THE H-REPORT

Some comments from Christian Huitema

Huitema is a member of the Internet Architecture Board and its former chairman, as well as a trustee of the Internet Society. He has been tasked by the ISOC with investigating how ISOC can support the technical infrastructure of the Internet

The two interviews you forwarded to me represent the views of individuals who certainly are not known for their benevolence toward the [Internet] Society (ISOC). Their feelings developed both recently, as in the case of Tony [Rutkowski, executive director of the ISOC], and a relatively long time ago, as in the case of Marshall [Rose, leader of the Internet Engineering Task Force (IETF)].

There is an error in fact in TonyÍs presentation of the POISED process. Itstarted after the IAB proposed ñ(OSIÍs) CLNP as the next IPî in June 1992, not before. In fact, the very notion that a POISED process was needed came from Vinton Cerf, then-president of the ISOC and one of the oldest boys in the Internet. This was proposed during the July 1992 meeting of the IETF in Boston, the very meeting where Marshall Rose distributed a provocative T-shirt where ISOC was initialized ñISO (silent) C,î hinting that it was to be as ñinternational,î bureaucratic and authoritarian as the International Organization for Standards (IOS). POISED proceeded nevertheless, and resulted in a redefinition of the IETF process and the roles of the IETF Steering Group (IESG) and the Internet Architecture Board (IAB).

Basically, we went from a self-appointed aristocracy„the old IAB that concentrated all the decision-making power„toward a decentralized democracy. In the old days, officials were nominated by their peers. Now, one half of the members of the IESG and IAB are selected every year by a jury [called] the nomination committee, [which] is drawn from the ranks of the IETF. There is a subtle chain of command whereby IESG nominations are approved by the IAB and IAB nominations by the trustees of ISOC, thus establishing a formal link between the groups.

The work conducted in 1992 is being reviewed now in the light of experience, by a new iteration of the POISED process. We should fix the handling of ñintellectual property rightsî as well as a few minor details.

The relations between the IETF and the ISOC have always been a mixed bag of hope and distrust. This is apparent when half of the participants to an IETF plenary raise their hands when asked whether they are dues-paying members of ISOC. Half of them had at least $35 worth of hope, but only half of them. I am one of those who still have hope, and this is indeed the reason why I placed my name in the ballot list and was elected an ISOC trustee last spring.

The Internet relies on the day-to-day actions of many Internet service providers that operate and interconnect their networks. It also relies on a number of Internet-wide services, some of which cannot be provided on a commercial basis. We need neutral authorities for developing and publishing standards, for allocating shares of the address space to registries and, in fact, for quite a lot of tiny management actions.

Today, we rely on volunteers and U.S. government contracts. But there is a limit to what can be asked of volunteers. They are willing to contribute their work and their ideas, but they donÍt necessarily have to overextend themselves and provide a free ride to millions of users. They need support and assistance from secretariats and insurance. There is also a limit to government contracts, which may be questioned when a budget is reviewed. In my opinion, this should be the role of a neutral and international organization, the ISO C. Neutrality, and moral respect, imply that it should be a not-for-profit membership association.

This [idea] is not new; it was the initial objective of the founders of ISOC. But, clearly, we failed to deliver on our promises during the first three years of operation. Many members of the IETF perceived the ISOC as being more interested in ñestablishing itselfî than providing services. There was a trend to spend the budget on luxury headquarters and large staffs. There was also a trend to ñdo it ourselves.î As part of that trend, the current government contracts would have been terminated or relocated. For example, the IETF secretariat or the IANA would have been run by ISOC instead of the Corporation for National Research Initiatives (CNRI) or ISI. Such a bid for power legitimately angered people like Paul Mockapetris of ISI and Bob Kahn of CNRI, without any perceived benefit for the rank and file of the IETF. The general idea of providing support was good, but the particular way in which it was executed was definitely inadequate.

I believe that the board of trustees, which I joined in May, is well aware of these problems. We decided to focus on our main role, that of servicing the community. Trust can only be re-established through concrete actions. The committee which I chair is currently investigating possible actions. It may be slow in coming out with results, because we are all overworked volunteers. But I expect to publish a plan by mid-October, to be discussed by the community first, (electronically), then during the December 1995 IETF meeting in Dallas, and then by the trustees, who will meet at the same place.

In between, we voted a budget line of $125,000, to be spent in the next six month on actions decided by the chair of the IETF. This is by no means our final word, mostly a placeholder, a concrete token of our good will. I hope that by next year, our contribution will be much more substantial.

Christian Huitema

October 15, 1995