[arin-ppml] Fwd: ARIN-prop-165 Eliminate Needs-Based Justification

Owen DeLong owen at delong.com
Tue Feb 28 06:25:32 EST 2012

On Feb 28, 2012, at 2:23 AM, Matthew Petach wrote:

> On Mon, Feb 27, 2012 at 1:46 PM, Tom Vest <tvest at eyeconomics.com> wrote:
>> Thanks for the responses Matt -- follow-ups inline...
>> On Feb 27, 2012, at 12:40 PM, Matthew Petach wrote:
>>> On Sun, Feb 26, 2012 at 7:02 PM, Tom Vest <tvest at eyeconomics.com> wrote:
>>>> Hi Matt,
>>>> You seem to be arguing the following:
>>>> 1. "IPv4 Speculation" is bad (or maybe good for the lucky speculators, but bad/unfair for everyone else).
>>> I was trying not to assign it a category of "good' or "bad"--simply stating
>>> it does exist today; and thus, one of the arguments *against* a free and
>>> unfettered market is not relevant.  The cry of "but if we allow anyone to
>>> buy and sell IP blocks, we'll be overrun with speculators" rings hollow
>>> when the market is already supporting dozens of speculators who operate
>>> outside the needs-based criteria.
>> But, notwithstanding rhetorical force of habit, is anyone actually crying about "allowing anyone to buy and sell IP blocks" i.e., as opposed to "allowing non-operators (or those who can't demonstrate the equivalent of an operator's "need") to buy IP blocks?" You're still trying to conflate the prerogatives of an(y) aspiring IPv4 seller with those of a(ny) potential IPv4 buyer -- but you really can't claim that the two are the same, or that the differences don't matter without explaining why. Membership in the class of "anyone that can sell IPv4" is bounded by the material fact that in order to sell IPv4 one must actually have some to sell -- and throughout all time to the present moment, 100% of the members of that class have already satisfied, at minimum, whatever definition of "operator" was officially recognized  at the time when they received their address resources (and in most cases, additional/more demanding eligibility requirements that came thereafter). By contrast, the class of "anyone that can buy IPv4" has no equivalent limiting factor.
> I suppose what I'm stumbling over is that today, we
> have large blocks of v4 address space that today
> are held be registrants who are not routing them
> as part of the global IPv4 routing table.  You seem
> to feel that they somehow qualified as "operators"
> in a way that has significance, in spite of them not
> routing their address space in a reachable fashion,
> whereas "speculators" would somehow be categorized
> as "non-operators".

Even under today's policies, there are legitimate purposes for obtaining a registration that do not involve a route appearing in the global table or any routing table you are likely to be able to see.

Just because their network isn't visible to you does not mean that it is not operating and/or is not connected to a network that is connected to the internet.

> Can I ask, what characteristic would you use to
> distinguish an "operator" from a "non-operator"
> that we can externally observe?  I see large
> chunks of address space assigned by IANA
> that are not operationally visible on the v4 internet;
> the CAIDA heat-map survey found similar results:

This question would seem to indicate that you are operating from an assumption that operating==visible in the routing table.

ARIN uses a variety of means to assess whether a network operator is actually operating a network or not. Sometimes that involves receiving the information under mutual NDA, so, it may not be externally observable.

Nonetheless, it is very possible to have a legitimate network operation that requires globally unique addresses that is not visible.

> http://www.iana.org/assignments/ipv4-address-space/ipv4-address-space.xml
> http://www.caida.org/research/traffic-analysis/arin-heatmaps/
> I would argue that your attempt to differentiate between
> "operator" and "non-operator" is specious and meaningless,
> as entities you claim to have qualified as "operators" are in
> fact externally indistinguishable from the category of "non-
> operators" in that none of the IPv4 address space assigned
> to them is routed or visible on the global IPv4 network.
> And if there is no distinguishable difference between
> "operators" and "non-operators", then the class of
> "operators", who managed to obtain IP space thus
> far, in spite of not actually routing it on the global
> internet, is functionally equivalent to the class of
> "anyone that can buy IPv4"--ie, non-operators.

