[arin-ppml] ARIN-prop-172 Additional definition for NRPM Section 2 - Legacy Resources

[Owen DeLong]

On Jun 16, 2012, at 2:03 PM, David Farmer wrote:

> 
> 
> On 6/16/12 12:09 CDT, Chris Grundemann wrote:
>> On Fri, Jun 15, 2012 at 7:54 AM, David Farmer <farmer at umn.edu> wrote:
>> 
>>> This is hindsight, but lets learn from it and fix the problem. "Legacy
>>> Resources will not be reclaimed solely for lack of use" should just become
>>> policy in the NRPM.
>> 
>> I can not disagree with you more David.
>> 
>> Mainly because there is no such thing as "legacy resources." There are
>> legacy resource holders and it has become common for some folks to
>> refer to the resources allocated or assigned to them to be called
>> legacy, but there is nothing (not one thing) special or different
>> about those resources.
> 
> Actually, it is neither a property of a resource or a resource holder independently. Legacy status a property of both a resource and its resource holder together, it is the relationship of the two that has legacy status.
> 
> New resources can be assigned to a resource holder that has legacy resources without the new resources having legacy status. A legacy resources may be transferred to a new resources holder (through 8.3) without the new resources holder having legacy status. Neither of these would be true if the resource or the resource holder interdependently held legacy status.  This is because legacy status is a property of the relationship between a resources and the resource holder, and of neither independently.
> 
> However, I contend that when a resource holder updates its name or changes its ownership through (through a 8.2 transfer) the relationship between the resource and the resource holder is not changed and therefore the legacy status is maintained.
> 

That has been traditionally true in some (though not all) cases. I am not sure it should remain true.

>> The fact that the registrant does not have a
>> contract is potentially a double-edged sword for the resource holder
>> but again, imparts no special qualities to the resources themselves.
> 
> I contend that while many legacy resources are not covered under a formal agreement (LRSA or RSA contract) all resources are are covered under an agreement between the resource holder and the registry that assigned them. In the case of many legacy resources this is an informal agreement between the resource holder and a predecessor registry of ARIN, for which ARIN is the legal successor.

The exact nature of said agreement(s) is, however, ill-, if not un-defined.

>> Also, we have been asked several times by the community at large to
>> stop mucking with IPv4 policy. Since there is no one arguing that
>> there is "legacy" IPv6 under any definition, I see no reason to muddle
>> about here, regardless of the definition of "legacy."
> 
> At one level I agree, however, this really isn't new policy, it is properly documenting current practice and dealing with an issue we have created by writing what should be a policy statement into contract.  I believe we have mistakenly crated a situation where parts of policy don't apply to some resources base on a contract terms.  I've recently come to believe this is a problem and dangerous.  I think the solution is to define Legacy Resources in policy clearly and move the policy "Legacy Resources will not be reclaimed solely for lack of use"  into the NRPM, along with other supporting policies that are needed to make it work the way we want it to.

You'll need to shed some light on what makes you think having this exception in contract language applying only to legacy holders that sign the LRSA is dangerous. IMHO, it's preferable to codifying it in policy for legacy holders that have not signed an RSA.

> Technically, if we transition ARIN to a new organization or radically change IPv6 policy we may have a legacy IPv6 situation in the future. So how with deal with legacy resources could matter even for IPv6 in the future.  I hope it is unlikely, but it not impossible.

Huh? How do you figure that? No IPv6 resources are issued without RSA. To me, the defining context for legacy resources is that they are not covered under RSA. I would argue that once resources are under LRSA, though they preserve some legacy protections, they no longer truly have legacy status (which is a good thing).

Owen