
> From:	Eric van der Vlist [SMTP:vdv@dyomedea.com]
> Sent:	Wednesday, July 10, 2002 10:30 AM
> To:	Jeff Lowery
> Cc:	'Biron,Paul V'; xmlschema-dev@w3.org
> Subject:	RE: What good is Restriction?
> 
> On Wed, 2002-07-10 at 19:24, Jeff Lowery wrote:
> 
> > I thought XPaths were suggested as an alternative in this case. Why was
> that
> > rejected? It's used for keydefs.
> 
> They can be a real mess in some cases... for instance if you use
> substitution group to substitute one of the elements on the path!
> 
To elaborate slightly...

Schemas are defined at the component level...not at the XML syntax level.
Thus, XPath's are at the wrong level of representation.

pvb

Received on Wednesday, 10 July 2002 14:31:38 UTC