Re: IPv7 Selection Criteria

[Noel Chiappa <[e-mail]>]

    In other words, Noel, if we cannot objectify application of the
    criterion, I believe that the criterion is inherently useless for
    the task at hand

In other words, if you can't measure it, it doesn't matter/exist? Sorry if I'm
getting snappy, but to me this is the single most important goal of a design
of a large, shared, communication system, and to ignore it because it's very
hard to measure seems to me insupportable.
However, I think your next phrase gives me the clue to what the real
disagreement is here, so let's go to that...

    since we will then just invite "yes it is"/"no it isn't" debates.

	You and I have clearly have a different model for what is likely to
happen with this thing. I doubt we are going to sit down, as an IETF comittee
of the whole, with this checklist in hand, and fill it out for each proposal,
and either vote on which one to pick, or go with the one with the most check
marks, after which we will all happily go work on that one.
	People who back scheme X, and truly believe in their hearts it is the
right thing, simply aren't going to go away because an IETF vote (whatever the
heck that is, and whatever the heck it means) went to something else. This is,
after all, the whole tradition of the IETF; a group of people who *ignored* a
formal vote on what the 'right thing' was, and proceeded to guerrilla deploy
their stuff into a major contender.
	What I see is most likely to happen is that each credible possibilty
will be designed, implemented, and deployed in some scale. This is not
necessarily a *bad thing*; done right, it can result (albeit at the cost of
some expenditure of design/implementation resources) in a better designed
thing at the end. (The challenge, as always, is to miminimize the bad feeling
between the design teams, and confusion to the outside world, while this
process is going on.)

	In this process, the criteria document more represents advice/input
from the user/engineering community to the design teams as to what are
perceived to be important goals for successful designs; a requirements
document, if you will.
	In that context, hard to qualify goals like "maximum design lifetime"
aren't quite as useless, yes?

	Noel