The Social Forces Behind the Development of Usenet News by Michael Hauben hauben@columbia.edu Right at this moment someplace in the world, someone is being helpful (or someone is being helped.) At the same time, others are participating in various discussions and debates. A new communications medium is currently in its infancy. Over the past two decades the global computer telecommunications network has been developing. One element of this network is called Usenet News (also known as NetNews), and this news' original carrier was called UUCPnet (or just UUCP). The rawest principle of Usenet News is its importance. In its simplest form, Usenet News represents democracy. The basic element of Usenet News is a post. Each individual post consists of a unique contribution from some user placed in a subject area, called a newsgroup. In Usenet's very beginning (and still to some extent today) posts were trans- ferred using UNIX's UUCP utility. This utility allows the use of phone lines to transmit computer data among separate computers. The network (UUCPnet) that Usenet News was transferred on, grew from the ground up in a grassroots manner. Originally, there was no official structure. What began as two or three sites on the network in 1979 expanded to 15 in 1980. From 150 in 1981 to 400 in 1982. The very nature of Usenet is communication. Usenet News greatly facilitates inter-human communication among a large group of users. Inherent in most mass media is central control of content. Many people are influenced by the decisions of a few. Television programming, for example, is controlled by a small group of people compared to the size of the audience. In this way, the audience has very little choice over what is emphasized by most mass media. However, Usenet News is controlled by its audience. Usenet News should be seen as a promising successor to other people's presses, such as The Searchlight, The Appeal to Reason, The Jewish Daily Forward in the U.S. and the Penny Press tradition in England. Like these other people's presses, most of the material written to Usenet is by the same people who actively read Usenet. Thus, the audience of Usenet decides the content and subject matter to be thought about, presented and debated. The ideas that exist on Usenet come from the mass of people who participate in it. In this way, Usenet is an uncensored forum for debate - where many sides of an issue come into view. Instead of being force-fed by an uncontrollable source of information, people set the tone and emphasis on Usenet. People control what happens on Usenet. In this rare situation, issues and concerns that are of interest and thus important to the participants, are brought up. In the tradition of Amateur Radio and Citizen's Band Radio, Usenet News is the product of the users' ideas and will. Unlike Amateur Radio and CB, however, Usenet is owned and controlled solely by the participants. Currently the range of connectivity is international and quickly expanding around the world into every nook and cranny. This explosive expansion allows growing communication with people around the world. In the 1960s, the Advanced Research Projects Agency (ARPA) of the Department of Defense began research of fundamental importance to the development and testing of computer communications networks. ARPA research laid the ground work for the development of other networks such as UUCPnet. ARPA conducted an experiment in attempting to connect incompatible mainframe computers.(1) It was called the ARPA Computer Network (Arpanet). ARPA's stated objectives were: "1) To develop techniques and obtain experience on inter-connecting computers in such a way that a very broad class of interactions were possible and 2) To improve and increase computer research productivity through resource sharing."(2) ARPA was both conducting communications research and trying to study how to conserve funds by avoiding duplication of computer resources.(3) A Cambridge, Mass. company, Bolt Beranek and Newman, Inc. (BBN), was chosen to construct the network, and AT&T was chosen to provide the communications lines. Arpanet was needed because it was found that a data connection over existing telephone voice lines was too slow and not reliable enough in order to have a useful connection.(4) Packet-switching was developed for use as the protocol of exchanging information over the lines. Packet-switching is a communications process in which all messages are broken up into equal size packets which are transmitted interspersed and then re-assembled. In this way, short, medium and long messages get transferred with minimum delay.(5) The Arpanet was a success. ARPA provided several advances to communi- cations research. Arpanet researchers were surprised at the enthusiastic adoption of electronic mail (e-mail) as the primary source of communication early on. E-mail was a source of major productivity increase through the use of the Arpanet.