Netizens-Digest Monday, April 2 2001 Volume 01 : Number 374 Netizens Association Discussion List Digest In this issue: Re: [netz] There is a need for online discussion of new DNS NAS Commmittee ---------------------------------------------------------------------- Date: Mon, 2 Apr 2001 14:36:34 -0400 From: "Howard C. Berkowitz" Subject: Re: [netz] There is a need for online discussion of new DNS NAS Commmittee >"Howard C. Berkowitz" writes: > >>Something that is urgently needed, and to which this list potentially >>could make a contribution, is proposing a generally acceptable >>definition, which does need to include research, informal >>collaboration, commercial activities both to the public and closed >>groups, etc. Unless there are some widely accepted definitions, each >>constituency is going to frame their debate with respect to THEIR >>Internet. > >Interesting that you suggest the importance of this. > >I recently made an application to the National Science >Foundation for a study of the Internet as a "mixed system". > >That is a term that Norbert Wiener used in the early 1960s >to describe how there was a need for computer development that >would take into account the relationship between the human and >the machine. > >JCR Licklider in a similar, but even more specific way proposed >that there was a need for research in the desirable relationship >between the human and the computer. He proposed as well, that >the relationship that would be most fruitful would be a symbionic >relationship, a relationship where the human and the computer were >recognized as dissimilar species but were recognized as each >being dependent upon the other. > >Licklider's notion of human-computer symbiosis was the basis for >his joining ARPA and setting up a research office to support the >development of computer technology to make this human-computer >symbiosis a reality. > >My proposal explains more of the backgound of these developments, >but essentially Licklider's vision set the foundations for >the development of first time-sharing, and then the creation >of the ARPANET packet switching network and eventually the Internet. > >Licklider was a neuroscientist who had studied the brain and was >similarly fascinated with computer technology and its potential. > >And he saw the analogies and differences between the human and the >computer. > >I will put my proposal online at > http://www.columbia.edu/~rh120/other/nsfprop.txt > >I didn't get the NSF grant, but I did get a very interesting review >of the proposal which I will also put online at >http://www.columbia.edu/~rh120/other/nsfreview.txt > >The reason I am raising this issue is that one can't just look >today at the Internet and understand or describe its nature. > >To understand the nature of the Internet it is important that >its origins and development be considered and the path of its >development be taken into account and considered. > >Otherwise we are in danger of replacing it with something that >is more like current tv or radio, or an electronic shopping >market, all of which would be a great loss. > >The kind of research I proposed would have helped in this process >but the NSF seems too busy, unfortunately, with funding or encouraging >commercial development of the Internet to recognize the need to >understand its origins in basic research and the need to maintain >a connection with the conceptual framework of that basic research. > > >>>I am not sure of the leaders but I do know that the Corporation for >>>National Research Initiatives (CNRI) has been developing a >>>directory system called the handle system and that they >>>have a RFC with the IETF on it. > >>>I don't know its strengths or weaknesses, but it is the effort >>>to provide a directory system that can help the Internet to >>>scale by providing that functionality. > >>I can't comment on the specific individuals, as I don't know their >>current work. But it sounds as if a focused letter to NAS, >>questioning the lack of directory expertise, is warranted -- copies >>to appropriate political and newsorganizations. > >>You might take a look at > >http://www.handle.net/introduction.html > >I am considering writing a comment on the provisional committee >appointments to submit to the NAS. > >It is hard though to feel it will not just be a waste of effort >as any efforts to make input into their process up to now have >only been a waste. > >What I am realizing is that there are two different models for >Internet development > >One of the models is a "feedback model" where feedback is encouraged >and welcomed to modify the activity so that there will be a way >of reaching the desired goal. This is the model I found operating >for example when tcp/ip digest was created to help with the cutover >from the NCP to the TCP/IP protocols. > >This is the model that I have found for the most part operating on Usenet >over the years that I have been on it, though there are those who >make an effort to squelch opposing viewpoints more recently. > >However, there is another model for Internet development. That model >is the "non-feedback" model. In that model feedback is discouraged >and even if it is offered, it will