[ppml] Policy Proposal: Resource Reclamation Incentives

On Wed, 4 Jul 2007 michael.dillon at bt.com wrote:
> 
> In any case, at some point we have to say enough is enough. We give
> people the opportunity to play ball but when they refuse to come on
> board, we have to act unilaterally. I believe that we are now at the
> point, due to IPv4 exhaustion being so near, where we have to give
> people ONE LAST CHANCE, and after that act unilaterally.

I think it is rather funny that people argue on one hand to loosen IPv4
allocation policy, and on the other to tighten reclamation.  Are you
trying to create a artificial crisis?

It seems to me that 'Resource Reclamation" is a policy in search of a
problem to solve.  When IPv4 runs out, people will use pure IPv6 and
gateways to old IPv4.  There is also a false dichotomy. IPv4 allocations
are already automatically IPv6 allocations, so there is no need for
anyone to exchange IPv4 for IPv6 [it would be kind of stupid, even].  
An IPv4 address is automatically an IPv6 address. There is only a need
to convert V4 wire protocols to V6 wire protocols. Address translation
is defined from IPv6->IPv4.  Traditional NAT is only needed for the
return path (IPv4 -> IPv6).  This is not an terribbly difficult problem,
and becomes easier as more ISPs are native IPv6.  And I think paths over
native IPv6 (e.g. IPv4 - IPv6 - IPv4) should be transparent.

So, there is no need for any unilateral actions for IPv4.  There may be
some need for some software/hardware for the V4-V6 gateways, but that
isn't anything ARIN can fix.  These gateways might not be easy, but that
difficulty isn't ARIN's problem, either.

I think the real issue and motivation is the size of the routing table.  
Some companies have always filtered /24s and anything smaller than /16
from classB nets, etc. They did this to reduce their table size. But
IPv4 resource reclamation won't solve that problem for more than a few
months, while people get IPv6 allocations, which take up even more space
per route, so this is a very shortsighted and contentious policy.

The recent example of running out of RFC1918 space on cable modems is
solved by using IPv6 space, behind a traditional NAT.  Again,
software/hardware, not policy changes, is required.

> We won't be acting arbitrarily because ARIN is an open organization
> and any actions will be discussed openly until we come to a consenus.
> This is our strength and I believe that when ARIN finally does act, it
> won't polarize anthing and will not create a mess.

Wasn't it claimed that the IETF/ISOC would be fair, open, and consensus
based? And before that, Nanog claimed to be open and consensus based.  
Yet both organizations have created false consensus by dirty tricks such
as blatantly lying about the consensus results, or creating false
consensus by silencing the opposition, or acting in conflict of interest
upto and including frauds.  These aren't really very fair or very open,
or very honest. None of what they did in those cases was in the public
interest. And you (Dillon)  participated in both organizations without
objection to blatant dishonesty and unfairness, so I have to wonder
about your assurance of a "fair and open" process.  If past experience
is any guide to what you apparently consider "fair and open", then I
have to conclude that we should all be very worried.

Examples of IETF/ISOC and Nanog misbehavior are available, and some have 
previously been cited.

		--Dean