RE: ID-ness is not the same as type (Re: [xml-dev] Re: determinin gID-ness in XML)

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

In relational dbs, one can autogenerate an ID, or choose 
a set member.  So the assignment of keyness is not the 
same as being a type.  Ok.  But just to be clear, why 
is that a mistake for XML?

len

-----Original Message-----
From: ht@cogsci.ed.ac.uk [mailto:ht@cogsci.ed.ac.uk]

Bob's comments remind me of one of the good things about the XML
Schema key/keyref mechanism -- it separates 'being a key' from 'being
a string/integer/date'.  SGML and then XML conflated these two, which
was with hindsight a mistake.

It would be nice if whatever solution we come up with didn't make that 
mistake again.  I _think_ the xml:idatt(s) proposal at least could be
understood to be _additional_ information about an attribute.