(6) By 1983, the Arpanet officially shifted from using NCP (Network Control Program) to TCP/IP (Transmission Control Protocol / Internet Protocol.) A key point to TCP/IP's success is in its simplicity. It is very easy to implement over various platforms, and this simplicity has accounted for its continued existence as a de facto standard of the Internet up to today. Arpanet's lasting contribution was demonstrating how a backbone infra- structure can serve as a connection between gateways. A gateway is a computer or part of a computer programmed to receive messages from one network and transfer them onto another network. Arpanet grew quickly to more than 50 nodes between Hawaii and Norway.(7) However, it did not extend to all who could utilize it. Computer scientists at universities without Department of Defense contracts noticed the advan- tages and petitioned the National Science Foundation (NSF) for similar connectivity. CSNET was formed to service computer scientists. CSNET was initially financed by the NSF. Very quickly the desire for interconnection spread to other members of the university community and CSNET grew to serve more scientists than just computer scientists at universities. CSNET became known as "Computer 'and' Science Network" rather than just "Computer Science network."(8) Arpanet was phased out by the Department of Defense, and was replaced by various internal networks (such as Milnet). The role of connecting university communities and regional networks was taken over by an NSF funded NSFNET, which originated as a connection for university researchers to the five National Supercomputer Centers. CSNET and NSFNET were made possible by the research on Arpanet. The NSFNet became the U.S. backbone for the global network now known as the Internet. Arpanet research was pioneering for communications research.(9) Researchers discovered the link between computer inter-connection and increased productivity from human communication. The sharing of resources was proven to save money and increase computer use and productivity. The development of packet-switching revolutionized the basic methodology of connecting computers. The source of these discoveries were the people involved. The personnel involved in the Arpanet project were very intelligent and forward- looking. They recognized their position of developing future technologies, and thus did not develop products that commercial industry could (and would) develop. Instead they understood that the communications technologies they were developing had to come from a not-for-profit body. ARPA researchers had no proprietary products to support, and no deadlines to meet. Either would have tainted, or made developing networks of incompatible computers impossible or limited. Current users of international computer networks are in debt to the pioneers of Arpanet. So Arpanet was successful in its attempt to connect various spatially remote computers, and thus more importantly the people who used those computers. However, these people were either professors at Universities that had Department of Defense research grants or employees of a limited number of Defense Industry companies. Eventually other Universities connected through CSNet, NSFNET, BITNET and other developing connections. There were still a mass of people who wanted a connection, but were not in a position to gain one. Duke University and the University of North Carolina at Chapel Hill were two such locations. It was in these underprivileged fertile grounds where the grassroots computer communica-tions breakthrough of Usenet origina-ted and developed. The UNIX operating system provides the basic tools needed to share in- formation between computers. UNIX(10) was developed as "a system around which a fellowship would form."(11) One of the programmers of UNIX, Dennis Ritchie, wrote that the intended purpose of UNIX was to "encourage close communication."(12) UNIX's general principles thus conceptually foreshadowed the basic tenet of Usenet News. How else should one go about designing communications programs, but on an operating system which was designed with a basic principle of encouraging communication? The UNIX utility UUCP (UNIX-to-UNIX CoPy) was developed in 1976 by Mike Lesk at Bell Labs. UUCP provided a simple way of passing files between any two computers running UNIX and UUCP. UNIX's popularity also arose from AT&T's prohibition to profit from other than their main business, phone services, under the terms of the 1956 Consent Decree. UNIX was thus available on a "no-cost" (or very low cost) basis. The operating system was seen as an "in-house" tool on DEC computers and was in use throughout Bell Labs. Many Universities used the same type of computer and were licensed by AT&T to utilize UNIX. It was thus easily accessible. Schools picked it up, and computer science students used it to learn about operating systems, as UNIX was a model of elegance and simplicity compared to most operating systems of the time. UNIX became a widely used operating system in the academic world. This paved the way for an international public communications system to form. Usenet News was created by graduate students Tom Truscott and James Ellis of Duke University in conjunction with graduate student Steve Bellovin of the University of North Carolina in 1979. A 5 page leaflet introducing Usenet News was distributed at the Winter 1980 Usenix UNIX Users' Conference in Boulder, CO. Later that year, at the Summer Usenix Conference in Delaware the software needed to participate in Usenet was put on the Conference tape. By this time, Stephen Daniel had rewritten the basic programs and it was called A-News. The software was immensely popular. Usenet was patterned to mean "UNIX Users Network." The developers thought Usenet would be used to discuss people's problems and to share experiences about UNIX. Usenet did provide a forum for people to solve problems with UNIX, as AT&T provided no support for UNIX. In an early handout, Usenet is referred to as a "poor man's Arpanet."