RE: Why XML Schema enforces UPA (was Re: a or b or both - mystery ..)

["Bullard, Claude L (Len)" <clbullar@ingr.com>]

I'm circling back to the concepts Henry talks about 
in his presentations discussed a while back, that 
we are *always* processing infoSet values.  Where we may 
differ is that if that is so, then the infoSet is 
not defined by the schema.  The schema may add 
information to the infoSet by adding values, 
and so may other applications (all of these 
schemas so far with the exception of DTDs 
are schemas).  The infoSet definition itself 
is immutable, or closed with respect to the 
extent of the process.  It can only be changed 
by the outer process; that of human specification.

Len 
http://www.mp3.com/LenBullard

Ekam sat.h, Vipraah bahudhaa vadanti.
Daamyata. Datta. Dayadhvam.h


-----Original Message-----
From: Murali Mani [mailto:mani@CS.UCLA.EDU]

On Wed, 18 Apr 2001, Bullard, Claude L (Len) wrote:

> I think Henry is right similar to the way the Grove guys
> were right.

Where is Henry right?
I say that we almost always end up with non-deterministic content models
during document processing.
I also say that unique particle attribution as in XML Schema is *very*
exaggerated, if we need UPA (i doubt if we need it), we need UPA as
provided by unambiguous regular tree grammars.

regards - murali.