Stress of the Search for Profit in a Commodity World

Understanding the Issues Behind IPv6 Transition

Walled gardens versus customer guidance and a comprehensive tour of the issues behind the IPv6 conversion. How to purchase this issue. $350 or $1400 group.

December 1, Ewing, NJ -- The January issue focuses on the difficult problems of the new to convert to IPv6 or suffer a fractured internet.

Executive Summary

Search for Profit in Commodity World

We begin with a look at three business models in the age of commoditization and open source. The first is corral and imprison you customer in a closed, silo-controlled ecosystem. One where many industrial age economies of scale no longer make sense in a world of the internet and mores law and the decentralized capabilities generated.


The second is a path to survival in an age of commoditized hardware and open source software being explored by large IT companies like Cisco and IBM. If you are an IT company by weaning yourself away from proprietary systems you can begin to develop a business model appropriate to the internet age that focuses on educating actual and potential customers on how to solve real problems that leaves government and enterprise IT staff in a state of confusion that comes from trying to integrate generally incompatible systems. We introduce and discuss Cisco's Internet Business Solutions Group in the context of the function of its strategic direction for Cisco and its role in the Connected Urban Development program on which we shall focus during the next couple of months.

Finally we point out a third model where an entity involved in publicly funded research, in this case Canarie, is taking open source tools it has developed over more than a decade and creating an innovative program where it serves in a coordinating role for public and private sector agencies in the funding of broadband solutions. The ability to search across disciplines as canaries has and come up with truly creative solutions depending on cooperation and collaboration will be one of the most important skill sets of the coming years.

The Looming IPv6 Transition

In a detailed interview with John Curran we learn why the global internet has no choice but to undertake a massive conversion of the core and the edge of the internet to IPv6 before allocatable large blocks of v4 addresses are exhausted by 2011.

The problem is primarily organizational and political without a technology solution. The impact is inextricable linked with the impact of v4 exhaustion on the size of the routing tables. This alone will mean that if every does not start to convert now and routable blocks become scare, more routes will be advertise to the core and ISPs will be faced with the choice of telling new customers go away the internet is full or begin or adding routes to the core in such numbers that routers falling over.

If this very difficult problem is not met with action NOW, then in 3 years the Internet is likely to fragment worse than it already has when large ISP are forced to cut connectivity by cutting what they route in order to keep on functioning. The problem is of similar magnitude to Y2K except there is no defined drop dead date and no immediate motivation to begin to invest. This is likely the most important interview I have published in the last five years. A brief excerpt:

COOK Report: And with the current deregulatory environment in this country the small ISP could do nothing?

Curran: It doesn't matter. Even if you had a benevolent dictator who was ready to step in, what are you going to tell someone? The big ISP must prioritize what it keeps in its routing table. They are going to have to drop people. How will you order them to keep the route in their tables when no force on the planet exists that will keep their routers upgraded fast enough?

COOK Report: In other words under this scenario, what is at stake is the continued scalability of the Internet?

Curran: Yes. And to keep running they will start dropping people based on what makes business sense to them. This doesn't bode well for the competitors of the people who have the deepest pockets.

Later Curran: The routing table issue turns out not be a motivator, because by the time you see routing table explosion from fragmentation the IPv4 address space, it's too late to begin your IPv6 transition. [snip] We have to upgrade the entire edge of the internet in three to four years with no immediate motivation for anyone to do it and no one knows that we have to start doing it NOW.


People have to see the decision and avoid its consequences - otherwise they are simply trapped.

Everything is at Risk

COOK Report: But from talking with you there seem to be so many uncertainties as to render a road map impossible?

Curran: The road map is awareness first followed by action.

IP v6 Follow on

An exchange between Paul Wilson and David Meyer

Wilson: What *is* needed is *deployment* on a very large scale, along with all the engineering design, planning and operational management that goes with it. And that first depends on business decisions being made, which are arguably the first and most intractable hurdles that we are seeing...

Meyer: Nice talk, but really, how exactly do you plan to handle the load of dual stack (RIB/FIB), and the dynamics exposed by PI allocations of the granularity we're seeing, or de-aggregation due to either the (further) rise of NAT of the emergence of IPvX markets. And of course, Randy isn't saying anything (in the above) about what's going to happen to the routing system if wide-scale deployment of IPv6 is successful (especially under current RIR "PI for all" policies). Here's a hint: its not going to be pretty; get real used to NAT. As Geoff points out, we're in a really bad place here, and getting to step one of the (need) 12 step program (head out of sand) would seem to be a real good idea (tm) at this point.

