[arin-ppml] *Spam?* Re: Discussion Petition of ARIN-prop-125 Efficient Utilization of IPv4 Requires Dual-Stack

[Matthew Kaufman]

On 12/30/2010 11:36 AM, William Herrin wrote:
>
> Then your answer is no, you're not capable of working the problem one
> piece at a time? You can't help the prop 125 guys come up with the
> least harmful force-IPv6-adoption proposal and THEN consider whether
> the resulting proposal is still inappropriate?
>
> Come on Matthew, tell me your better than that. Let's help the prop
> 125 guys find the least objectionable version of their proposal and
> put off the talk about adoption or rejection of the general concept
> until later.

When I was young and naive, I signed all ballot proposition petitions 
that were stuck in front of me, figuring "every idea should be allowed 
to come to a vote, no matter how flawed".

I've since learned that there's a whole lot of things that get the 
popular vote that aren't actually good.

So no, I'm not interested in helping them make this bad idea more 
palatable to the masses. It is a bad idea. I've pointed out why it is a 
bad idea. As I've pointed out it is an *especially* bad idea if applied 
to transfers... but I don't want them to take transfers out of the 
proposal either, because doing so will make it *more likely* that it 
will pass.

And there's already a great way to force everyone to adopt IPv6, really, 
and this isn't it. The only force that will make people put IPv6 on 
servers is when there are eyeballs that have IPv6-only service... and 
the only force that will make eyeballs demand IPv6 is when there are 
services they want to reach which are only available on the IPv6 
Internet. And none of that happens by giving out IPv4 addresses to 
people who are deploying IPv6. In fact, giving them IPv4 addresses 
*delays* the inevitable.

The sooner runout happens, the better. That means that delaying IPv4 
runout by forcing people to jump through more hoops is a *bad* thing.

So... you want a policy to "force IPv6 adoption"? Stop giving out *any* 
IPv4 addresses. The only way for ARIN to do that is for ARIN to not have 
any more, as I doubt we could get a "reserve all the rest, immediately" 
policy through soon enough to matter, and this policy is the opposite.

Matthew Kaufman