Re: IPv7 Selection Criteria

[Noel Chiappa <[e-mail]>]

<Catching up on mail....>

    >> * Last a long time (20 years).

    > I agree that it's hard to quantify the lifetime of an architecture.
    > However, don't you think that this is a critical goal. I.e. wouldn't you
    > agree we've failed if the thing we pick is obsolete in 5 years?

    Would you care to offer some examples of concrete tests that can be used
    to determine whether a spec satisfies this requirement?

	Perhaps you'd be happier if it said something like "maximizing the
system lifetime" as opposed to putting a concrete number on it?
	I agree, it's hard to have a test. On the other hand, look at the
possibility of failing to meet this goal: wouldn't you agree we've failed
miserably if the thing we pick is obsolete in 5 years?

	Projected design lifetime, like robustness, is hard to measure
concretely (for many of the same reasons, since to some degree it involves
predicting the future), but again, I claim good engineers can do this.
	Just like you sort of get a feel for what is robust and what is not,
you develop the same feel for what kind of design methodology leads to
long-lived architectures: "clean design" and "flexibilty" are some things that
lead to this. Doing a lot of thinking forward to future needs and
possibilities (even if the mechanisms for doing them aren't completely
specified in the current specification) is something else.
	For instance, is capability C met by adding a kludgy bag to the side
of mechanism P, or have mechanisms P and Q been reorganized into X,Y, and Z,
so that capability C is now provided in a natural and flexible fashion? That's
the kind of thing you look for...

	Noel