Congress and Regulation

Searls: Tell him [Markey] nobody "consumes" the Net. Tell him the Net was designed so anybody can produce and anybody can consume. And so anybody can connect with anybody for as close to zero cost and hassle as possible. And that turning it into yet another telco-like or cableco-like "system" optimized for "billing events" and subscription services is tantamount to killing it.

Comcast Tackles Bit Torrent

Scott McCollough: I still maintain the way to crack this nut is the security/privacy angle. Quit calling it Net Neutrality. Utterance of that phrase leads to immediate cessation of all rational discourse. We have a white paper on privacy and security I'm still honing and summarizing. You can see the current public version at t=6519741393Frankston: My solution is locally owned infrastructure funded as locally as feasible.

I need to preface this by emphasizing that I am taking an opportunistic approach. Once the idea takes hold and we have lots of examples, it's easy but till then we have to find opportunities. I'll give a set of semi-connected answers to your question but if you have more specific questions I'll try to respond more specifically. Paying first. (1) Initially those who benefit like in your house or a clueful city (2) Second clueful cities Interconnect (3) Later general revenues but paying then becomes the wrong term -it's a savings and investment out of general revenue. Spector: What we do need is connectivity and access, unfettered by the ultimate failure producing (to consumers and society) grasps of those who gain only by throttling, garroting, and controlling in order to produce anti-economic scarcity. A transport mechanism open to all comers, competitive and societal, on an equal basis and not monopolized by the transport mechanism's czars is AN, if not THE answer. We need a plan toward that end.

Two Internets?

What do you do if the people paying for the infrastructure don't seem to value the end-to-end constraint? Vest: Actually, I think we answered that back in the 1970-1980s. First we invented a logical/overlay control plane that featured end- points which could choose their own applications, connection methods, etc., and we made those resources available to anyone who could put them to "good" use. As the pool of potential "good users" expanded beyond the research community, the fact that some ends chose to stay isolated was trivial and irrelevant to majority who could reach each other "transparently." Then we decided that the results were so promising that the stuff required to connect one end to another should not be left exclusively in the hands of jealous monopolists who had inherited the one comprehensive facilities platform needed to deploy the new system. So in 1976 we collectively (through FCC 60 2D) mandated that the monopolist would no longer be at liberty to arbitrarily ration point-to-point links. Then in 1984 we collectively (through the federal lawsuits that led to the MFJ) decided that the previous intervention wasn't enough, and forced the monopoly to break into functionally and geographically disjoint facilities segments. Then in 1996 we decided again that the previous intervention was still insufficient, and mandated that arbitrary rationing of the last mile segment would also be forbidden. The fatal flaw of the last intervention was the adoption of language suggesting that it was just temporary, in place just long enough so that that would-be new entrants could build their own substitute phone company platforms. Too bad that this makes no sense and has never happened in practice anywhere/anytime in human history (cable was just another legacy platform designed to deliver a completely different "native" service, which just happened to be compatible with the logical overlay too). I suppose we're lucky that such language wasn't incorporated into the 1976 act -- or else the Internet would have never gotten any further than a slightly larger number of university campuses.

Too bad we forgot how we got here. I guess we're lucky that a small but growing number of countries (Japan, UK, NZ are just the ones I have first-hand experience in) took note and years later followed very similar paths; they'll be out there -- loudly growing, innovating, attracting FDI and talent, and basically eating our lunch -- to act as a constant reminder of the price of forgetfulness.

End-to-end requires both middles and ends; both are necessary, neither is sufficient. Both are in jeopardy today. Fixing one without fixing the won't make any difference, because when it's all over if you're only left with an "e" or a "2" you're not going to be able to do very much with it.

Internet2 and NLR

University CIOs Aren't What They Used to Be

In short we have a convoluted situation where for close to three years many university CIO's who understand only the university financial bean counting side of things have been grumbling in frustration that they are paying money to (the horror) TWO research network organizations. Strangely they seem to see the new seven year agreement that they signed onto with Level 3 for managed services as giving them the equivalent technology that NLR offers its members. They don't seem to rebel at all against the fact that a lightwave via managed services costs on the order three times what a lightwave on NLR costs.

On October 4, NLR sent I2 another memo elaborating on the three points and proposing solutions. The Internet2 Board again turned down any further discussion. Accept the agreement “as is.” For the sad remainder of the maiting dance see Unfortunately the issue has degenerated to a state of internecine warfare among the US R & E community where, on both sides, emotions and not logic are in control.

Meanwhile XO Declines to Renew its IRUs and Goes into Competition with Level 3 And just to shake up the deck -- this year XO Communications jumped clear of Level 3, bought Infinera gear, and lit its own dark fiber over its 18,000 route mile foot print.