(13) Stephen Daniel told me that people who didn't have access to the Arpanet were hungry for similar opportunities to communicate.(14) Usenet News has been full of surprises from the beginning. The originators of Usenet News underestimated the hunger of the people. As the initial intentions were to produce an easy method of communicating with other users at the same site, the writers thought people would want to have local bulletin boards.(15) However, people were attracted by the possibility of communicating with others outside the local community. Even today, the wide- spread communication is part of what makes Usenet so enticing. It was also thought NetNews would be useful as a method of communications at individual locations, and between sites close to each other.(16) Usenet grew as a grass- roots connection of people. The people who utilized NetNews wanted to communicate, and communicate they did! People have a fundamental need to communicate and Usenet News aptly fills the bill. (See, e.g., Gregory G. Woodbury's "Net Cultural Assumptions") Early in 1980 or 1981 the gap between Arpanet and Usenet was bridged.(17) The University of California at Berkeley had connections to both Arpanet and Usenet News. This allowed another pioneer, Mark Horton, to bring discussions from Arpanet mailing lists into Usenet newsgroups.(18) This was a significant achievement. Communities other than ARPA sponsored researchers were finally able to see what the Arpanet had made possible. The gatewaying of Arpanet mailing lists into Usenet attracted a wave of people. These people became attracted to Usenet News when two Arpanet mailing lists (SF-LOVERS and HUMAN-NETS) began to appear on Usenet.(19) These lists provided interesting material and discussions. The size of the news feed (i.e., the raw data of Usenet News) thus became larger and provided more for people to read. Later other sites would serve as gateways to even more discussion lists from the Arpanet. NetNews was also seen as a superior method of holding discussions. Gatewaying these fa (i.e., From Arpanet) newsgroups proved to be politically courageous. The Arpanet was only accessible by a certain group of people, and these gateways challenged that notion. The effect on the Arpanet was important as Steve Bellovin wrote: "The impact of Usenet on the Arpanet was more as a (strong) catalyst to force re-examination (and benign neglect) on the strict policies against interconnection. Uucp mail into the Arpanet became a major force long before it was legit. And it was obviously known to, and ignored by, many of the Powers that Were."(20) The network made possible by UUCP expanded to connect people across the entire country. Rather early UUCP expanded internationally when the University of Toronto Zoology Department joined the Net in May of 1981.(21) Two companies proved helpful to this communication by distributing NetNews and electronic mail long distance. Each UUCP site had to either pay the phone bill to connect to the next system, or arrange for the other system to make the phone call. System Administrators at AT&T and DEC did the footwork in order to take e-mail and news where it might not have reached. These people went through the trouble in order to try to see the system work. However, easy connections were not always available. In one example, Case Western Reserve University graduate students had to route mail across the continent twice in order to send mail through UUCP to reach their professors who were connected to the Arpanet next door.(22) Usenet News seems to have introduced the idea of connectivity to the Arpanet, as gradually the Arpanet connected to other networks until it became more known as a backbone to other networks than a self-contained network.(23) Voluntary effort is the crucial foundation of UUCPnet and Usenet News. On one side, there are those who donate time and energy by contributing to Usenet's content - writing messages and answering messages or participating in a debate. Without the time and effort put in by the users of Usenet News, Usenet News would not be what it is today. Also important to Usenet's success are the system administrators who make the running of Usenet News possible. Resource-wise, NetNews takes up disk space on computers throughout the Usenet, and phone calls often must be made to transfer the raw data of the news. In particular, system administrators at AT&T and DEC found it worthwhile to transport the News across the country. Certain sites emerged as clearing houses for Usenet News and UUCP e-mail.