[address-policy-wg] global coordination (was: 2012-01)

On Apr 16, 2012, at 4:12 PM, Scott Leibrand wrote:

> On Apr 15, 2012, at 1:03 PM, Milton L Mueller <mueller at syr.edu> wrote:
> 
>> Three in a row: three consecutive, mutually reinforcing indications of bad faith.
>> 
>> Scott Leibrand: "we shouldn't have a global policy because of 'differences in the different regions' situations...'" In other words (to translate from the original American), "my region has a self-interest in creating trade barriers."
> 
> I think you're confusing me with someone who opposes liberalizing
> (inter-regional) transfer policy. My main accomplishments in the realm
> of ARIN policy have actually been to help get transfer policies
> adopted: first a local one, and more recently an inter-regional
> transfer policy that reduces "trade barriers" between ARIN/APNIC and
> any other regions that choose to participate. To use an economic
> analogy, we first created regional trade areas (with needs based
> "licensing" requirements in some regions, which we've liberalized to a
> 24 month supply in North America). We then expanded that to a
> multilateral "trade agreement" between North America and the Asia
> Pacific region, and also offered the same terms to the rest of the
> world.  (Of course, none of what we're going is actually trade
> negotiation in the traditional sense, as there are no governmental
> trade barriers to be removed, so that is only an analogy.)
> 
>> Yeah, I know that, Scott, that's my point. And your second point seems to boil down to this: those of us currently running RIRs feel more comfortable doing bilateral deals with our buddies in other RIRs. Yeah, I know that, Scott, that's my point. So you don't think there's anything wrong with that? So you don't want to let anyone else in the game? Please reconsider.
> 
> The regional registries register the addresses, so they have to be
> involved if you want transfers recorded. If you want to involve other
> multilateral organizations as well in coming up with a proposed global
> policy, I have no problem with it.  It might in fact allow you to come
> up with language for a global policy that has a better chance of
> getting consensus.  But I still think there are enough regions that
> don't want to participate in inter-regional IPv4 address transfers
> (yet) that the chances of getting such a global policy adopted in all
> regions is low, so I'm focusing my efforts on allowing inter-regional
> transfers between regions with an interest in doing so.  But if you'd
> like input on crafting a global policy, I'd be happy to participate
> electronically, or in person at an upcoming ARIN or LACNIC meeting.  I
> probably won't make it to any other upcoming fora in person, though.
> 
> And, as always, everyone is welcome to participate in the RIRs' policy
> processes, either by subscribing to mailing lists like this one, or
> showing up at public policy meetings.  I for one believe we need more
> input from a broader array of participants.  (It was good to meet you
> at a recent ARIN meeting, and I hope you'll come to future meetings as
> well.)
> 
>> Let's lay this out even clearer for all to see:
>> 
>> Any time someone expresses dissatisfaction with current RIR approaches to address policy, out comes the broken record: "make a policy proposal and put it before the RIRs." And as soon as someone starts to do that, and asks you to get serious about cooperating with them, you come up with a dozen lame excuses to tell them, as Randy colorfully put it, to "foad"
> 
> You are correct that new policy proposals tend to be opposed by some
> segment of the community.  I've attempted to counter that by working
> with proposal authors to take the community's feedback and use it to
> make revisions or craft a new proposal that will engender less
> opposition and more support from the community.  I don't think the
> opposition is a symptom of an insular policy process, so much as a
> manifestation of the conservative approach many network operators and
> other interested members of the IP addressing community take to
> proposed changes to rules/policies.

Hi Scott, 

As someone who has frequently spoken in opposition to the market liberalization initiatives of the last few years, I cannot say whether your assumptions about a widespread "conservative predisposition" toward policy making is right or wrong. I can only speak to my own perspective, which I would describe not as "conservative," but rather as "mindful" (as in the opposite of "forgetful") of two major lessons drawn from operational experience during the 1990s - 2000s: 

1. The last major/industry-wide disruption in network input prices was one of the factors that contributed to the first Internet bust. Subsequently, that cost/price shock also contributed to the creeping "financialization" of some kinds of protocol number resources (e.g., ASNs), which in turn have created new cumulating challenges to Internet scalability and security of a kind that may ultimately turn out to be quite resistant to private, industry-based coordination. 