Here is the operative paragraph from the Sept 2007 Capacity Magazine: “Just over a year ago, XO Communications' long-haul network was mostly built on wavelength IRUs from Level 3. Despite having its own fibre as a result of the network joint build with Level 3, XO lit up only 30% of its footprint between 2001 and 2006, concentrating investment instead on building out its metro coverage.

Verizon FiOS Plans

November 9th, 2007 COOK Report: Moral of story that follows: if you sign up for online automatic credit card billing be very very careful to keep written notes of what you are doing and especially your user id and password which will NOT be variants of your Normal usual ID and passwords. If you don't do this you will be condemned to a journey through Verizon back office systems hell.

The complete saga is at

Verizon FiOS Saga Continued

November 11th, 2007 by Gordon Cook Last night I wrote to my list Interesting. I just got a phone call at 5 pm on a Saturday from a Velma Stoner. i think the office of customer relations - we are in ivan Seidenberg's office she said. She has a copy of my first mail list complaint forwarded to her by Fred Briggs Exec VP of operations. She wanted to make sure that my issue was resolved. She asked if i would talk to Fred Briggs later in the week and help them figure out what they were doing wrong. I would. She led me to believe that they would give me a human name and contact number from whom i could get help in the future. That idea I like.

Doc Searls My own tale of woe with Verizon FiOS begins here… … and gets worse. Long story short, the company's online sign-up service for FiOS is so wacky and bad that I ended up with two orders that collided, resulting in a cancellation of service that could only be corrected, finally, when a tech support guy in Rhode Island actually gave me his personal number to call back, explaining that he was breaking the rules because, I'm not kidding, he's from Massachusetts. "I'm from Boston. You're in Arlington. I try to look out for people from Mass." I didn't tell him that I had just moved here from California, or that I'm really from New Jersey.

FCC Move Against the Cable Industry

Marcus: What happened in the last few years is, first, the technological and market evolution that puts IP into the core of the network; and second, a series of FCC decisions that specifically exempted IP-based access (which would normally be the area where regulatory oversight is most needed) from the coverage of the statute.Marcus: What happened in the last few years is, first, the technological and market evolution that puts IP into the core of the network; and second, a series of FCC decisions that specifically exempted IP-based access (which would normally be the area where regulatory oversight is most needed) from the coverage of the statute.Marcus: What happened in the last few years is, first, the technological and market evolution that puts IP into the core of the network; and second, a series of FCC decisions that specifically exempted IP-based access (which would normally be the area where regulatory oversight is most needed) from the coverage of the statute.

Suddenly, there is no administrative accountability. The FCC has no statute to follow - it is free to play to the political galleries.

In parallel with this, you have a Bush administration that sought to politicize all of those agencies that historically sought to operate more apolitically. You see very much the same trends at the Department of Justice, and also at the FTC. If you look for example at antitrust - I mean REAL antitrust, not what you sometimes refer to as antitrust - those agencies had well articulated standards by which to judge horizontal mergers. Today, it is pretty clear that they reach a political decision whether to permit a merger, and at most pay lip service to their own guidelines.

Figuring out Where Innovation Comes From

Editor: This was an interesting exercise in what Benkler calls peer production

John Waclawsky: Hi everyone, I am asking for a little help from those with long memories. I am investigating where innovation (surrounding networking) has had its roots over the past 25 to 30 years. To gain some perspective, I am attempting to catalogue innovation and looking at it as parts of two eco-systems: Telco/Cableco innovations and Internet innovations. I am building a chart to show an innovation perspective between these two eco-systems. [Later]

Waclawsky: Ok, I can see some disconnect and I want to limit or eliminate complex interactions and narrow things down to make the main point about economics. I usually define "innovation" as the fruition of an idea in an available product, solution, service or even new business model.

Therefore, one should consider innovation to be more like a process that needs to be complete before declaring success and assuming the label of "innovator". So technological innovations begin life as ideas until they reach the market refined and validated as in a product or service. As you point out, many companies and individuals have the propensity to initiate innovative technological activities (ideas, standards etc.), directed towards delivering a market ready product or service, and then abandoning it, often because there wasn't a market, the idea was too complex to implement ...IMS :-) ..or someone else comes up with a better mousetrap. For example I think all the presentation layer stuff in the past has been replaced by the Web, so they are irrelevant from a money making perspective - no one would use them today, I think :-).

Innovation is about making money (IMHO) so to me this is an important distinction (I believe Tom Vest would agree). Consider some of the past e-mail and other ideas and activities that you point out. You seem to be telling me they have come and gone, mostly being replaced by innovation arising in the Internet. I see people spending a lot of money on ideas and related standards and even though they thought of the idea and did some development, it didn't get to market or eventually maybe some did (which is a point for many to argue) after the Internet eco-system massaged it, opened it or embedded it into other things. I believe that SIP will eventually be in this category. Today people are spending a lot of money on it and it is mostly being ignored in the market and other things like XMPP, IAX etc. are becoming more and more popular.:-)


For the complete issue you must subscribe.