While you may not be able to make the distinction, ARIN staff usually can make said distinction. As such, your assertions are not correct.

> So...again...please help me understand the nature
> of the category of "operator" that allows one to
> obtain globally routable IP space with no plans
> to route it on the global internet, from the category
> of "speculator"?

Speculator, at least for me in the context of this discussion means someone who does not intend to use the addresses to operate an actual network (whether connected to the internet or not) and is merely attempting to profit from price differentials in address trading and/or to manipulate the price of address resources.

> I ask, because the one line I keep hearing over and
> over again in defense of a needs-based assignment
> system is that without it, IP space will be hoarded
> up by speculators.  And if we can't actually tell the
> difference between a "speculator" and an "operator"
> from an external viewpoint, that defense becomes
> meaningless.

We as a community may be unable to tell the difference, but, ARIN staff has proven rather adept at it.

>> Needless to say, you have every right to argue (i.e., give *reasons* why) we should henceforth ignore these real-world material and historical distinctions -- but simply claiming that, suddenly, the differences don't exist is not only unpersuasive, it's not even an argument.
> Ah.  Agreed, the historical differences may have
> existed at one time, or so people claim; not having
> a time machine, we cannot go back to verify such
> claims.
> What I *am* asserting is that at the present moment
> in time, there is no observable characteristic that
> distinguishes them; and thus, at the moment, the
> difference exists only based on a historical record,
> not on any observable difference that can empirically
> be observed.

Observable is a function of the position of the observer and the visibility from that vantage point. I would argue that ARIN staff likely has a position of much greater visibility than you do and that as such, your assertion about what can be seen based entirely on your own perspective is probably not an accurate measure by which to judge overall visibility.

A flee on the back of a dog may be unaware that the dog has a belly because he cannot see it. This does not mean that the dog does not have a belly.

> You read far too much into what I said.
> I'm simply trying to come up with a definition
> of what an IP Speculator might be, as people
> are very concerned that an unfettered market
> would Lead To Rampant IPv4 Speculation--with
> the attendant implication that such a situation
> would be a Bad Thing(tm).
> I'm postulating that one such definition might
> be that an IPv4 Speculator is an entity that
> obtains more globally routable IPv4 address
> space than they actually need for running their
> business.

That is a partial definition. See above for a more complete definition.

A speculator must be someone that obtains more address space than they need.
However, not everyone who does so is necessarily a speculator.

> I then point out that based on such a definition,
> it is clear that there are entities that already
> meet that definition.  For those who are unsure
> of who those entities might be, I again refer
> them to:
> http://www.iana.org/assignments/ipv4-address-space/ipv4-address-space.xml
> http://www.caida.org/research/traffic-analysis/arin-heatmaps/
> from which they can easily see large swaths
> of IPv4 space which has been allocated for
> decades, but is not being used on the global
> IPv4 internet.

First, not being used visibly on the IPv4 internet != not being used.
Second, even if that is true, it is not sufficient to define them as a speculator.

> And, given the existence (as shown) of IPv4 non-operating
> speculators that have existed throughout the entire history
> of this mailing list, it is clearly foolishness to now suddenly
> object to allowing free and unfettered transfer of resources
> on the grounds that it might lead to IPv4 speculation.

Since this depends on the earlier failed assertions, it is, of course, the product of failed logic.

