THE H-REPORT

Interview with Marshall Rose

by Christy Hudgins-Bonafield

Q: What is your perspective on the organizational problems now occurring with the Internet?

A: There is a general problem in that the cost of entry to participate in the Internet Engineering Task Force (IETF) is very low. You just need to be able to do e-mailıIt allows for the greatest amount of input and diversity. It allows the genius and the moron to participate equally, but of course, the geniuses keep their mouths shut.

Are there problems? Well, in the absence of a technical dictatorship you have an open process, which is always subject to abuse and the more you tighten the rules, the more you invite tyranny. Some aspects of the process are open to abuse by people with commercial or egotistical agendas.

Q: Is a split in IETF and the Internet Society (ISOC) imminent?

A: Itıs hard to say. I donıt think itıs imminent. If the IETF was a corporate entity, theyıd say this is going to happen and everyone would march, but the IETF is 100 generals and two privates.

Q: How does POISED (a process for developing procedures for the adoption of standards and dealing with intellectual-property rights) fit into what is happening?

A: There was a palace revolt three years ago against the Internet Architecture Board (IAB)ıWhat happened is that we instituted a process for nominating and confirming people to IAB and IESG (the IETF Steering Group) and simultaneously started POISED. The POISED effort was supposed to develop rules by which the IETF would opera te. It would formalize how people were nominated and confirmed and formalize some issues which were very important, for example, the relationship of the IETF and ISOC and issues of legal liability with respect to inclusion of intellectual property during the standardization process. POISEDıs charter was to formalize and add a level of written record or legitimacy to the Internet standardization process.

POISED produced a couple of documents which were never formally ratified and a lot of people lost interest about a year and a half ago. Since then, the issue of the relationship has come to the fore again. What tipped the scales was that the ISOC was claiming ownership of the IETF and that the benefit to the IETF was that ISOC would fund the IETF secretariat and provide liability insurance to the people approving the standards-making process.

In both efforts ISOC failed horribly. It was an absolute disaster. It got the wrong kind of insurance. It covered the wrong people in the IETF for the wrong kinds of liability. If you had the three stooges in charge of the Internet, they couldnıt have done worse. Vint [Cerf] wasnıt responsible. He was doing his job as the first president, beating the campaign trail to get sponsors. The day-to-day operations were left to the staff.

I have a lot of problems with ISOCıs claim of ownership. I don't recall granting my consent to have ISOC own the IETF. I was never asked and the same is true of every other IETF and ISOC member. The frustration level got to where a number of people said, ıWe were told there would be benefits to IETF; they never materialized. Enough of this. Goodbye.ı

Now every year, every month, every day, the value of the Internet increasesıand the amount of tension and pressure in and on ISOC and IETF about who controls the process and provides assurances like liability protection is causing a split. The thing that is rather interesting is that in the past, Christian Huitemaıhas been a moderating voice; now it would appear that even hi s patience is not infinite.

I'm not sure what the outcome will be. Part of the problem is that if ISOC doesnıt do it, who does it? How does it get funded? This is a general problem with the decentralized Internet. It has a number of core services, but no one has the right of core ownership. Take the InterNIC. A couple of months ago it changed its policy for how domain names are handed out. From a technical standpoint, it doesnıt own the name spaceıthe community does. They are the stewards of the name space.

I donıt see a white knight on the horizon, but Iıd be happy to be proven wrong.

Q: If the IETF simply went its own way, how would it fund its activities?

A: Funding could be done through membership, say through a surcharge on meeting-attendance fees. This has been discussed for several years as a possibilityıIt might be an adequate short-term solution.

October 15, 1995