[arin-ppml] Draft Policy 2011-6: Returned IPv4 Addresses

[Owen DeLong]

On Feb 23, 2011, at 6:51 AM, William Herrin wrote:

> On Tue, Feb 22, 2011 at 11:01 PM, Chris Grundemann
> <cgrundemann at gmail.com> wrote:
>> On Tue, Feb 22, 2011 at 20:41, William Herrin <bill at herrin.us> wrote:
>>> On Tue, Feb 22, 2011 at 10:33 PM, Chris Grundemann
>>> <cgrundemann at gmail.com> wrote:
>>>>> merging legacy
>>>>> considerations with address space that is clearly controlled by ARIN
>>>>> is likely to result in confusion.
>>>> 
>>>> Actually the opposite is true. Separating IPv4 address space along
>>>> substantially arbitrary lines is likely to cause much more confusion
>>>> than making clear policy that applies to all IPv4 addresses.
>>> 
>>> You're both right, but in different contexts.
>>> But here's the context: legacy addresses are an exception.
> 
>> There is no such thing as "legacy" addresses.
> 
> Chris,
> 
> I stand corrected. Marty had the right of it: you're denying the facts
> on the ground.
> 
No, Chris has it right....

There is no such thing as legacy addresses. There are addresses which
currently have legacy status, but, if those resources are returned,
transfered, or otherwise reregistered or freed, they are no longer
in a legacy status.

> It's not unreasonable to write policy which changes those facts but if
> the changes are to be clear, they must reference the facts as they are
> both before and after the change. The facts today are: there IS such a
> thing as legacy addresses and they're presently immune to most
> ARIN-region policy. Moreover, that policy rarely says so explicitly.
> Hence policy which should impact legacy addresses will be more clearly
> understood if it explicitly says so.
> 

The part I think you are missing from Chris's point is that these are
legacy registrations. The resources themselves are only in a legacy
status so long as the current registration remains in effect.

Owen