[ppml] IPv4 "Up For Grabs" proposal

[Owen DeLong]

On Jul 5, 2007, at 1:53 PM, Ted Mittelstaedt wrote:
>> No.  I'm saying that launching a denial of service attack against
>> them is an
>> even worse thing.
>>
>
> And I'm not advocating that.
>
Up for grabs is _EXACTLY_ that whether you realize it or not.

>>
>>>
>>> Oh I forgot, you were advocating that
>>> legacy holders don't even have to sign an RSA to get IPv6.
>>>
>> Please show me one place where I have advocated such a thing.
>> I have never advocated that and your accusations here are baseless
>> and misleading at best.
>
> I already e-mailed you this off list, maybe you didn't read it:
>
> In your proposal:
>
> "...	1.	If the organization does not currently pay ARIN
> 				fees, they shall remain fee exempt...."
>
> You didn't say "fee exempt for a certain amount of time", nor did you
> say "does not currently pay ARIN IPv4 fees they shall remain
> fee exempt for IPv4 only" both of which would have closed a huge hole.
>
Nor did section 1 discuss IPv6 address space.

Read section 4. which is the only part of the proposal that speaks of
anything related to IPv6 space being issued...

> 			4.	All organizations returning space under this
> 				policy shall, if they meet other eligibility
> 				requirements and so request, obtain an
> 				appropriate IPv6 end-user assignment
> 				or ISP allocation as applicable, with no fees
> 				for the first 5 years.  Organizations electing
> 				to receive IPv6 allocation/assignment under
> 				this provision must sign a current RSA and
> 				must agree that all of their IPv4 resources are
> 				henceforth subject to the RSA. Organizations
> 				taking this election shall be subject to end-user
> 				fees for their IPv4 resources not previously
> 				under an ARIN RSA.  If they are already an
> 				ARIN subscriber, then IPv4 resources
> 				affected by this process may, instead,  be added to
> 				their existing subscriber agreement at the
> 				address holder's discretion.

Note that it very specifically requires them not only to sign an RSA  
for their
IPv6 resources, but, also, in order to get IPv6 for free for 5 years,  
they _MUST_
bring all IPv4 resources under RSA _AND_ start paying appropriate fees
for them.

>>
>> Hardly.  I just want to recognize what is and isn't possible and  
>> do the
>> best we can with the situation we have.  There really is little  
>> point in
>> wasting the monumental amount of effort and capital that it would  
>> take
>> to (probably fail in the) attempt to revoke legacy IPv4 resources.
>
> How does the RIR's ceasing to track IPv4 that isn't under an RSA  
> with them
> post-IPv4 runout constitute a revoking of IPv4?
>
Removing then from the registry isn't revocation, but, it's also not  
helpful
and there's really no point to doing it.

>> By the time all the court battles were done, the reclamation of
>> legacy IPv4 resources would probably not be of substantial benefit
>> to the community.
>>
>
> This isn't about reclamation.  This is about getting people that
> aren't paying IPv4 fees to an RIR, out of the tracking system once
> IPv4 runout has happened and a significant number of orgs have
> switched to IPv6.

What's the point of doing that?

> Specifically, my suggestion wouldn't even take place until IPv4
> was effectively useless for new assignments - even if it was
> available.
>
I'm not convinced this assertion is accurate.

>> Marking the addresses as "up for grabs" and having a policy  
>> discussion
>> on record describing "up for grabs" the way you already have would
>> certainly hold up as "encouraging".
>>
>
> Except that this isn't a policy discussion since no policy has been
> proposed and your not even discussing the items in the post anyway.
>
Sorry... If this isn't a policy discussion, it doesn't belong on this  
mailing list.
This list is for the purpose of discussing and developing ARIN policies.
Whether the policy has been proposed or not, this _IS_ a policy  
discussion.
You can have a policy discussion without a formal policy proposal.

>> Actually, besides the RIRs there are a number of other resources that
>> track these, including, but, not limited to completewhois.
>>
>
> So go join the fringe at http://www.opennic.unrated.net/ and quit
> bothering the rest of us.  How many ISP's do they have now?  7? 8?

