Netizens-Digest Saturday, March 31 2001 Volume 01 : Number 371 Netizens Association Discussion List Digest In this issue: Re: [netz] National Academy of Science's new Internet DNS study Re: [netz] National Academy of Science's new Internet DNS study [netz] There is a need for online discussion of new DNS NAS Commmittee Re: [netz] There is a need for online discussion of new DNS NAS Commmittee Re: [netz] There is a need for online discussion of new DNS NAS Commmittee Re: [netz] There is a need for online discussion of new DNS NAS Commmittee Re: [netz] There is a need for online discussion of new DNS NAS Commmittee ---------------------------------------------------------------------- Date: Mon, 19 Mar 2001 23:55:39 +0100 From: Dan Duris Subject: Re: [netz] National Academy of Science's new Internet DNS study rpc> I am working on a paper about a governance model that set the rpc> foundations for the Internet and would welcome comments on a draft rpc> that I hope to have available soon. The ICANN fiasco has shown rpc> that there is a need for a constructive model for Internet governance. rpc> I welcome hearing from those who feel this is a need and who rpc> are willing to try to collaborate toward this goal. Count on me. Send me the draft, when it's ready. dan - ----------------------------- email: dan@netcommodities.com ICQ: 17932727 *- "ye shall not rob from the house i have built" thief1 -* ------------------------------ Date: Mon, 19 Mar 2001 18:12:38 -0500 (EST) From: ronda@panix.com Subject: Re: [netz] National Academy of Science's new Internet DNS study ------------------------------ Date: Fri, 30 Mar 2001 14:12:36 -0500 (EST) From: ronda@panix.com Subject: [netz] There is a need for online discussion of new DNS NAS Commmittee >From article in Telepolis: "National Academy of Science Enters the DNS Controversy New Committee Established to Do Study for US Congress "It is helpful to understand the origins and reason for such a study to be requested of the NAS. The National Academy of Science was created [5] by the US Congress in 1863. The Academy is mandated to "investigate, examine, experiment, and report upon any subject of science or art" when called upon by any branch of the US government. Early work of the Academy included surveying the "uniformity of weights, measures and coins" in connection with the needs of domestic and international commerce, recommending the need for a new scientific government agency, and in a third situation, recommending the establishment of a permanent national forest service to oversee and patrol the public forests. The mandate of the NAS is to study problems involving science and technology. Considering this background and the official role of the NAS to advise the US government on scientific matters, what expertise is needed to fulfill this request by the US Congress? Does the definition of the problem and the composition of the NAS committee demonstrate an understanding of the challenge? " URL: http://www.telepolis.de/english/inhalt/te/7248/1.html - -------------------------- A process that made it possible to build the Internet was a process that learned from what had been created toward determining what was needed for future development. Is the composition of the new DNS NAS committee such that there is the expertise to learn the needed lessons from the current achievement of the Internet as a general purpose interactive and international human computer communications metasystem? Judging by some of the biographies of the provisional members of the committee, there seems little indication that an understanding of the development of the Internet as an expertise was the criteria for the selection of several of the members of the committee. The public has 20 days from the date that the initial appointments were made to make comments on the selection of members. The initial appointments were supposedly made on March 16. This would mean that comments on the provisional appointments would need to be submitted to the NAS by April 4. In such a controversial matter, it seems a serious problem that there is has been so little public discussion or awareness of the creation and composition of this committee thus far. It would be good to see public discussion of the nature of the scope of the committee and its composition before the end of the comment period. This committee is charged with writing a report for the US Congress which will serve as advice for the US policy toward the domain name system development and its future. This is an area of technical development that will impact all users of the Internet, and yet a small group of people, a number of whom seem to have limited knowledge of the nature of the development of the Internet, are being charged with advising the US Congress on the Internet's future development. It is hard to know how the public can have any impact on this process, but one can predict what will be the outcome if the online community continues to be excluded from the activity of the NAS in the formation and workings of this committee. See article in Telepolis URL: http://www.telepolis.de/english/inhalt/te/7248/1.html Ronda ronda@panix.com http://www.columbia.edu/~hauben/netizens http://www.columbia.edu/~rh120/other/birth_internet.txt ------------------------------ Date: Sat, 31 Mar 2001 00:13:57 -0500 From: Philip Busey Subject: Re: [netz] There is a need for online discussion of new DNS NAS Commmittee Dear Ronda, This regards the composition of the Committee on Internet Searching and the Domain Name System: Technical Alternatives and Policy Implications, as listed at http://www4.nas.edu/webcr.nsf/CommitteeDisplay/CSTB-L-99-07-A?OpenDocument I am not quite sure what you see should be represented on the committee that's not already there, which you describe as having, "limited knowledge of the nature of the development of the Internet." They seem to be individuals covering most aspects of law, business, and technology of the Internet. Perhaps there should be some scientists? I would assume that the historical knowledge could be available to the proposed committee members, even if they were not personally there when it happened. Please explain what backgrounds you think should