A Practical Navigator for the Internet Economy

Zero Sum Internet Business Models Vie With Internet Cooperative Culture

Use of Lawyers, Charging for Routes, RBOC Moves & Role of Robert Kahn in Context of ISOC & CNRI Surveyed pp. 1-10

Cooperation or conflict? Is the commercial Internet to remain cooperative, best-effort in its service model or will business interests turn it into a zero sum - I win, you lose - affair? If the latter happens, can it remain unregulated? We examine four stories.

First is what we chose to call the litigious business model of AGIS/Net99. This was the original case that caused us to explore the framework of this article. AGIS was started by interests who seem to see the cooperative culture of the Internet as out-of-date. When the recently departed Net99 backbone engineer asked AGIS on a mail list recently about its purported tendency to use lawyers, or threaten their use, in building its Internet business relationships, Peter Kline AGIS's second in command jumped right in with a justification. Kline had written to us: "We understand the culture of the Internet very well. We also understand business very well. Real business is not made of culture, handshakes, verbal agreements, and the heroic deeds of "gurus", it's done with competent staff, lawyers (yah, I know), contracts, and money, things that the business folks behind and in AGIS understand very well."

We reprint the discussion that ensued between Kline and the backbone engineer. Herewith a short sample: Siegel: How long would it take to resolve problems in the Internet if the above [use of contracts and lawyers] were necessary? Kline: Probably less time once the agreements were in place because then there might be a legal requirement to reciprocate. Siegel: How successful would you be at an attempt to sue a major provider for their lack of response to a problem that is costing you business? Kline: We may find out some day soon.

Three of the questions we then asked Kline follow: Peter are you suggesting that the Internet can be better run by putting up litigious walls? What effect would this have beyond the driving up the cost of doing business for everyone? Is there any danger that the price of getting involved in several suits could make you not competitive? We never received any answers.

Our second area of examination is the continuing RBOC attempt to move into the Internet. Access Indiana, rather than respond to our questions about its policy switch to include private business, has removed us from its mail list which it now censors to keep out opinions contrary to the party line. However we obtained the addresses of the list members and have now established our own version of the list which we are in the fourth week of operating. Dissident reports now come to us and we post them to our version of the list. Meanwhile the program finds its effort to fund community networks, for content only, rapidly collapsing as several communet teams have left in dismay and the efforts to recruit customers for Ameritech and Sprint seem not to have been overwhelmingly successful.

Meanwhile, at the request of the Maine Public Advocate, we wrote a set of comments on NYNEX's school/library proposal for it to file before the Maine PUC on September 29. The NYNEX proposal was to take $20 million in over charges and use it for a five year program to fund the attachment of all Maine K-12 schools and public libraries to the Internet at the "princely" speed of 56 kbps. Not surprisingly the NYNEX proposal is telco centric with Internet access ensconced at the center of a frame relay network that NYNEX intends to outsource to IBM. Unlike Indiana, Maine has taken the very positive approach of soliciting a wide selection of public interest advice. The state seems likely to move network intelligence and control to the local schools and libraries. If it is successful, it will have set a path for the rest of the nation to follow.

The third story we explore is Sean Doran's imposition of route filtering for network addresses that are insufficiently aggregated. The technical details are in the full story. We offer here one of our conclusions: We have seen no evidence that Sean Doran has done this for anything other than technical reasons. If only Sprint does it, it can be used as a sales tool against Sprint by the other providers. But, if all the other providers did it, it could become tantamount to a device that would either put small ISPs with non routable addresses out of business or force them to get an aggregatable addresses and go to the time and expense of renumbering that would include some period of down time for them and their customers. And unlike coordinated action to impose settlements, which could be regarded as price fixing, and subject to anti-trust action, a coordinated response on not routing small sets of IP numbers (address blocks with large prefixes) might well be regarded as a "socially responsible action" by the large providers to ensure the continued stability of the Internet.

The fourth story focuses on Robert Kahn and CNRI in the context of the commercialization of the Internet and the pressure for new business models. The IETF standards process is one of the crown jewels of the Internet. There is no doubt of the continued and properly jealously guarded independence of the engineers and their working groups. However, for these processes to function smoothly, the "glue" provided by the IETF Secretariat is quite necessary. For several years the Secretariat has been housed within Robert Kahn's Corporation for National Research Initiatives where it has received about $500,000 a year from the Federal mission agencies through an NSF grant. CNRI has been regarded by the community as an acceptably neutral site for funding and operation.