not function to modify the >the activity. This "non-feedback" model is what has tried to dominate >Internet development since the commercialization, privatization >process was encouraged in the mid 1990's. > >This "non-feedback" model is what ICANN was founded upon. >And in my efforts to contribute to the NAS process, I have >found that they are also acting on this "non-feedback" model >of Internet development. > >I have tried to make input into their process and only find it >has had no affect. > >Thanks for the further information about Bill Mannings research. >It sounds like he would have been a good person to have on the >NAS committee if it was trying to solve a real problem about >the scaling of the Internet. > >>What is Bill's quarterly walkthrough? Is it online? > >>See http:www.isi.edu/~bmanning/in-addr-audit.html (wincing a bit at > >I will take a look as soon as I have the chance. > >At 10:59 PM -0800 3/5/01, Paul A Vixie quoted from a press release: >>New.net (http://www.new.net), a domain >>name registry created to meet the market demand for Web addresses with >>logical, easy-to-remember extensions that make Internet navigation easier, >>today officially released its first 20 new top-level domains. New.net has >>developed a novel, proprietary approach, using software technology deployed > >Is Paul Vixie involved with this "proprietary" web addressing enterprise? A resounding NO. He posted the material to NANOG, and generally was appalled. > >I haven't had a chance yet to look at the web site. > >>New.net acted unilaterally, creating their own set of top-level >>domains. > >It does seem very problemmatic to have a corporate entity acting >unilaterally this way. > >>>I have seen in several places in my research that the original >>>Internet was in fact conceived of as a public utility. > >>I'd have to disagree. Public in the accepted academic and research >>community, but certainly not acceptable to the general public. In >>the seventies, while doing legitimate commercial research (networking >>and/or medicine), I could not get access without explicit sponsorship >>from a government or academic organization. > >This is an example of why the history of the Internet is so important. > >There was *no* Internet in the 1970's. There was an ARPANET, one >big network that one had to get permission to join by contacting >BBN or whoever from the DoD was in charge. > >The Internet, begins with the creation and implementation of >the TCP/IP protocol. (originally called TCP). The cutover to >the TCP/IP protocol was in January 1983 and in the fall of 1983 >the ARPANET was split into an ARPANET (a research network) and >MILNET (an operational network for the DOD) RFC 760 described the first version of IPv4 (the current variant), but that lasted not more than a year because scalability problems had cropped up. Imagine...people wanted to interconnect more than 200 networks! RFC 791 is the principal IPv4 specification, although still pretty primitive. RFC 950 introduced subnetting, but in a way we wouldn't have done without 20/20 hindsight. >These were able to >communicate via TCP/IP because the design of the protocol was >to make it possible for dissimilar networks to be able to function >under different political or administrative authorities and to >create their networks based on the constraints of their own situations, >yet be able to communicate with other networks. TCP/IP isn't essential to interconnecting networks, although it was the first reasonably scalable protocol suite to do so. The original IETF work always considered network interconnection, first called the "catenet" model of "concatenated networks." > >Only after the implementation of TCP/IP could dissimilar networks >connect to the Internet and hence communicate with each other. "Internet" called such, no. But interconnected heterogeneous networks certainly did exist, under the ARPANET umbrella, in 1971 or so. The first equivalent to access servers/routers were the IMPs operated by BBN, and you did indeed have to go through BBN to get connected. Networks were mainframe-centric, but variously used PDP-6/10 TENEX, IBM CP/CMS, GE 645 MULTICS, IBM OS/TSO, etc. These used IMP-specific protocols, not IP. For an excellent, and hysterically funny, history of many of these early efforts, see M.A. Padlipsky's _The Elements of Networking Style_ (Prentice-Hall,1985. ISBN 0-13-268111-0). > >In the 1970's this was impossible. One had to essentially become >part of the BBN controlled ARPANET in the 1970's. It was still the case in the mid-80's. I was the first technical staffer at the Corporation for Open Systems, specifically charged with accelerating interoperability. Admittedly, we were OSI/ISDN, and there were still wars with TCP/IP (I claim the war ended with the typical fraternization and intermarriage). We used TCP/IP, but our mail connectivity was UNIX "bang style," not ARPANET. We never did get into ARPANET, MILNET, etc. even though we were a legitimate not-for-profit research organization. I seem to remember that while we had government members, the military (DCA) and civilian (NIST) members got into a turf war on who would sponsor us. In the late 1970s, the