I wouldn't take the same liberty as you did in
interpreting Sally's reply or with the IAB's original
intent, not even in a wider context. I wasn't nitpicking at
the wording either. The wording was just fine in saying what
it apparently meant to say -- single-int, because it would
have made no sense to require the end user to address
multiple intermediaries at the IP level. (Wouldn't an
OPES pipeline with the end user connecting to multiple
intermediaries separately look out of whack?)
What I've been driving at is whether clause 2.2 serves any
practical purpose at all, unless making life difficult for
OPES per se can be thought of as a practical purpose.
Anyway, the question had originally been addressed to the
IAB and the answer I got was frankly what I had expected
I'd get -- no pleasant surprise there. So be it then.
Thanks for having chimed in.
Joe Hui
Digital Island, a Cable & Wireless company
===============================================
-----Original Message-----
From: Ian Cooper [mailto:ian(_at_)the-coopers(_dot_)org]
Sent: Friday, December 07, 2001 5:18 PM
To: Joseph Hui; ietf-openproxy(_at_)imc(_dot_)org
Subject: RE: IAB document
--On Friday, December 7, 2001 16:10 -0800 Joseph Hui
<jhui(_at_)digisle(_dot_)net>
wrote:
-----Original Message-----
From: Ian Cooper [mailto:ian(_at_)the-coopers(_dot_)org]
Sent: Friday, December 07, 2001 3:33 PM
To: Joseph Hui; ietf-openproxy(_at_)imc(_dot_)org
Cc: Sally Floyd; Patrik Fältström; ned.freed-mrochek.com
Subject: RE: IAB document
[snip]
If implementors could argue their case that
use of multiple intermediaries
had the same overall effect as multiple proxylets on the same
intermediary then I think this would sufficiently demonstrate
that the recommendation had been addressed.
However contriving with proxylets, they still can't satisfy the
clause if the intermediaries, each being a one-trick pony,
run on different IP hosts (as I shown in my use cases).
But surely there's an implementation reason as to why each
intermediary is
a "one-trick pony", in that it's easier to deploy it that way
for some
reason?
I agree there's an issue with the strict wording of the text.
What I said
above was an attempt to address the *intent* of that text for
implementors.
From Sally's reply I think that's along the lines of how it
was intended to
be interpreted. (I think you've identified a problem in the
wording that
wasn't anticipated when it was written; that problem seems
tangential to
what IAB was actually addressing when they made the recommendation.)