First, I would hardly put www.completewhois.net in the same bin
with opennic.unrated.net.  Second, I have a long history of constructive
participation in the ARIN policy process including a number of proposals
I have either written or participated in the development of which are  
now
policy.
> If you really thought that someone else tracking these was a serious
> problem you wouldn't be using that as an argument, because since an
> alterantive would be available, it would make what the RIR's did a
> non-issue.  The only reason your bringing it up is because you know
> it's a bogus argument.
>
I'm bringing it up because it further demonstrates the extent to which
your proposal accomplishes nothing positive and all of its limited
potential effects are just disruptive.
>>
>> The RIRs are not governing bodies.  It amazes me that you are so
>> thoroughly
>> ignorant of the law.  The governing bodies that govern what is or is
>> not allowed
>> on the highways are actual GOVERNMENTs.  The RIRs are NOT GOVERNMENT.
>
> Yup - and so, what requirement does an RIR have to continue to record
> a legacy assignment?  They have no contract and as you point out they
> aren't a government, so why do they have to keep doing it?
>
I believe they made an agreement to do so with IANA as part of the  
process
of their formation. Other than that, I suppose, perhaps, they don't  
need to,
however, there's also no gain to anyone for them to stop doing so.


>> Further, you can take a horse and buggy onto an interstate highway.
>> While you
>> can't do it in most places, your blanket assertion does  prove false
>> if you look
>> at Pennsylvania, parts of Ohio, and a number of other rural areas
>> where the
>> Interstate replaced earlier roads and would render places
>> inaccessible to
>> horse and buggy if they could not traverse said interstate.
>
> http://www.commonsensei69.org/damage.htm
>
> "...Because the Amish travel only by horse-and-buggy, they cannot
> travel on or across interstate highways..."
>
That's only between two cities within Indiana which is not one of the
states I mentioned.  I know that the Amish are allowed to operate
horse and buggy on Pennsylvania interstates, and, I know that
in some parts of Ohio this is common practice and the LE folks
seem to ignore it even if it is against the law.

Unfortunately, neither Ohio nor Pennsylvania put their laws on the
web in such a way as to make it easy for me to give you the exact
statutes (or even to read them myself).

>
>>> Fundamentally I am saying let's make that policy right now.  If you
>>> believe
>>> that 20% IPv6 adoption isn't sufficient enough to call IPv4
>>> obsolete, then
>>> what about 40% IPv6 adoption?  If that's not enough, what about 60%
>>> adoption?
>>>
>> What does it matter?  What's the point of calling IPv4 obsolete at
>> the RIR
>> level?  When ISPs start derouting it, it will be obsolete for any
>> meaningful
>> definition.
>>
>
> That is nothing more than setting up circular logic.  As long as the
> RIR's track IPv4, those assignments are official, and anyone who has
> one can claim that an ISP cannot deroute it.  (you claim they can file
> a court injunctions over this kind of thing, well there you go)  So  
> the
> ISP's aren't going to deroute them.
>
Nobody can claim an ISP can't deroute an IPv4 address today.  Lots of
IPSs deroute lots of addresses every day now.  No ISP has to carry your
IPv4 route unless you or someone else has a contract with them that
says they will.

> Your saying the RIR's aren't going to consider the assignments  
> obsolete
> until the ISP's start derouting them.  Until they consider them
> obsolete they will still track them.
>
Pretty much.

> So in summary, the RIR's will never stop tracking them and the ISPs
> will never stop routing them.  So explain why again that a legacy
> holder who pays no fees to an RIR for IPv4 would choose to go to IPv6
> and start paying fees?
>
I don't accept your premise.  I firmly believe that ISPs will begin
charging more and more for IPv4 connectivity and eventually will
terminate IPv4 services on an ISP by ISP basis.  I believe that
when there is no longer a critical mass of IPv4 connectivity, IPv4
will rapidly fall into disuse on the public internet and that at that
time, the RIRs can put obsolescence policies in place to sunset
the tracking of IPv4 registration data.

An IPv4 legacy holder who wants to talk to the rest of the internet
will move to IPv6 because he will have to in order to talk to the
rest of the internet.  It will be the ISPs that provide this forcing
function, however, and not the RIRs.

>>>
>> Doesn't matter.  Eventually, the legacy holders won't be able to get
>> an ISP
>> to route their IPv4 addresses.
>
> Then why are you so opposed to setting a date in advance that we will
> all say this is going to happen?  If it doesen't matter, then why  
> argue
> against this?
>
Because the date should be decided on a case-by-case basis between
the ISP and the address holder, not by some RIR policy decision without
any visibility into the real world of what is happening.  Because there
is no benefit to doing so, only cost.

> Sounds to me like the boyfriend objecting to his girlfriend visiting
> churches and looking at wedding packages, while at the same time  
> insisting
> that he's going to marry her... eventually.
>
Only if we accept all of your broken assertions.  I don't.

Owen