Re: initial issues

[Thomas Narten <narten@raleigh.ibm.com>]

Eliot,

> I half agree with you on these points, and I might even agree with you
> 100% depending on what you meant.  As we saw in the BoF in San Diego,
> multihoming can mean different things to different people.

Which is why a good starting point would be to gather requirements, do
a taxonomy, etc. We need to understand what the constraints of the
problem are before we look too much at specific approaches.

> What it boiled down to was this: multihoming at what level?  I
> believe we should develop proposals that address this in their
> entirety so that we do not make the same mistake we did years ago by
> excluding ideas that merely needed time to mature or otherwise
> mutate into things more useful.  I wouldn't want to take forever,
> but I also wouldn't want to move in haste.

We might want to see if we can walk before we try to run. :-)

To me, reasonable short-term goals would be to see if we can come
together on what multihoming actually is, then look at some general
approaches and see if we can home in a small number of the more
promising ones that we should pursue in detail. Beyond that, let see
where we are when we get there. I.e., the WG can be rechartered to do
more when/if it becomes clear what the more is, or perhaps some sort
of work should be sent to the IRTF, ...

My $0.02 anyway.

Thomas