RE: languages and tags

["Miller, Robert (GEIS)" <Robert.Miller@geis.ge.com>]

Hi All,

Catching up following two weeks on the road, I waded through some but not
all (I'm only human!) of the messages related to this topic. It seems late
to respond to individual points made in the thread, so I'll just express my
opinion on the topic.

1) Element tag names may but need not be 'readable'.
	- Attributes & Presentation scripts can be used to provide
readability on demand
	- XML is intended to be processed by machines, not by humans.
That's why we view documents with browsers rather than with simple text
viewers.
	- Element names may sometimes be derived from human 'readable'
sources, such as field names in a database

2) The relationship between Element Tag Name and 'Semantic Identity' may not
be 1-1
	- The same 'semantic' element may be known by different or multiple
names across industries and countries.
	- A given 'semantic name' may relate to multiple 'Semantic
Identities' across all industries/countries, even though it relates to a
single 'Semantic Entity' within the scope of its industry/country use.
	- Where multiple sources contribute to the definition of element
names (e.g., multiple applications used to store and process data), XML
Namespace prefixes may be required to avoid name ambiguity

3) (Unique) 'Semantic Identity' is important to canonical simplicity.  Every
semantic unit should be uniquely identifiable through the value of a single
well-known attribute of an XML element.

4) The value of a well-known 'Semantic Identity' attribute can provide
direct access to a rich storehouse of metadata information concerning the
referenced entity.  

Cheers,
        Bob