[ppml] Posting of Legacy RSA and FAQ

michael.dillon at bt.com michael.dillon at bt.com
Mon Oct 15 23:48:12 EDT 2007

> These are political issues, not technical issues.  Sure, from 
> a technical standpoint, anything is possible.  Technically, 
> legacy holders could indeed be offered a succession of RSA's, 
> one after another, until all of them sign.
> But, politically it won't happen.  Trust me on this.  It 
> won't.

I am with Ted on this one. In politics, timing is everything, and if
this legacy RSA fails, then the exhaustion of IPv4 addresses is so close
that it doesn't really matter anymore. Just about everyone is now
working on their IPv6 transition strategies, and the energy will all go
into getting that right, because those companies who do get it right,
will have smooth sailing a couple of years from now.

> For a legacy holder to turn it's back on the current effort 
> is politically unwise.  It is, in fact, political suicide.  

Yes. I can't see why anyone would put any more effort into bringing
legacy IPv4 holders on board if they get slapped in the face over the
current proposal.

> Well, that's your opinion.  I thought my tone wasn't 
> antagonistic enough.
> Dean and Randy and others like them aren't going to 
> understand that they cannot go against the rest of the world 
> and win, until they try it and suffer contusions.  But, some 
> of the folks that they are leading around by the nose may 
> come to their senses after reading my statement and decide it 
> is in their best interest to work with the system rather than 
> fighting it.

>From a political point of view it might be advantageous to have Randy
and Dean opposing your position. What many engineers fail to understand
is that politics does not demand perfection. If anything, politics
prefers imperfect solutions because they tend to be optimal. Engineers
often think that an optimal solution means that it is best according to
some engineering criteria. In fact, the engineering approach should be
to understand the problem as one of bringing a system into a state of
equilibrium which can be maintained with low energy. If there are
several local minima that are close to optimal, then a random choice is
usually made. The "energy" being measured here in the system is the
"heat" of opposition. An optimal political solution is the one which
most people hate the least. There may be a few people which hate it a
lot, but that gets averaged out by the other members of the policymaking

Therefore an excellent political solution that is BEST FOR THE OVERALL
COMMUNITY is still likely to have one or two vehement opponents.

> And, I apologize to any and everyone 
> that is uncomfortable with it, but I am merely relating what 
> has happened thousands of times before in these kinds of scenarios.

Unfortunately, a lot of the players in the ARIN community come from a
technical background and are used to finding the kinds of consensus
solutions that are usually possible in the technical arena. They often
forget that the second P in PPML stands for "policy" and that means that
ARIN policymaking is a political forum, not a technical engineering

And in any case, it will all be moot about three years from now when
growth shifts to the IPv6 Internet. Let us not forget that.

--Michael Dillon

More information about the ARIN-PPML mailing list