2. Thanks to the ever-falling cost of other core material inputs (i.e., everything but power and "turf") and the relatively low barriers to entering most Internet content and services markets, the worldwide geographic pattern of Internet infrastructure deployment and competition has been strongly shaped by the opposing commercial strategies of "bottleneck" (i.e., retain customers, deter competitive entry) and "bypass"  (i.e., shop around, find an alternate deployment/delivery path). To date, the rapid pace of Internet deployment/adoption and Internet content and services innovation/diversification is a testament to the overwhelming dominance of the "bypass" strategy over the "bottleneck" strategy -- at least in many/most though not all places. One of the main reasons for that dominance is the fact that aspiring service providers have always been able to leverage different levels of the TCP/IP "stack" to surmount different/shifting problems encountered along the service delivery path and over time, e.g., as they expand out and scale up. In a future of asymmetrical access to globally interoperable IP addresses, however, that flexibility could cease to exist, or more likely become the exclusive purview of the inheritors of RIR-era IPv4, who simply by virtue of that inheritance will be able to "bottleneck" and/or "bypass" any future aspiring post-RIR era new entrant with relative ease. Either way, "bottleneck" strategies are likely to become easier for future incumbents, while "bypass" options for future new entrants will either dwindle or disappear. As time passes, commercial incentives to exploit that strategic advantage for as long as possible (e.g., by implementing IPv6 in ways that, at best, indefinitely perpetuate the advantages of IPv4 possession) are likely to become quite tempting for some commercial operators, just as incentives to commercially exploit various inherited territorial privileges and/or legacy physical assets have been a perpetual feature of the fixed telecoms sector. 

In short, I have a keen appreciation of how incentives matter, and a fair bit of experience observing how similar changes to incentive structures have played out in the not-so-distant past. My own reluctance to see the community race down this path is thus not based on some vague attachment to the status quo for its own sake (ala "conservatism"), but rather on what seems to me to be a fairly concrete and grounded understanding of several potentially serious risks that this path will impose on us all -- current and future direct stakeholders, number resource registry and industry self-governance institutions, and the Internet more generally. 

If the consensus of the community continues to discount these and all other/similar risks, I can live with that; that's how the current, consensus-based system works. But I would like to make sure that folks are actually clear about what it is that they're discounting.

Regards, 

TV

> I usually involve myself mostly in my home region (ARIN), but if
> anyone in the RIPE region would like to work on proposing an
> inter-regional transfer policy along the lines of what ARIN and APNIC
> have adopted, I'd be happy to collaborate.
> 
>> The workshop will happen. There will be a global proposal. You CAN make it fail and you probably will. But in the not-so-long run, that could end up being a major defeat for you, not me and the others you think you are fending off. A lot of people are starting to watch this.
> 
> If a global policy fails, it is because the interested parties who
> choose to participate in the RIR policy processes don't all think it's
> a good idea.  In fact, the way the global policy process is
> constructed, there has to be a consensus in all five regions, so the
> chances of anything controversial getting adopted that way is fairly
> low.  And I'm not sure that's a bad thing.
> 
> -Scott
> 
>>> -----Original Message-----
>>> From: address-policy-wg-bounces at ripe.net [mailto:address-policy-wg-
>>> bounces at ripe.net] On Behalf Of McTim
>>> Sent: Saturday, April 14, 2012 10:27 PM
>>> To: RIPE Address Policy Working Group
>>> Subject: Re: [address-policy-wg] global coordination (was: 2012-01)
>>> 
>>> On Sat, Apr 14, 2012 at 8:10 PM, Milton L Mueller <mueller at syr.edu>
>>> wrote:
>>>> 
>>>>> -----Original Message-----
>>>>> 
>>>>> As has been noted, there already is a successfully working process
>>>>> for submission of global policies in the Internet number registry
>>>>> system In fact, I pointed that fact out specifically to Milton the
>>>>> other day -
>>>>> <http://www.internetgovernance.org/2012/04/06/the-coming-trade-war-in
>>>>> -
>>>>> ip-address-blocks/#comment-1031>
>>>> 
>>>> [Milton L Mueller]
>>>> 
>>>> Yes, John and I have had many conversations about this.
>>>> 
>>>> I will be putting together a workshop at the Baku IGF on precisely
>>>> this topic. Why don't we use that opportunity to make it a real
>>>> WORKshop and attempt to develop consensus around a global policy that
>>>> could be submitted
>>> 
>>> so you want to develop a policy proposal in Forum A which can only be
>>> decided upon in Fora B, C, D, E and F??
>>> 
>>> As John has indicated, it wouldn't be appropriate for RIR staff to
>>> represent their respective community's position since in some cases they
>>> don't know what it would be unless a proposal is put in front of their
>>> community, and it is not the role of the RIR staff to make global
>>> policy.
>>> 
>>> You would have more success (I think) if you used the argument "we
>>> should all run out at the same time" to push for a global transfer
>>> policy.  I don't think an appeal to free market purism will win many
>>> folk, at least that is my sense from my experience as a co-chair of the
>>> AfriNIC PDWG.  We rejected a free market based proposal less than a year
>>> ago.
>>> 
>>> --
>>> Cheers,
>>> 
>>> McTim
>>> "A name indicates what we seek. An address indicates where it is. A
>>> route indicates how we get there."  Jon Postel
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 
>>> following the procedure John has outlined? Are you in John?
>>>> How about a RIPE-NCC representative? Any takers?
>>>> 
>>>> 
>>>> 
>> 
>> 
>