be added to the committee, perhaps even some name suggestions. Phil Philip Busey veld@veld.com http://earthfire.com ------------------------------ Date: Sat, 31 Mar 2001 10:16:44 -0500 From: "Howard C. Berkowitz" Subject: Re: [netz] There is a need for online discussion of new DNS NAS Commmittee I'm also a little unclear about the concerns here. While I don't know all of the individuals on the committee, I do know some, and can see relevant background with others. My general sense is that this committee contains far more relevant experience than did the original ICANN board. Paul Vixie certainly is one of the DNS authorities in the Internet, with experience both in developing the most commonly used DNS software, and in continued operations. I actively participate in the IETF and operational forums such as NANOG and RIPE. While I am a routing, not a DNS specialist, there are recognizable Internet technology experts on the list. There certainly are others who could be added, such as Bill Manning. I would hope, however, there isn't a desire to include people who actively oppose the current structure that, with all its flaws, works. Eugene Kashpureff (sp?), I believe, is not available since he's in jail for hijacking the DNS root. The marketeers at new.net also have taken unilateral actions that jeopardize a working system, in the interest of pure profit. Howard Berkowitz hcb@clark.net ------------------------------ Date: Sat, 31 Mar 2001 13:16:32 -0500 (EST) From: ronda@panix.com Subject: Re: [netz] There is a need for online discussion of new DNS NAS Commmittee Dear Phil Good to hear your questions about my post and the Telepolis article. Philip Busey wrote: >This regards the composition of the Committee on Internet Searching and the >Domain Name System: Technical Alternatives and Policy Implications, as >listed at >http://www4.nas.edu/webcr.nsf/CommitteeDisplay/CSTB-L-99-07-A?OpenDocument >I am not quite sure what you see should be represented on the committee >that's not already there, which you describe as having, "limited knowledge >of the nature of the development of the Internet." When I first heard that there were plans to form the committee I had the naive expectations that it would be a committee of people with expertise in Internet research. > They seem to be individuals covering most aspects of law, business, > and technology of the Internet. This is indeed part of the problem. It is hard to understand why someone with her expertise in the law of "corporate governance" has been appointed to this committte. This is someone who helped to squelch discussion and alternative points of view in the IFWP and then at at least one ICANN related meeting. (IFWP was the International Forum on the White Paper and were the meetings that preceeded the formation of ICANN. I was at the IFWP meeting in Geneva where differences of opinion were squelched and consensus declared when there were profound and important differences among people with what the leadership of the IFWP was pushing thru.) In the ICANN situation, legal and business folks have shown they are incapable of finding a way to solve what is really a technical and social problem with the DNS system. A committee of researchers might ask for advice from legal people, but to have at least 3 on the committee is an indication of a problem. And it has been surprising that the technical community wasn't able take up the challenge of ICANN more directly. So it has seemed that those from the technical community, when in the a situation with legal and business folks are even weaker to take on to identify and find a way to solve the problems of scaling the Internet and its infrastructure. > Perhaps there should be some scientists? Yes precisely :-) This is what I had expected when this was to be a committee set up by the National Academy of Science. I originally expected the commmittee would be made up of those with scientific and research expertise. In my research about the early development of interactive computing and networking, I have seen that there was a need to understand how complex systems operate and how to provide the needed feedback mechanisms for complex systems. The Internet is an amazing achievement. It has battered down walls that one would never have expected to have come down. It has made it possible for people many countries around the world to communicate in a way that is very special. It is important to do the research to understand how this has happened and how to build on it. JCR Licklider, the first head of IPTO, and sometimes called the "grandfather" of the Internet gave talks at conferences on creating self-organizing systems. He explained the importance of having an interactive collaborative research environment to be able to identify the difficult problems that would develop in self-organizing systems (servo mechanisms). The Internet is built as such a system and it needs such a collaborative research environment to support and identify the problems and to find a way to solve them. The problem demonstrated by the DNS system seems to be that the kind of collaborative interactive research environment may be lacking and so there isn't the mechanism to collaboratively investigate and solve the problems when they arise. This new NAS committee, it seems, has been set up to view the Internet as a business product, not as a complex general purpose international interactive human computer communications system. The problems with the DNS can't be solved by those looking at narrow definitions aimed to serve business. The people looking at the problem need to have a broad perspective that is concerned with the whole of the Internet and its millions of users. And the people need an expertise that relates to an understanding of how the Internet was built. There are some on the committee with a technical background who are likely to have some understanding, but there are many on the committee who don't have any such background. > I would assume that the historical knowledge could be available >to the proposed committee members, even if they were not personally >there when it happened. Please explain >what backgrounds you think should be added to the committee, perhaps