>> Alternately, one could embrace and reconcile both sorts of grievances by proposing a policy that would unilaterally strip all IPv4 possession rights from all past and present IPv4 holders holders without exception, and compel everyone who still wants IPv4 for future use to rebid for that privilege in a neutral open auction.
>> That, at least, would be a coherent/consistent response --  though I doubt many people would call it "fair."
> If you go back through the history of the ARIN debates about
> evaluating number resource usage, that was very close to
> the heart of the policy language that was included in early
> years allowing for annual audits of number resource usage;
> in essence, if you obtained a resource, and it was later found
> you no longer had justifiable need for it, the resource could be
> reclaimed and returned to the pool for others to request.
> At the Eugene meeting, the microphones were occupied
> frequently by advocates on both sides; those who felt that
> once a resource had been granted, it was forever yours, and
> those who felt resources should be re-justified on an ongoing
> basis, or be forfeit back to the pool for others to use.

Technically, anyone with an RSA can be subject to an annual audit and have unused space reclaimed as per section 12.

> The policy language to support that model was proposed and
> discussed, and eventually a variant that allows for reclamation
> was finally agreed on, and is at the heart of every contract
> ARIN members sign:
> "If ARIN determines that the number resources or any other
> Services are not being used in compliance with this Agreement,
> the Policies, or the purposes for which they are intended, ARIN may:
> (i) revoke the number resources; (ii) cease providing the
> Services to Applicant; and/or (iii) terminate this Agreement."

This is also backed by NRPM section 12.

> (from https://www.arin.net/resources/agreements/rsa.pdf )
> ARIN has never proposed doing a massive, all-at-once audit
> and reclamation back from everyone it deems to not be in
> compliance--but that ability is there in the contract, so it's
> not too far-fetched to consider that as a possible scenario.

First, not feasible without significant additional resources.
Second, not at all what was proposed in the alternative solution above.

I would not oppose the audit-based reclamation of which you speak. I would be very opposed to a mass reclamation of all resources requiring all registrants to bid for what they previously had without any indication that they were not properly utilizing it. (which is what was proposed above).

>>>> 3. Therefore, anyone/everyone (today) should have an equal right to be an IPv4 speculator.
>>> Given that there are many players already who hold IPv4 resources that wildly
>>> flaunt any needs-based justification whatsoever, I propose that anyone and
>>> everyone should likewise be able to participate in the market to that same
>>> degree.  If we're going to turn a blind eye towards holders of large blocks
>>> and not require any needs-based justification for them, then be fair and
>>> do away with the needs-based justification for everyone.

Continuing to make this assertion does not make it correct.

>> This is the same straw man argument coupled with the same, completely orthogonal proposed remedy.
>> You really need to come up with a better way to frame this argument than "Someone else can wildly flaunt their indifference, but I can't, and that's not fair"
> So, let me make sure I understand your position on this;
> from your perspective, it is entirely proper to have different
> rules that apply to different entities within the ARIN region.
> There is no requirement to have the rules equally apply to
> everyone?

The rules apply equally to all requestors receiving resources under ARIN stewardship. At best, you can make a claim that some people received resources prior to ARIN's existence that were granted under different policies. In some cases, those entities may not meet current policies. Applying current policies to those organizations may be cost-prohibitive to the organization.

The fact that there are historical recipients who received addresses under different policies is not in my mind different from the fact that recipients in the other RIR regions also receive addresses under different policies. Sometimes policy changes are not practical to apply retroactively. Much like the smog laws do not force 1970s era vehicles to comply with modern emissions standards.

Nonetheless, anyone receiving resources today, whether through transfer or by allocation or assignment from ARIN should have to comply with the same qualification criteria and policies and they should be applied equally. If you know of a situation where that is not happening, you should probably inform ARIN of it so that they can take appropriate action. If you feel that you are not getting a sufficient response from ARIN, I suggest you discuss the matter with the ARIN CEO or any member of the AC or BoT.