(24) These machines served as major relay stations of both news and e-mail. A structure grew that was considered the "backbone" of "the net." Backbone sites formed the trunk of the circulatory system of news and e-mail. A backbone site would connect to other central distribution computers and to numerous smaller sites. These central backbone sites provided a crucial organization to the Usenet communications skeleton. People formed the center of these connections. For example, ihnp4 at AT&T existed mainly because of Gary Murakami's effort and only partially from management support. Usenet services and support were not officially part of Gary's job description. After Gary left ihnp4, Doug Price put time and effort to keep things running smoothly. Certain System Administrators in Universities also picked up the responsibility for distributing News and e-mail widely. Often these individuals would find ways of having their site pick up the phone bill. Sometimes sites would bill the recipients. However, others who received a free-connection often exchanged that for spreading what they received to others for no charge (e.g.; Greg Woodbury & wolves off of Duke, and plenty of others.) Initially, expansion of sites receiving Usenet News was slow. Some statistics are shown in the table. Year # of Sites Articles/day 1979 3 2 1980 15 10 1981 150 20 1982 400 50* 1983 600 120 1984 900 225 1985 1300 375 1MB+/day 1986 2500 500 2MB+/day 1987 5000 1000 2.5MB+/day 1988 11000 1800 4MB+/day *This was after Arpanet mailing lists were gatewayed into Usenet. (Gene Spafford, Usenet History Archives from the Mailing List) [from Gene Spafford, Oct. 11, 1990, based on presentation on Oct 1, 1988 for the IETF meeting.] Why did this happen? Initially Usenet was only transported via UUCP connections. Besides UUCP, other resources were used, such as weekly airmailing of mag-tape data to Australia to provide connectivity.(25) Today, Usenet News travels over all types of connections. The evolving Arpanet (and now the Internet) provided a faster way of transporting news. However, a large number of Usenet News recipients only have connectivity via UUCP. Universities and certain businesses can afford to connect to the Internet, but many individuals also want a connection. Today 60% of Usenet traffic is carried over the Internet via the instantaneous Network News Transport Protocol (NNTP), but 40% of Usenet News is still carried through the slower UUCP connections. From my own research using Usenet News, I have heard of several examples of various types of connections using UUCP. These representatives of the "fringe" give a clue to what the origins of this communication must have been like. The number of sites receiving Usenet News continually increased (as already illustrated) and this clearly demonstrates its popularity. People were attracted to Usenet News because of what it made possible. People want to communicate and enjoy the thrill of finding others across the country (or today across the world) who share a common interest or just to be in touch with. Besides the common thrill, it is possible to make a serious relationship. Usenet News makes this discovery possible because it is a public forum. People expose their ideas broadly. This wide exposure makes it possible to find compatriots in thought. The same physical connections which carry Usenet News often also transport electronic mail. Interactions and discoveries are only made possible by the public aspect of Usenet News. Mailing Lists have as wide a range of discussions, but are exposed to a much smaller sized group. The appeal of Usenet can become tiresome at times(26), but it is rare that anyone leaves Usenet permanently. Unless, of course, someone can't find the time to fit Usenet into his or her life. As more universities, businesses, and individuals connect, the value of Usenet News grows. Each new person eventually can add his unique opinion to the collection of thoughts that Usenet already has. Each new connection also increases the area where new connections can be made through cheap local phone calls. The potential for inexpensive expansion is limited only by the oceans and other natural barriers. Arpanet has been supplemented and eventually replaced by networks like CSnet and its successor NSFnet. Both were created by the United States Government in response to research scientists' and professors' pleas to have a similar connection to the Arpanet. The NSFnet was also created to provide access to the five supercomputer computing centers around the country. And now NSFnet as the backbone of the global provides another route for Usenet News to be distributed. Similar to the Arpanet, NSFnet is a constant connection run over leased lines. NetNews is distributed using the NNTP protocol over Internet connections. This allows for News and e-mail to be distributed quickly over a large area. Internet connections also assist in carrying news and mail internationally. The Internet-class networks and