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

[Matthew Kaufman]

On 12/30/2010 8:29 AM, Chris Grundemann wrote:
> On Thu, Dec 30, 2010 at 08:52, Kevin Kargel<kkargel at polartel.com>  wrote:
>> Umm, If you say that the only way I can have any IPv4 is if I have working IPv6 seems to be an attempt to force me to adopt IPv6.
> Well, prop-125 states that if an org wants any *more/new* IPv4
> addresses, they need to show that they are actively deploying
> production IPv6.
Why? And why apply it to transfers? (Which I assume are "old, used IPv4 
addresses")

There are any number of reasons (including my hypothetical "legacy 
computing museum" example from yesterday) why one might have a 
legitimate need for IPv4 addresses and no need, reason, or even *way* of 
deploying production IPv6 for those same devices.
>   The requirement is to use new IPv4 addresses in the
> most efficient way (the policy does not force anyone to request new
> IPv4, only to use it in the communities best interest if they do).
There is nothing other than your (somewhat suspect, see below) value 
judgment determining that deploying IPv4 addresses without parallel IPv6 
addresses is somehow "inefficient".

Why is putting 100 legacy machines on the public network of less value 
than putting 100 new machines on the public network? As I asked 
yesterday, who is to decide whether an SSL-protected porn site is more 
or less valuable than a publicly-reachable VAX running 4.3bsd or an 
original NeXTstation, or a sensor device that the local flood control 
district needs to reach and won't be replacing for 10 more years? It 
shouldn't be me, you, or ARIN.
>> What are you going to do about the multitude of networks that have no access to native IPv6?
> What am I going to do personally? Well, for one, I work for a backbone
> provider that offers IPv6 throughout the US.
Aha. So you have a financial interest in having quicker deployment of IPv6.
>   For another, I am working
> to connect folks who need to deploy IPv6 with those who can help:
> http://www.theipv6experts.net.
And an interest in having people who don't need or understand IPv6 being 
forced to somehow show their IPv6 expertise.
>   I am also supporting prop-125 which has
> the potential to encourage more access to native IPv6.
After the first two reasons, this is no surprise then.
>   No one who needs more IPv4 is going to get all that they need
> at this point.
Agreed. There's a limited supply coming from ARIN, which I hope will 
continue to be available under policies that aren't continually shifting 
as misguided attempts to optimize the last few pieces such as this one 
change the policies repeatedly during a time when ARIN should be focused 
100% on IPv6.

And there's a transfer mechanism, which shouldn't be muddied up by 
ridiculous IPv6 requirements that make no sense.
> The question that remains is who should we reward with
> the crumbs: Those who did not act, who still have not acted _or_ the
> folks who listened and who did deploy (or at least are now deploying)
> IPv6?
We shouldn't be "rewarding" either. They're both trying to get IPv4 
addresses, which are running out. Not one of us knows which of those two 
groups has the more compelling reasons to get space. Yes, it does make 
sense to set aside *some* of the space for people who need IPv4 
addresses *specifically* to operate their IPv6-IPv4 transition 
technologies, but that's it. Once you've set that aside there is nothing 
that makes any one need for IPv4 space more important than another 
unless you want to have a "homeowners association" that starts reviewing 
each application on its public-interest benefits and choice of web page 
background color.
> In both scenarios organizations will be hurt, perhaps even
> destroyed but that does not make the two choices equal.
I disagree. With the sole exception of IPv4 space reserved for 
transition technologies, the choices *must* be equal in the eyes of ARIN.
>   Leave the
> trees for a moment and do the right thing for the forest.
Often the best thing for the forest is to *not meddle*. I say we leave 
IPv4 policy as-is and worry about the forest on the other continent.

Matthew Kaufman