> Much like with the shared wireless spectrum, I anticipate
> ongoing adjustments being made to allocations, with the
> entire community having to make concessions and
> adjustments, not just new entrants to the field.  As spectrum
> became more and more in demand, and the push for trying
> to pack more uses into the limited spectrum range increased,
> the adjustments were applied across the board.  It didn't matter
> if you had been using analog TV broadcast frequencies for 50
> years; when the push for reallocation of spectrum came along,
> you had to give up your swath of analog spectrum and move to
> new frequency ranges.  Nobody got to say "but I qualified for this
> spectrum under the rules from 10 years ago, so you can't touch
> it."  Likewise, I think in the IPv4 space a similar situation could
> exist; as the limited resource comes under increasing demand,
> reapportioning across the board could become necessary.  We
> have a nice precedent for it in the analog TV spectrum, so I don't
> think it's an outrageous model to propose at all.

Having watched much of the spectrum issues over the last 20 years (KB6MER, Extra Class FCC Licensee), I have to say that I think taking a similar approach to IP addressing would be one of the worst possible outcomes we could consider.

> I think the analogy I'm looking at is more along the lines
> of someone who has inherited land being allowed to sell
> that land to overseas investors, completely tax-free, whereas
> everyone who did not inherit their property must sell the
> property with government involvement, including restrictions
> on the zoning of the land being sold, and taxes on the transfer
> of the land.

I'm not sure what makes you think that legacy holders are able to transfer their addresses to recipients that don't meet current policy. ARIN will not recognize such a transfer and the recipient in such a transaction is at risk for ARIN reclaiming the resources and issuing them to other parties if ARIN discovers that the original organization no longer exists or is no longer using the addresses.

>>> However,
>>> I'm not advocating that we take up pitchforks and torches and
>>> storm the castles of the wealthly landed gentry; instead, I'm
>>> advocating removal of the unfair and unreasonable rules that
>>> deny anyone else the ability to gain access to similar levels
>>> of resources.
>> Sorry, but I can't see how your proposal adds any additional "fairness" to any element of the system that is not already on the winning side of broader fairness/unfairness divide(s) that are already inherent in this situation, while I can see all sorts of ways that it would be likely further exacerbate those otherwise unrelated problems...
> I'm not actually proposing anything, I'm simply advocating
> for the removal of needs-based allocations.  I've just been
> pulling out more analogies than other supporters have,
> which might make it seem like I'm advocating for more
> than they have.

Yes, it does seem that you are advocating strongly for proposal 165 which is tantamount to proposing something. There's nothing wrong with this at all. But it does limit the veracity of a claim that you are not proposing anything.

>>> The old boy's club has their wealth and power and prestige
>>> in the form of large, unjustified swaths of IPv4 space; let's
>>> bring equality and fairness to the market, and allow
>>> everyone the *opportunity* to participate in that mode.
>> Please stop using the using the term "participate" to erroneously (or misleadingly) conflate the separate and distinct prerogatives and material capabilities of all hypothetical IPv4 buyers and IPv4 sellers. Make a positive argument for why the differences should not matter, and/or why ignoring all of the current differences that you'd like to ignore going forward -- including "operators" vs. "non-operators" -- would make the future better or "more fair" than a future in which those distinctions continue to be honored. Proof by assertion is not strengthened by continued repetition (!).
> I've already indicated up above that the distinction
> between operators and non-operators doesn't exist,
> and cited links to concrete evidence.  I didn't realize
> you needed me to connect the dots for you; I apologize
> for the omission the first time around, and hope that
> now it's clearer that this is not simply proof by assertion.

No, you've shown that the distinction cannot be perceived from your vantage point.
You've also shown that you don't entirely understand the definition of operators.

Everything that follows from that is an assertion based on those two inaccuracies.

>> Thanks again for the detailed response, but no it didn't provide any new information.
>> Heavy and elaborate reinforcement of the same still-unrelated propositions, yes, but the gaping holes where a positive argument connecting reality > problem > solution > reality should be are still there, waiting to be filled...
> Is this helping fill your holes at all?

It's certainly not helping to fill mine. I think i have been pretty blunt as to why and how above.


More information about the ARIN-PPML mailing list