connections include the established government and university sponsored connections. However much of the way individuals are connected at home is through the phone lines and various versions of UUCP. There are also commercial services that exist now for a fee that serve to provide connections for electronic mail and Usenet News access, as well as access to the Internet. Much of the development of Usenet News owes a big thanks to restrictions on commercial uses. Where else in our society is the commercial element so clearly separated from any entity? Many other forums of discussion and communication become clogged and congested when advertisements use space. On UUCPnet, people feel it wrong to assist any commercial venture through the voluntary actions of those who use and redistribute news and e-mail. When people feel someone is abusing the nature of Usenet News, they let the offender know through e-mail. In this manner users keep Usenet News as a forum that is free from the monetary benefit for any one individual. Usenet is not allowed to be a profit making venture for any one individual or group. Rather, people fight to keep it a resource that is helpful to the society as a whole. On what was the Arpanet and what is now the NSFnet and the Internet, there are Acceptable Use Policies (AUP) that exist because these networks were initially set up, founded and financed by public monies. On these networks, commercial usage is prohibited, which means it is also discouraged on other networks that gateway into the NSFnet. [Unfortunately, the NSF is now en- couraging privatization of the NSF backbone. See e.g. the U.S. Office of Inspector General's Report on NSFnet, April, 1993 -ed] However, the discouragement of commercial usage of the global Usenet News is separate and developed differently from the AUP. The social network that Usenet News represents supersedes the physical connection it rides on. The current NetNews rides on many of the physical networks that exist today. However, if need would ever be, Usenet could re- establish itself outside of the current physically organized networks. Usenet News' quality is such that it will survive because of its users will. As a peer to peer network, Usenet draws its importance. People who use Usenet News wish to communicate with others. This communal wish means that people on Usenet find it in their own and in the community's interest to be helpful. In this way, Usenet exists as a world-wide community of resources ready to be shared. Where else today is there so much knowledge that is freely available? Usenet News represents a living library. Usenet News is only a part of the worldwide computer networks that are "part of the largest machine that man has ever constructed - the global telecommunications network."(27) Usenet News began with the spirit that still exists today. On several newsgroups I posted a message with the following subject: "I want to hear from the four corners of the Net - That means YOU!" In return I received numerous wonderful answers. One new pioneer was going to use packet radio to send e-mail up to the CIS's orbiting Mir Space Station in the heavens. One person criticized Japan's lack of understanding the computer technology they supposedly "lead". Another user from France told me how the government charged a lot of money to access e-mail and Usenet News, and how there were at least two other "unofficial" connections. Since the government didn't recognize these other gateways, e-mail was to be sent via the United States in order to reach others across the street! Certain cities (e.g., Wellington, New Zealand and Cleveland, Ohio) have free public connections to Usenet News, e-mail and other network resources. Others in Krakow in Poland, Australia and the ex-USSR sent me information about their connection. Some told me of how they made other connections possible. One user in South Africa told me how he distributed news and e-mail and was trying to gain access to a satellite in order to set connections up with the interior of Africa that lacks the otherwise needed infrastructure. The world is still in the infancy of this communications interconnectivity! The very nature of Usenet News promotes change. Usenet News was born outside of established "networks", and transcends any one physical network. Currently, at this time, it exists of itself and via other networks. It makes possible the distribution of information that might otherwise not be heard through "official channels." This role makes Usenet News a herald for social change. Because of the inherent will to communicate, people who don't have access to News will want access when they become exposed to what it is, and people who currently have access will want News to expand its reach so as to further even more communication. Usenet News might grow to provide a forum for people to influence their governments. News allows for the discussion and debate of issues in a mode that facilitates a mass participation. This becomes a source of independent information. An independent source is helpful in the search for the truth. Administrators and individuals who handle the flow of information have been predicting the "imminent death of the net" since 1982.