even >some name suggestions. This historical knowledge doesn't come from having experience. Being involved in the developments themselves is helpful, but one then only has one's own experience to go by. It is important that there be a broader understanding of the development of the Internet. Thus far there has been very little support for such research and documentation. We have tried to do some of this in the chapters of "Netizens" and I have also been working on exploring the creation and importance of the Information Processing Techniques Office at ARPA. The NAS report "Funding a Revolution" showed how even some minimal effort to do such research is very helpful. (The committee there conducted interviews to learn the history.) In doing the research I have done, I have found that the early researchers realized the importance of looking at what had developed as the nuggets and finding a way to support their further development in the next stages of the research. So that one research stage built on the previous research stage. That is some of what is needed for the continued development of the Internet now. But there doesn't seem any awareness on the part of whoever chose the committee members that this is a need for scaling the Internet. For example, in the Internet development I have studied, mailing lists taking on the problems of scaling were critical to figuring out what the problem was and how to solve it. One example is the TCP/IP digest that was created shortly before cutover from the ARPANET protocol NCP to the protocol TCP/IP which would make an Internet possible. The discussion on the list helped to identify what problems people saw had to be understood to be able to make the cuttover. Similarly MsgGroup Mailing list ws formed in 1975 to figure out how to create mail and conferencing programs for a network. There doesn't seem to have been any concern to create the committee in a way that there would be an expertise which would be able to draw on how the problems with Internet scaling have been solved in the past toward determining what is needed now. >Phil Ronda ronda@panix.com http://www.columbia.edu/~hauben/netbook/ http://www.columbia.edu/~rh120/other/birth_internet.txt ------------------------------ Date: Sat, 31 Mar 2001 15:57:22 -0500 From: "Howard C. Berkowitz" Subject: Re: [netz] There is a need for online discussion of new DNS NAS Commmittee At 1:16 PM -0500 3/31/01, ronda@panix.com wrote: >Dear Phil > >Good to hear your questions about my post and the Telepolis article. > >Philip Busey wrote: > > > >When I first heard that there were plans to form the committee >I had the naive expectations that it would be a committee of people >with expertise in Internet research. There's a fundamental problem with the way we use "Internet," and I believe that is confusing this discussion. In my writings, I've started using the more generic term "Internet Protocol Service Provider," of which a subset is involved in the public Internet. Many IP-based services cannot be delivered with production quality in a public environment. The term "Internet" loosely includes: 1. a collaborative environment for networking research. I believe this is closest to your definition, and it is one in which I participate with my IETF and related work. While it is important in the work of many researchers, it is not expected to be of "production quality". It includes various research overlays, such as Mbone and 6bone, which again are not of production quality. It is public, but with controlled access to various experimental services. 2. a distributed environment for sharing information among individuals and organizations. This is quite achievable within the public Internet. Again, it isn't what I'd call production quality. An environment in which people without much network or system administration experience, or the resources required for fault-tolerance, is good for informal sharing. Insisting on production quality would constitute an economic bar to entry. Again, I think this falls into your usage. 3. commercial services characterized by having a known set of public servers, but without the clients/users being predefined. This is the basic model for business to consumer applications. Servers and their access networks need to be of production quality to be reliable. This category blurs into a range of applications where clients can reach servers from arbitrary locations, but the client has to establish a relationship with the provider before application processing can begin. Think of online banking or investing. 4. commercial services with both clients and servers known before communications begin, with assumptions about security (i.e., it is a closed environment) and possibly about quality of service. This is the domain of intranet and extranet virtual private networks. While these services use the Internet Protocol, and indeed may share parts of the same provider infrastructure that do public applications, they are in the domain of "IP Service Providers" rather than "ISPs." > >> They seem to be individuals covering most aspects of law, business, > > and technology of the Internet. I've arranged these four categories in roughly increasing order of how "corporate" they are. Only the first three categories need to participate in the public DNS, although all four use DNS protocols. Much of the DNS controversy centers around categories 2 and 3. Often forgotten in DNS discussions is that DNS was never intended to be a general-purpose directory system for resource location. It was simply a means to resolve human-readable names to IP addresses (and the reverse). These name assignments were intended for the use of system administrators, not for arbitrary users to find machines. Trademark and related intellectual property/brand recognition issues were not at all a consideration in developing the DNS. I'm a specialist in IP routing. One of the realities with which we deal is that the basic design of the global routing system did not consider many of the functions we now demand of it, and it is creaking badly. We are developing a number of short- and medium-term fixes, but the consensus is that the long-term solution will require a new routing