Library of Congress couldn't get connected. > >>I always regarded CIX as parallel to the academic/research structure, >>and complementary to it. There's a very key principle here: from a >>pure technical standpoint, once we got beyond the basic ARPANET and > >NSFNET, there was/is no central Internet core or central Internet >>authority. It's a collaboration: "be liberal in what you accept, be >>conservative in what you send." > >The problem with CIX and other commercial networks is that >with the privatization of th NSFNet in 1995, CIX wasn't a network per se -- it was a single exchange site for traffic that did not meet the NSFNET AUP. >the US govenrment >stopped any protection for those who weren't commercial, >and the ocmmercial networking needs took over and dominated >the networking development. Huh? While I freely admit that many of the active players in the IETF work for commercial organizations, there are still lots of academics (indeed, people go back and forth). Internet II, Abilene, CANARIE, Renater (or whatever the European network calls itself), etc., are all principally academic research networks. There are also a number of research overlays onto ISP-built networks, such as Mbone for multicast and 6Bone for IPv6. > >>>I do indeed consider as one of the great achievements of the Internet, >>>the interactive, particatory process that the Internet encourages >>>and rewards. >> >>>This is something very special. > >>I can only say that I saw the worldwide free information exchange >>take place AFTER there were alternatives to the AUP-based >>communications. In my own experience, when the ARPANET was under >>AUP,. I couldn't join protocol mailing lists. > >I disagree. That is because I have found I guess that Usenet, being >transported in part on the Internet in 1992 was a place where there was >free and open exchange among people from around the world. The >research my co-author of "Netizens: On the History and Impact of >Usenet and the Internet" did in 1992 via Usenet and the Internet >and which he documented in Chapter 1 of Netizens, demonstrates >this open and valuable communication. We get into definition here, but Usenet certainly was not initially deployed over ARPANET or Internet facilities until the Internet was quite privatized. The USENET connectivity I used in the eighties was implemented through bilateral agreements with other organizations participating in netnews. The connectivity mechanism was UNIX dialup UUCP (Unix-to-Unix Copy Program), not Internet NNTP (the network news transfer protocol in the TCP/IP suite). > > >>>And further that this process is the nature of the Internet. >>>Efforts to make the Internet into a network where >>>the user is passive and someone's "customer" will represent a >>>fundamental change in the nature of the network. > >>But can this be a parallel, coexisting change rather than a >>disruptive one? I don't see them as mutually incompatible. > >No it hasn't worked that way. > >The commercial and private entities have no regard for the educational >or research entities. Again, it depends. Networking vendors have discounted a great deal of equipment and consultation to education and research. This isn't necessarily altruistic. Indeed, it often mirrors the very wise strategic decision of Bell Labs to make UNIX widely available to universities, so a generation of UNIX-literate computer scientists developed. >And they also put pressure on government >to support their commercial objectives at the expense of the social objectives. Not challenging you, but what are the social objectives and where is the consensus on them? I recognize that certain cultures/communities have sets of them. > >Compuserve was a commercial network. It didn't develop an Internet. >The public Internet did develop. There is an important lesson to >be learned from this fact with regard to what kind of leadership >is needed to have a socially valuable Internet continue to develop >and scale. > >>Exactly. There's a very close correlation between "customer" as >>described above, and "for profit." Clearly, there are a substantial >>number of people who don't want to participate in the technical >>details, and think of the Internet as an alternative to the telephone >>or television. There is a legitimate desire to use online buying -- > >But if the user is to become passive, the Internet will become >the next "television" *not* an *alternative* to television. Yep. Is there a consensus that being passive is necessarily socially bad? I'm not much of a TV watcher in general. Under normal circumstances, there are a few dramas I'll watch when not in rerun (West Wing, Boston Public, and ER), as well as some PBS cooking and British comedy shows. It is my firm contention no one can be a good network architect without solid Monty Python experience. At the same time, I freely admit that after a 16- to 20-hour day on the road, when I flop down in my hotel room, I can find what I'll generously call trash TV as just the thing for unwinding. Married with Children has its place! > >The tv industry ruined the promise that television held for >social good. Is the only purpose of TV to be of social good? I'm afraid