(28) The software that handles the distribution of NetNews has gone through several versions to handle the ever increasing amount of information. People who receive News have either had to decrease 1) the number of days individual messages stay at the site, 2) the number of newsgroups they receive; or they have had to allocate more disk space for the storage of News. Despite all the predictions and worries, people's desire for this communication have kept this social network floating. Brad Templeton once wrote, "If there is a gigabit network with bandwidth to spare that is willing to carry Usenet, it has plenty more growth left."(29) Brad, and everyone else will be happy to know that such a network does exist! Various research labs (including the NSF Center for Telecommunications Research at Columbia University in New York) are close to producing usable gigabit networks. Usenet News is a democratic and technological breakthrough. The computer networks and Usenet News are still developing. People need to work towards keeping connections available and fairly inexpensive, if not free, so as to encourage the body of users to grow. There are several cities and governments across the world where the public has access to network services as a civic service. This direction is to be encouraged. Exclusive arrangements for access are to be discouraged. The very nature of Usenet News means people are going to be working for its expansion. Others will be working for the expansion for their own gain, and I wouldn't doubt that some forces will be an active force against expansion of Usenet. I can only ask that people attempt to spread this document in an attempt to popularize and encourage the use and fight for Usenet News. Footnotes 1. "In September 1969, the embryonic one-node(!) Arpanet came to life when the first packet-switching computer was connected to the Sigma 7 computer at UCLA. Shortly thereafter began the interconnection of many main processors (referred to as HOSTs) at various university, industrial, and government research centers across the United States." (Kleinrock, "On Communications and Networks," IEEE Transactions on Computers, vol. C-25, No. 12, Dec, 1976, pg. 1328) 2. F. Heart, A. McKenzie, J. McQuillan, and D. Walden, Arpanet Completion Report, Washington, 1978, pg. II-2 3. Alexander McKenzie et al, "Arpanet, the Defense Data Network, and Internet" in The Froehlich/Kent Encyclopedia of Telecommunications, vol. 1, pg. 346 4. Lawrence G. Roberts, The Arpanet and Computer Networks, pg. 145 5. Leonard Kleinrock, "On Communications and Networks", IEEE Transactions on Computers, vol C-25, No. 12, Dec., 1976, pg. 1327. 6. Alexander McKenzie, pg. 357 7. F. Heart, pg. ii-25 8. Alexander McKenzie, pg. 369 9. "For many of the people in government, at the major contractors, and in the participating universities and research centers the development of the Arpanet has been an exciting time which will rank as a high point in their professional careers. In 1969 the Arpanet project represented a high risk, potentially high impact research effort. The existence of the net in practical useful form has not only provided communications technology to meet any short term needs, but it represents a formidable communications technology and experience base on which the Defense Department as well as the entire public and private sectors will depend for advanced communications needs. The strong and diverse experience base generated by the Arpanet project has placed this country ahead of all others in advanced digital communications science and technology." (Arpanet Completion Report, pg. II-109.) 10. UNIX was born in 1969, the same year as Arpanet. 11. D. M. Ritchie, "The UNIX System: The Evolution of the UNIX Time- sharing System," Bell Systems Technical Journal, vol. 63, No. 8 (October 1984), pg. 1578. 12. Ibid. 13. Stephen Daniel, James Ellis, and Tom Truscott, "USENET - A General Access UNIX Network," Duke University, Durham, NC, Summer 1980. 14. Stephen Daniel, 1992, a personal communication, November 1992. 15. Bellovin, Steve. M. and Mark Horton, "USENET - A Distributed Decentralized News System", an unpublished manuscript, 1985. 16. Ibid. 17. KEY POINT - The first gateway of Arpanet mailing lists to Usenet was an early force to have gateways with Arpanet. Gateways to Arpanet were on the side and in all likelihood not officially sanctioned. However, this provided the impetus for future gateways into Arpanet. This was the first pressure on the Arpanet to provide service to a larger number of people - a first step to transforming of the Arpanet to become a part of the backbone on the Internet. 18. Comment from Steve Bellovin, Oct. 10, 1990, Usenet History Archive: "Correct. The original concept was that most of the traffic would be the form now known as UNIX-wizards (or whatever it's called this week). Growth was slow until Mark started feeding the mailing lists in because there was nothing to offer prospective customers. Given a ready source of material, people were attracted." 19. