• From: Dave Pawson <davep@d...>
• To: xml-dev@l...
• Date: Thu, 2 Dec 2010 08:13:08 +0000

On Wed, 1 Dec 2010 16:45:04 EST
BillClare3@a... wrote:

>  
>  
> Several proposals have been made here to  integrate a variety of 
> fundamental XML standards.   
> Most take an approach of jamming some subset  of a disparate variety
> of standards into a common structure that maintains all  their
> idiosyncrasies and inconsistencies.

Not the source of this thread Bill, see James blog entry. 


>  An alternative approach is to abstract
> from competing  specifications and generalize their essential
> substance.  This can provide simpler and more flexible foundations
> with which to rebuild higher level concepts. 

JC had a target, without one you'd have 42 different definitions of
'essential'? 



>  With such
> foundations,  compatibility with existing name spaces can be
> maintained but deprecated.

I'd go further and not require any full level of compatibility,
following the SGML to XML model. The goal is a cleaner smaller
better integrated subset suitable for the web/mobile/2010. 
That would drop lots of Amelias hard to explain bits. 

> Among these fundamental concepts for language foundations  are : 
> ·         Data Types – with  nesting, inheritance, extension,
> constraints and executable methods 

Is that really a required part of a new SXML? I'd put Schemas 
and datatype validation as an XML application. 


> ·         Data Attributes –  especially extended specifications for 
> metadata, storage, communication,  presentation, dynamics, etc. 
Growing XML? In which case I don't understand this item?
> ·         Data structures –  atoms, lists, hierarchies, networks 
As above? How to justify making SXML more complex than XML?
Wheres the market for this?

> ·         Expressions –  arithmetic, logical, selection, iteration,
> path, set, query, etc. - with basic  and extendible types.  
 Again an application? Not a part of basic SXML? Based on xpath?


I  don't understand the rest of your post in the context of JC blog.
It seems to move in the direction of XML applications more and more.





-- 

regards 

-- 
Dave Pawson
XSLT XSL-FO FAQ.
http://www.dpawson.co.uk

• References:
  • Re: RE: James Clark: XML versus the Web
    • From: BillClare3@a...