Sorry to have not been involved in the disucssion. Vacation and all..
Based on the discussion with Graham I am at a loss as to how to fix
the document to satisfy your concerns. It seems that most of your
concerns are more to do with the entire W3C promoted web architecture
than with anything in particular with this proposal other than the
desire for the individual syntactic elements to be as semantically
free as possible.
I do have reservations with IETF endorsing the "W3C promoted web
architecture" and this proposal seems (deliberately or not) to be
encouraging us to move in that direction.
Is there anything that can be done to fix this document or are you
opposed to even the intended purpose of it?
I can see utility in this proposal. I can also see it causing problems.
I don't know how to get the utility without introducing the potential
for problems. Some proposals are like that.
I did include some concrete suggestions along with my original comments
that I thought would improve the document, and that I hope IESG
(and you) will consider.
Graham also made at least one suggestion I thought was valuable (there
may have been others that I don't recall offhand) - that the document
contain an explicit statement saying that these URIs are not to be used
for the purpose of translating existing protocols into XML, or at least,
that such a practice is nonstandard and discouraged. (I don't recall
his exact wording, but something like this could be very helpful.)
I continue to believe that if people want to import protocol elements
from an IETF protocol for use in another protocol that the proper
way to do this is to write an document about it, describing how those
data elements are used in the new context and (if appropriate) assigning
new names or URIs for those protocol elements -- NOT to assume that the
model can be imported into (say) XML merely by giving each protocol element
a URI without any change (intended or not) to the semantics of those
protocol elements.
Keith