Summary of changes coming in model-09.txt

["Mark Day" <markday@cisco.com>]

The main comments that I received on the model draft were from Phil Rzewski
and Ian Cooper in messages sent to the list. I have incorporated all of
Phil's changes (8 in all) and most of Ian's (14 of 17). In this message I
just highlight the changes Ian asked for that I didn't do, plus point out a
couple of things that I had to change as consequences of the comments and
anticipating movement toward RFC status.

Things Ian asked for, but I didn't do

1. I left PUBLISHER unchanged, since I think we did leave the definition
somewhat vague deliberately.

2. I left "There are limits to how large..." instead of making it
"There are limits as to how large..." since the latter didn't strike me as
clearly superior.

3. I haven't changed any of the "Note:" parts in section 3.  I am not
convinced that those should be removed or modified before the draft goes to
RFC.

Some other changes

1. It was pretty easy to arrange the text and references so that the model
document can go to RFC and not get hung up waiting for other drafts to
become RFCs, with one exception. That one exception is the architecture
document, which is referenced both in the security considerations section
and as a source for additional technical detail about the terms being
defined.  So something we'll have to consider on the list and in Salt Lake
City is whether we can/should make changes to the model document to
eliminate those references to the architecture document. Otherwise, the
model will have to wait for the architecture document to go through another
round or two of revision.

2. After re-reading Ian's comments and parts of RFC 3040 a couple of times
each, I think that the CDI definition of surrogate is compatible with the
definition of that term in 3040, and I put that claim into the draft. This
may be controversial, I'm not sure.

--Mark