On Thu, Dec 02, 2004 at 07:50:12AM -0500, Elliotte Harold wrote:
> This is *a* data model. It is not *the* data model. There is no one XML 
> data model, and there will not be.

I certainly think that's true for the forseeable future.  "Never is
a long word, even for an ent", but...

People doing a lot of work in DOM might find it very natural; people
programming with GConf might find that model natural; people using
JDOM or SAX or XML::Simple or XML::Twig might find those models
more convenient.  Someone using XML::Simple would find moving
to a DOM-style API (based on trees rather than unordered hashes)
a major change to the complexity of the code.  The simpler data
model doesn't work for arbitrary XML, but it makes the programming
easy enough that people were happy to use XML for configuration
files instead of inventing their own formats.

The value of XML is that people can interchange structured data
without limiting or defining exactly how that data is to be processed.
The extent to which XML documents are self-describing is the extent to
which people can take XML documents and process them in new ways.
Part of that processing is to build and manipulate some sort of
data structure, defined by a set of types and operations and often
called a data model.

There's no limit to the number of different ways in which a given piece
of XML can be processed.

So it's a strength of XML that there is no single data model and
no single API.

Liam

-- 
Liam Quin, W3C XML Activity Lead, http://www.w3.org/People/Quin/
http://www.holoweb.net/~liam/