postalsympos.txt Talk (1) The Internet: its International Origins and Collaborative Vision by Ronda Hauben rh120@columbia.edu (2) graphic I - Introduction I am happy to be part of this symposium on the history of communications sponsored by the National Postal Museum and the Mercurians. My paper is on the International Origins of the Internet. You may wonder why I say "International origins". The common perception is that the Internet is synonymous with the development of the ARPANET. The ARPANET began in the U.S. and was under the ownership and control of the U.S. Department of Defense. The ARPANET was an important development. It is, however, *not* the Internet. Similarly the origins of the Internet are not the same as the origins of the ARPANET. (3) (ARPANET 1968-9) The ARPANET began in the 1968-9 period. The history of the Internet begins in the 1972-73 period. One had to get permission from the US Dept of Defense to be part of the ARPANET. The Internet made it possible to connect by setting up a gateway and adopting the tcp/ip protocol. The ARPANET was a US development. The ARPANET connected different computers and different operating systems, (4) (Internet 1972-3) The history of the Internet begins in the 1972-73 period. The Internet connected different networks. The Internet was an international development. By 1973 there were various packet switching networks either being developed or in the planning stages in countries around the world. (5) For example there is a memo with a diagram which shows three of the early packet switching research networks.. The memo is written by a US researcher. It is dated 1973. The three networks are the ARPANET - USA the UK's NPL and France's CYCLADES (Host) / / / ( ) ( ) ( ) ( ) ( ) ( ) ( ) (Host)---(CYCLADES)----(gateway)---( ARPA )-----(gateway)----( NPL ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) \ \ \ (Host) Each of these networks was under the ownership and control of different political and administrative entities. Consequently, each differs technically in order to meet the needs of the organization or administration that controls it. (6) The problem that pioneering Internet researchers would have to solve can be described as the Multiple Network Problem. The question being raised in this period of the early 1970s is how to interconnect dissimilar packet switching networks? One of the pioneers of packet switching, the UK's Donald Davies, presented a paper in 1974 considering "The Future of Computer Networks". In the paper, he writes: (7) "To achieve...the interconnection of packet switching systems....a group including ARPA, NPL, and CYCLADES is trying out a scheme of interconnection based on a packet transport network with an agreed protocol for message transport.... (Davies, "The Future of Computer Networks", IIASA Conference on Computer Communications Networks, October 21-25, 1974, p. 36) (8) A protocol in computer networking vocabularly is a set of agreements to make communication possible among enties that are different, as, for example, entities who speak different languages. Davies was explaining that there was an effort to make communication possible among these diverse networks. (9) The conference where Davies presented this paper was held at an detente era research institution. It is called the International Institute for Applied Systems Analysis or IIASA. I first learned about IIASA at a conference in Berlin, Germany in October 2001 when I met a Professor who had taught at Humboldt University in the 1970s and 1980s. He told me that Humboldt University had been on the ARPANET. I had never heard of anything like this and so was happy when he asked me if I would like to see some information about this. The next day he brought me a copy of this IIASA publication "Workshop on Data Communications", September 15-19, 1975, The Professor from Humboldt University said he had been at a workshop on computer networking in 1975 at IIASA in, and had been one of the researchers presenting a paper.(a) a)Man/Computer Communication: A Problem of Linking Semantic and Syntactic Information Processing The list of those at this 1975 workshop included researchers from Austria, Belgium, France, the Federal Republic of Germany, the German Democratic Republic. In addition, he brought me a copy of this map. (11) The map shows a computer network connecting sites in the US, Austria, Poland, the USSR, and the IIASA connected by a satellite.(b) (b)http://www.ais.org/~ronda/new.papers/3.pdf The map sparked my curiosity, and demonstrated that there was something important to understand about the efforts to develop a computer network linking the countries of the East and West. I needed to learn more. Also in this 1975 workshop proceedings, I was surprised to find an article by British researchers describing the early development of a British, Norwegian, U.S. research collaboration to make it possible to have the Internet. (12)graphic from Kirstein's paper This disagram is from the 1975 Laxenburg workshop. The diagram shows the University College London (UCL) research network with its connections to the ARPANET, the U.S. packet switching network, and NORSAR the research site for the Norwegian researchers who were part of the international research developing the TCP/IP protocol. (c) Kenney, and Peter Kirstein "The Uses of the ARPA Network via the University College London Node" http://www.ais.org/~ronda/new.papers/2.pdf This diagram, from a workshop at IIASA, just one year after the workshop with paper by Donald Davies that described the effort to create an interconnection between the French, British and U.S. packet switching networks, shows something quite