RE: [xml-dev] Re: determining ID-ness in XML

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

Right.  We use well-formedness because of the medium.  
What is biting us is the medium is based on addressing 
and it is unreliable in well-formed only documents because 
reliable addressing depends on typed information.

Ok, rising to the bait...

This has nothing to do with SGML. It is a system vocabulary 
problem.   The same problem is there for any notation processor. 
Understanding that simple fact was the brilliance of Hytime 
and failing to account for it is the arrogance of XML.

len

-----Original Message-----
From: Champion, Mike [mailto:Mike.Champion@SoftwareAG-USA.com]


> -----Original Message-----
> From: Bullard, Claude L (Len) [mailto:clbullar@ingr.com]

> Ever inscribe Arabic numerals into stone?  Requirements 
> count.

Thanks for elucidating a requirement that Roman numerals fufill better than
Arabic!  I hadn't thought of that ... and it does explain why you tend to
see Roman numerals on cornerstones and tombstones...  

Anyway, SGML *is* best for inscribing documents into "stone" so that they
can be read in 20 or 50 years.  XML evolved from it to meet use cases
requiring messages enscribed in "paper", and finding and cross-referencing
information is more important than persisting it forever.  xml:id is being
proposed because the requirement of cross referencing information easily
outweighs the requirement to leverage the SGML standards in many XML
applications these days.