this characterization reminds me unpleasantly of an Orwellian telescreen, a source of propaganda and social conditioning I can't turn off. > >Unless government can be persuaded to act forcefully to prevent >a similar disaster for Internet development, there is the danger >of it going the way television went. > >>Amazon as a good example, but I increasingly use the net to track >>down source for obscure gardening and cooking supplies. There is a >>desire to use the net for entertainment. These desires are not >>illegitimate. > >But all the Amazon.com's in the world don't compensate for losing >access to the kind of cooperative activity and technical cooperation >that the Internet makes possible. Ronda, in my daily life, I see more, not less, opportunity for cooperative technical research, collaboration in lifestyle interests, etc. I really don't see who is cutting me off. > >It is important to determine what is most important and what is >secondary and to have public policy that supports such development. > >> >>See http://www.columbia.edu/~rh120/ >> >>Once you were online. Outside government and academia -- and I'm >>talking about the seventies and eighties, it was quite difficult to >>get online. Even when I worked directly for the Library of Congress, >>we either needed to make substantial investment or get a government >>sponsor to connect. > >I got on the Internet in 1988 via the NSFNET in Michigan. It was >just being built at the time. So there wasn't an Internet really >until the development of the NSFnet which was by the 1990's. >(NSF took over networking development from DARPA in 1987 or so) Than what was I doing in cooperative networking from 1970 or so? Incidentally, there were cooperative networks that were not part of ARPANET/INTERNET/etc. I worked in medical and military research networks to which access was very appropriately controlled. > >So when I got on Usenet via the Internet and the Cleveland Freenet >in 1992 that was actually the early Internet. I don't think you'd find the majority of IETF oldtimers agree that a Freenet is the "classic" Internet. A freenet is one of many ways for interaction. > >And the rush was on by then to commercializa and privatize it, >even those the Free-Nets had begun to spring up and they were >a helpful model to make inexpensive access available to all. > >By 1995 and the privatization, there was a real effort to shut >up those who were challenging that as the road to the development >of the Internet. Specifically, who was trying to shut down freenets, as opposed to offering alternatives to them? > >And from that point on the discussion, at least in the US, has been >muted as to what is the form that Internet development should take. > >This is a public question but it is being treated as a private >perogative. Public? In the US, then that would imply some legislative authority. In the absence thereof, there is a fairly basic assumption that the private sector and market will govern. Don't get me wrong -- I find there is far too much effect of corporate government, but I don't feel all that restricted by it either in research or social contexts. > >But that was because there wasn't yet an Internet. There were >ways of supporting Internet development by the late 1980's >and the early 1990's that would have encouraged the development >of public access. By the mid 1990's you could have had access >to the mailing lists via USENET or via Free-Nets. > > >So the problem wasn't that the Internet was *not* commercial >in the 1980's. How can you say there was no Internet when the IETF document series (RFCs) begins in 1969? Many of the researchers involved worked in for-profit companies. There is NO UNIVERSALLY ACCEPTED DEFINITION OF THE INTERNET. If you want to define when your idea of the Internet started, please give a definition. I have the impression that your definition somehow involves commercial versus noncommercial, but I can't articulate it. > >The problem is that this is a public issue but only a very small >set of interests are encouraged to be part of the discussion. > >That is what happened with the privatization in 1995 and then >with the creation of ICANN and now with the creation of the NAS >committee. Quite frankly, I find ICANN largely to be a bad joke, other than with respect to DNS. > > >>Certainly, some of the newcomers do. Others want the model to evolve >>to something that accomodates both commercial and noncommercial >>interests. Increasingly, in my publications, I make the distinction >>between providing "Internet" service and providing "Internet >>Protocol" service. The former has to exist in the public environment, >>but, while the latter can coexist within the service provider >>environment, is more for intranets and extranets > >To accomodate both takes a protection for the noncommercial, because >the commercial otherwise dominate all. > >Good you are making an effort to recognize that there is something >besides the commercial. And it's not a binary distinction between commercial and noncommercial. There are dimensions of academic and research interests, which may be restricted. The freenet model is yet another. ------------------------------