[ppml] Policy Proposal: IPv4 Soft Landing

[David Conrad]

Ed,

On May 14, 2007, at 1:16 PM, Edward Lewis wrote:
> I.e., from this is seems that PI is nothing compared to PA when it
> comes to draining v4.  That is, if I understand the statistic.

That was my understanding and why I decided to focus on PA.

> Maybe it has too many steps.  I'm sure that can be sanded down.

Yes, perhaps.

> I think the auditing cost ought to remain on the LIR/ISP as this is
> their request for space.

OK.

> I assume that this isn't retroactive.

I'm not sure what this means.

> I would like this to be a global policy (i.e., headed for IANA), if
> the global policy process would take too long then there there is a
> problem with that process.

I personally believe there is a problem with the process in that it  
takes so long, but dealing with that is a bit outside the scope of  
the "Soft Landing" proposal.  Ideally, I would agree that if there is  
consensus in the ARIN region that this policy is appropriate, it  
should apply to other regions since the obvious downside of not being  
a global policy is that it will strongly increase the incentive to  
"registry shop".  However, I thought it important to see how at least  
one community reacted to the policy proposal before considering  
pushing it global.

> Now for my soapbox...what's good about this is that it forces ISP's
> into the realization that they can either stay v4 and stop *growing*
> when numbers run out or they have to switch to something with a
> bigger address space (v6 is the one alternative at hand) and continue
> to be able grow customers.

Actually, it isn't a question about stopping growing so much as the  
cost of growth will no longer be predictable.  Those desperate enough  
(and with enough money/power) will always be able to obtain the  
address space they need.  However, I believe at some point, the cost  
of obtaining IPv4 (either monetary or in terms of bureaucratic red  
tape) is going to outweigh the cost of migrating to IPv6.  Part of  
the goal of the "Soft Landing" policy proposal is to gradually  
increase the (bureaucratic red tape) costs of obtaining IPv4 over  
time instead of going from (essentially) $0 to $<large, likely  
variable number> over night, thereby giving people time to adjust.

Rgds,
-drc