RE: Schemas Article

[Dylan Walsh <Dylan.Walsh@Kadius.Com>]

> From:	Bullard, Claude L (Len) [SMTP:clbullar@ingr.com]
> Sent:	Thursday, June 07, 2001 2:37 PM
> To:	vdv@dyomedea.com; xml-dev@lists.xml.org
> Subject:	RE: Schemas Article
> 
>If all TREX does is validate, how will that be 
>any better than a DTD?

1. Namespace aware
2. XML syntax
3. AND in content model (from reading the interview)

I've been slightly perplexed by all the negative comments on XML Schema,
but I've found James Clarks interview
(http://www.ddj.com/articles/2001/0107/0107e/0107e.htm) to be the most
persuasive. He makes a strong case for seperating these parts:
1. making changes to the infoset (general entities in DTDs, PSVI in XML
Schema)
2. markup validation
3. advanced features like OO structures, relational constraints and
datatyping. In particular this is an area where you can't please all of
the people, all of the time.

Perhaps XML Schema 1.1 should modularize the standard in the same way
that XHTML 1.1 does. That said, I suspect that a lot of the negative
reactions are coming from people who have to implement it. For every
programmer who uses XML, what percentage have written a parser? It is a
very small proportion, and I think it will go down better with the
developer community at large than it has with the core people developing
the tools.