Earlier this year something rather unusual happened. A Bell Atlantic official and George Edwards, President of the Alliance for Telecommunications and Information Solutions (ATIS) - an RBOC IXC standards group, invited Bob Kahn to talk to the full ATIS Board about the IETF. We were able to get remarks from three key people about what happened at the meeting.

COOK Report: Was there some thought of changing the function of the CNRI based IETF secretariat? Bell Atlantic official: "We were trying to see whether the whole thing should be changed to facilitate dialog or not. Instead of having two separate groups moving off in two separate directions, we felt collectively that it would be to our country's advantage to evolve the next generation of the network in a way that supported TCP/IP type activities and other isochronous traffic will be carried on the Internet." COOK Report: What would changing the Secretariat in a way to support dialog mean? Moving it to telephone company support? Bell Atlantic official: "That's what we talked about, but we didn't do anything of interest to move it that direction."

When we asked Vint Cerf for his point of view he replied: "I came to the meeting with my Internet Society hat on, not my MCI hat. My recollection is that the ATIS group invited Bob Kahn and I suggested that ISOC be represented as well and offered to attend. I only got involved when I learned of the agenda item under discussion (a proposal for ATIS to fund all of the IETF secretariat function). As I recall, Paul Mockepetris also participated in that meeting. My conclusion was that such a move would be ill-advised and probably misunderstood and, so far as I am aware, it did not progress beyond that one discussion."

Bob Kahn, when we asked him, stated: "The only idea on the table that could be attributed to me was the need to establish a broader base of support. I was interested to learn if they would be willing to participate in support of the IETF. The answer seems to have been no, since nothing ever came back to encourage any other view." [We shared the complete draft text of this section with Bob Kahn asking for his comment. He disagreed vigorously with nearly everything we had written. However, when we offered to publish his disagreements as a side bar, he refused to give his permission.]

Certainly telephone industry funding seems inappropriate as the source of support for the IETF secretariat. We are glad that Vint Cerf was quick to recognize this. The problem is that some entity does need to be responsible for this. Recognizing this Vint Cerf said to us: "It seems to me possible for the Internet Society to facilitate the evolution of additional structures and the coordination of discussions leading to their formation, whether the result functions within or outside the current scope of Internet Society activity."

We responded to Vint that we would like to see that happen, adding that we are getting a strong sense that ISOC is as flawed as the CIX in its ability to claim the universal allegiance of members of the Internet community. We don't claim yet to have a full and completely comprehensive picture. But the one we do have portrays resentment of the three charter members (Terena, Educom and **especially CNRI**). We think an absolutely critical issue is establishing a means of taking the profits from domain name registrations and using them for the benefit of the intellectual infrastructure of the Internet, and especially to ensure the continued independence of the IETF and its secretariat. We do not see a way for ISOC to control these new funds without CNRI having unacceptable control over them.

The NSF, in its Domain Name charging agreement with NSI, has created a cash flow that, if wisely used, could support the IETF Secretariat and the most of the remaining parts of the intellectual infrastructure of the Internet. Unfortunately neither ISOC, nor CIX, nor CNRI enjoy solid enough support to serve as the agency for disbursing these funds. Tony Rutkowski is leaving ISOC. We think that where he lands may be a key to some urgently needed redesign of the Internet landscape.


PSI Satisfied With Cooperative Best Effort Internet Business Model, pp. 11-13

Bill Schrader, in an interview with COOK Report, says phone companies want settlements to compensate them for their own failed Internet business plans. PSI is making a fair profit and has no desire to change the rules of the Internet he adds.

Thoughts On Internet Business Models, pp. 14-17

Sean Doran writes about Internet economics and other issues discussed in last month's COOK Report.

National Science Foundation Domain Name Charges, pp. 18-19

Don Mitchell outlines financial implications for Network Solutions and explains NSF rationale behind the actions of the charging decision.

NII: The Dark Side In Washington State Part 3, pp. 20-22

Interview with Al Huff, outgoing Director of WSIPC.