RE: what the "scope" disagreement is about

Tony,

Having picked on Keith in public a few days ago, it is your 
turn, and I'll make the same apology in advance.  Even on the 
parts of this with which I agree, the content and style of the 
discussion is no longer helpful. So, with the understanding that 
I'm just a random member of the community who has no authority 
to tell anyone to do, or not do, anything, I want to make two 
observations and a few suggestions to you and everyone else who 
is posting to this list.  Fwiw, if I hadn't had enough 
experience working with both you and Keith to value the 
experience and perspective of both of you when you are calm and 
thinking and explaining yourselves clearly, I wouldn't bother.
First of all, you have managed to convince me and, I think most 
others who have been listening, that "get rid of SL" is an 
answer to the wrong question.  Whether it is the right answer or 
the wrong one is less important than identifying the right 
question and then sorting things out in that context.  I think 
you have been trying to identify the right question, but your 
doing so has gotten repeatedly lost in the noise.
Second, it is clear to me that you and Keith have a profound 
disagreement about the character, state, and evolution of the IP 
architecture. The precise nature of that disagreement has been 
sufficiently obscured by differences in vocabulary and arguments 
about side-issues that I imagine that most readers of the list 
can't crisply identify the disagreement.  I'm not even sure you 
and Keith can do that at this stage; I suspect, from the level 
of mis- and non-communication, that you cannot.  I can try to 
explain it, but not this week (see below).
For whatever it is worth, I'd like to encourage everyone to take 
about a week off from this.  The arguments that are being made 
today don't differ very much from those that were being made in 
late March.  Anyone who hasn't been convinced yet --one way or 
the other-- is, IMO, much more likely to be convinced by quietly 
thinking about the arguments and statements, and generally 
letting things sink in, for a week or two than by more traffic 
that repeats the arguments yet another time.
And, for several other participants... There really are some 
subtle and very important architectural issues here (a subject 
on which I think Tony and Keith agree).  _Please_ invest the 
time to try to read and understand the postings and the issues 
they are trying to address. They are not simple, or at least the 
right solutions aren't: if they were, I really hope we wouldn't 
be having discussions this lengthy or heated.   When you read a 
few sentences of a posting and then fire off a response based on 
what you think that poster, or previous ones, are suggesting -- 
and more than a few postings have clearly resulted from that 
behavior-- you mostly just add to the noise.  And the noise is 
getting seriously in the way of looking carefully at those real 
problems.
       thanks,
          john


--On Friday, 02 May, 2003 12:44 -0700 Tony Hain 
<alh-ietf(_at_)tndh(_dot_)net> wrote:
> Keith Moore wrote:
> > NO, Tony.  We need for the network to do routing, not the
> > hosts.
> >
> > Why is this so hard for you to understand?
> It is not hard to understand, because I agree there is no need
> for the host or app to do routing. At the same time, the host
> or app doesn't need to be passing around topology information.
> If any process chooses to pass around values that are
> explicitly expected to be directly used as topology locators,
> that process needs to understand the topology it is
> describing. This isn't rocket science ...
>
> Tony