modest suggestion for how to proceed

[Jari Arkko]

Keith Moore wrote:
> what you're suggesting is that it's more important to "hit" the market
> window than it is to produce something that actually solves a problem.

Dave Crocker wrote:
> far better to hit the market with a perfect product that is too late.

Obviously, both of you are right... what the IETF produces is useful
only if it both solves a problem AND comes out at the right time. If
its not useful it ... well... doesn't help anyone. And if it comes out too
late, vendors and users have already gone and done something else
Not good either.

So, we really have to do both. The question is, how? In my
experience the answer lies in the *scope* of the work. For
instance, in the case of this WG, do we want to a complete
understanding of *all* problems and the *full* situation
before trying to solve the first piece?

Jari