Eliot Lear wrote:
First, I think we are far afield if we are simply arguing about whether
working group consensus is necessary but not sufficient for protocol
success. Procedurally that is what we have to work with. Right?
Right. We don't have to wrry about the braoder IETF community taking
exception, the Security area shooting holes, or AD Discusses. And heaven
knows, we don't have to worry about actually market uptake of our output or
whether the protocol actually works.
More to your text above, there are many factors as to why IPv6 is not
broadly deployed. The simplest factor, however, has nothing to do with
complexity OR functionality: it's simply cheaper to not change
operationally until it's not.
Quite a few people consider the feature-bloat of IPv6 and its incompatibility
with IPv4 to have been critical barriers to adoption. Were IPv6 made a fairly
minor upgrade to v4, then its initial adoption would have been a fairly minor
upgrade to existing software.
We're down to a basic Occam's Razor discussion. What functionality do
you wish to eliminate that makes the state machine simpler, and does
that make the solution to the problem we are attempting to solve simpler
than is possible (meaning are we no longer solving the problem we agreed
Yup. Nicely said.
One wrinkle is that it's clear different people are trying to solve different
problems. This exercise is often useful to uncovering incompatible goals.
NOTE WELL: This list operates according to