XML-LINK questions

I have started to try to understand the XML-LINK draft with a view to 
authoring prototype documents and changing my DTDs.  Not having a background
in hypermedia I found parts of the draft hard going (and have failed to follow
all of the WG discussion).  I have posted my current understanding to
xml-dev (archived at http://www.lists.ic.ac.uk/hypermail/xml-dev) and would
be grateful for feedback as to whether I have 'got it right'.  If I have, there
seems to be a number of possible implications for the draft, including the
responsibility of the parser, and the question of validity of documents
containing XML-LINK constructs.  In that case, I might repost them here.


Peter Murray-Rust, domestic net connection
Virtual School of Molecular Sciences