The Stress of the Search for Profit in a Commodity World

Siloed and Walled Garden Brands on the One Hand -- A Movement to Customer Advisory and Education Programs on the Other

Both Tempered by the Capability of Peer Production, and Cross Silo Services p. 1
A Canadian Approach p. 1
A Strategy for Large IT Companies p. 2
Educating Large Enterprise's in Technology Adoption and Strategy p.3
The Scarcity Walled Garden Model p.4
Whither the COOK Report in 2008? p. 5

Understanding the Issues Behind IPv6 Transition

Interview with John Curran Likens Magnitude of Problem to Y2K but without the Known Drop Dead Date

Transition to IPv6 at Core and Edge must Start Now to Keep Internet from Severe Fragmentation as Routable IPv4 Blocs Are Exhausted by 2011 Address Problem: Definition and Vendor Adoption Happened Deployment Did Not p. 6
V4 Allocation Is on Documented Need p. 7
“End of the IPv4 World is Nigher!” p. 8
Trying to Keep IPv4 Running - the Impact on Routes & Peering p. 9
Scarcity of Blocs Means Pressure to Route Smaller Blocks p. 11
Too Many Routes Eventually Kills Routers p. 12
Adding Customers by Dropping Routes p. 13
The Alternative includes Nat-PT p. 14
The Problem - Huge Change Without a Roadmap p. 15
A Problem Without a Technology Solution p. 19
Everything Is at Risk p. 20

V6 Issues - follow on to the John Curran Interview p. 23

Symposium Discussion: October 16 - November 17 2007

Congress and Regulation p. 28

Comcast Tackles Bit Torrent p. 29

Two Internets? p. 33

France Telecom Shares its Ducts - What is Likely Behind the Press Release p. 36
Understanding the “Ultimate Communications
Experience” p. 31

NLR and Internet 2 Merger Denouement

Are the State Networks the Major Prize in the Power Struggle? p. 39
University CIOs Aren't What They Used to Be p. 40
Internet2 Responds November 6th, 2007 p. 40
Some Unsolicited Advice: NLR Should Blow its Own Horn p. 41
Meanwhile XO Declines to Renew its IRUs and Goes into Competition with Level 3 p. 41

My Journey Through Verizon FiOS Back Office Billing Hell

Verizon FiOS Saga Continued p. 43

Scott Marcus Explains FCC Move against Cable Industry

"It effectively eliminates administrative accountability for the agency" p. 45

Figuring Out Where the Innovation Comes From p. 48

Executive Summary p. 55

Symposium & Interview Contributors to this Issue

Affiliation given for purposes of identification - views expressed are those of the contributors alone


Jim Baller, Partner in Baller Herbst law firm & Expert on Municipal Networks
Mike Bookey America at the Internet Crossroads, Principal Pachen Light Consulting
Vint Cerf, Chief Internet Evangelist, Google
David Conrad, Vice President of Research and IANA Strategy, ICANN
Susan Crawford, Asist Prof Cardozo Law school, Board Member ICANN
Bennoit Felton, Senior Analyst Yankee Group Paris and proprietor of the Fiber revolution blog
John Curran, Board Chair ARIN the RIR for North America
Lee Dryburgh, SS7 specialist, Doctoral Candidate University College London
Johannes Ernst, CEO of NetMesh, and authority on OpenID
Bob Frankston developed Visicalc and Lotus and later home networking at Microsoft
David Isenberg, author of the Stupid Network and proprietor of
Scott Marcus is a former FCC Official now working within the Euiropean Regualtory System
Kevin Marks, now with Google, Coauthor Micro Formats, Quick Time Developer
Scott McCollough, Texas Regulatory Attorney
David Meyer, Director Internet Technologies Cisco Systems, and IETF Board
Andrew Odlyzko, Director Digital Technology Center, University of Minnesota
Craig Partridge, Chief Scientist at BBN Technologies and works in the Internetwork Research Department
Ed Pimentel, CTO AgileCo, Alpharetta Georgia
Doc Searls Editor Linux Journal and Doc’s IT Garage blog, ClueTrain Co-author
Mike Spector Attorney and long time friend of Bob Frankston
Bill St Arnaud, Director Ca*Net4, Canada's high speed research network
Tom Vest, Senior Analyst, Internet Economics & Policy CAIDA
John Waclawsky, Chief Software Architect, Motorola
Paul Wilson, Director General of APNIC, the Regional Internet Address Registry (RIR) for the Asia pacific region.