different. The graphic shows an actual networking collaboration to create the tcp/ip protocol. (13) Involved in this research were British researchers at the University College of London, in the UK, American researchers developing the ARPANET, and . Norwegian researchers at NORSAR in Norway, UCL - UK ARPANET - U.S. NORSAR - Norway (NDRE - Norwegian Defense Research Establishment - Forsvarets Forskningsinstitutt) (NORSAR is the Norwgian Seismic Array Facility which is a research installation just outside of Oslo in Kjeller, Norway) The Professor from Humboldt University also workshop remembers discussing the possibility of a connection to the UCL network in Great Britain for Humboldt University researchers. Further research would need to be done, however, to learn whether there was ever any computer network link from Humboldt Univ in the former East Germany to London. (look for longer description significance workshop because showed that links were being possible) ----- My paper focuses on the collaborative research on the development of the TCP/IP protocol which was done by the researchers from the UK, US and Norway. This research centers around the creation of a satellite packet switching network called SATNET. Also involved in this networking research were German and Italizn researchers for shorter periods of time. (14)This is a graphic of SATNET one of the German researchers sent me: Graphic of SATNET This diagram shows the collaboration between German, Italian, US, UK, and Norwegian researchers. There was also collaborative research creating a packet radio network, but I haven't been able to pursue this area of the TCP/IP research yet. Explaining the difficulty of involving different countries in the research process, the Norwegian researcher Paal Spilling writes: The start of the development and experimentation with SATNET was considerably delayed. The idea was to use one 64 kb/s channel... with ground stations in Norway, England, Germany, Italy and the USA. Spilling goes on to describe how there were no regulations that made it possible for the telecoms from the different countries to collaborate in the way being requested. He recalls that the director of the US research agency Bob Kahn, "spent a long time...more than a year -- to have them accept this new mode of operation." To solve the problem of connecting dissimilar packet switching networks, the political problems were essentially turned into the technical problem that was at its essence. (15)The technical problem was how to connect the ARPANET, a terrestial packet switching network, PRNET (a packet radio network) and SATNET, a packet satellite network. These were all different technologies. Tackling the problem of interconnecting dissimilar packet networks was another way of tackling the problem of interconnecting dissimilar packet networks under the control of different political and administrative entities. (16)Looking back at this period of Internet development in the 1970s and then into the 1980s, two of the researchers, French researcher i Louis Pouzin who had led the creation of Cyclades, and Peter Kirstein note the importance of the European Unix community in creating the basis for the Internet collaboration, and for spreading tcp/ip among the European Unix community. Interestingly, the first UNIX (version 5) installation in Europe was at IIASA in January 1975, according to Jim Curry, who headed the Computer Services at IIASA from 1974-1978. Curry also describes how networking researchers at IIASA "were fooling around with various ad-hoc telecommunications linkups (mainly to Eastern Europe) as early as 1975." However, he suggests that email was developed a little later. My research has begun to investigate the role of Europe in the early creation of the Internet. This history is basically not documented and not understood. Both Vint Cerf and Bob Kahn, the inventors of the TCP/IP protocol, acknowledge that the creation of tcp/ip was international from its earliest days. This international collaboration is reflected in the SATNET graphic. The desire for international networking is also reflected in the maps of IIASA networking proposals. These demonstrate that there was the desire for internetworking in the 1970s and the actual research for the technical foundations of the Internet was achieved. This brings me to propose some research questions that I feel would benefit from collaborative investigation. 14) research questions 1- What is the European role in the creation, development and dissemination of TCP/IP and hence of the Internet? This could include: a) SATNET development b) PACKET radio development c) The transfer of computer networking from the ARPANET to SATNET. d) UNIX community collaboration in developing and then disseminating tcp/ip e) The actual networking developments in a spectrum of countries that made it possible to connect to the Internet. 2- Has the desire for European integration had any role in the shaping of the Internet? --------------------- I later learned that this map was documenting a 3 week experiment in computer networking for scientific collaboration between the Russian scientist Gennaudij M. Dobrov, the American scientist Robert N.Randolf and the Austrian scientist W. D. Rauch. They describe the importance of computer networking to make possible needed scientific collaboration and describe a 3 week effprt to explore how researchers from different countries could use the network to participate in an online conference on a common problem. The title of the article was "International Networks for International Team Research" and it was published in 1978 in International Forum Inf Doc., 1978, vol. 3, p.3-13.