architecture. In like manner, there's just so much that can be done to the fundamental DNS architecture to make it meet requirements it was never designed for, corporate governance or no corporate governance. Many of the new commercial requirements will need new architecture and protocol development, which may either overlay the existing DNS or map directly to the IP routing system. Regulators and assorted other governance types have dictated telephone number portability. The reality is that doesn't work very well, and the performance requirements of telephone number portability are vastly less than service location in the Internet -- contrast how often new telephone calls are placed (and numbers looked up) with how often new resources need to be located during a web search. The reality, however, is that funding for many of these services have to come from somewhere, and corporate participation is a reality. It's a classical political situation, where a satisfactory result may look like everyone being equally dissatisfied. > > >In the ICANN situation, legal and business folks have shown they >are incapable of finding a way to solve what is really a technical >and social problem with the DNS system. I don't pretend ICANN is remotely ideal. It's a placeholder at best. I've always been amused that ICANN spends a great deal of effort identifying the role of "stakeholders," although no one seems to have a definition of stakeholder. As far as I can tell, the first use of "stakeholder" in this context was by Ira Magaziner, around 1998, used informally. > >A committee of researchers might ask for advice from legal >people, but to have at least 3 on the committee is an indication >of a problem. But it may be a political necessity. Paul Vixie, for example, is a researcher, but not at all shy. > >And it has been surprising that the technical community wasn't able >take up the challenge of ICANN more directly. Some of the technical community is focusing on more long-term solutions to present to ICANN, rather than arguing about things that won't much change in the short term. > > >In my research about the early development of interactive >computing and networking, I have seen that there was a need to >understand how complex systems operate and how to >provide the needed feedback mechanisms for complex >systems. There is a difference between what you do in production software and network engineering, and in developing experimental self-organizing systems. When I consider the four categories of Internet above, I have to say these kinds of development have to work in parallel. > >The Internet is an amazing achievement. It has battered down >walls that one would never have expected to have come down. > >It has made it possible for people many countries around the >world to communicate in a way that is very special. > >It is important to do the research to understand how this has >happened and how to build on it. > > > >This new NAS committee, it seems, has been set up to view the >Internet as a business product, not as a complex general purpose >international interactive human computer communications system. Perhaps we can agree that it focuses on certain aspects of a broadly defined Internet, and concentrate on helping it define its scope. > >The problems with the DNS can't be solved by those looking >at narrow definitions aimed to serve business. DNS isn't going to become a full business system, no matter what anyone decrees. >The people looking >at the problem need to have a broad perspective that is concerned >with the whole of the Internet and its millions of users. DNS isn't the solution for millions of users. It is system administrator friendly, but not friendly to nontechnical users. > >And the people need an expertise that relates to an understanding >of how the Internet was built. Actually, human factors experience may be more relevant than knowing the evolution of IMPs, fuzzballs, etc., in the early ARPANET and NSFNET. > > >It is important that there be a broader understanding of the development >of the Internet. Thus far there has been very little support for >such research and documentation. We have tried to do some of this >in the chapters of "Netizens" and I have also been working >on exploring the creation and importance of the Information Processing >Techniques Office at ARPA. > >The NAS report "Funding a Revolution" showed how even some minimal >effort to do such research is very helpful. (The committee there >conducted interviews to learn the history.) > >In doing the research I have done, I have found that the early >researchers realized the importance of looking at what had >developed as the nuggets and finding a way to support their >further development in the next stages of the research. > >So that one research stage built on the previous research stage. No argument. But research does not equal production. Production typically uses older, better understood technology, while research continuyes. > > >But there doesn't seem any awareness on the part of whoever chose >the committee members that this is a need for scaling the Internet. I'm rather intimately involved with the routing aspects of Internet scalability, but I don't pretend to be an expert on service location and directory scalability. Presumably, if this is within the committee scope, the members can call on appropriate specialists. > >For example, in the Internet development I have studied, mailing >lists taking on the problems of scaling were critical to figuring >out what the problem was and how to solve it. And NANOG, RIPE, IETF, IRTF, etc., mailing lists are actively dealing with these issues today. > >There doesn't seem to have been any concern to create the committee >in a way that there would be an expertise which would be able to >draw on how the problems with Internet scaling have been solved in >the past toward determining what is needed now. > As a serious question, is the committee the place these issues are going to be addressed in any technical level of detail? I'd see the IETF as the place for this, with the committee more helping deal with funding models. ------------------------------ End of Netizens-Digest V1 #371 ******************************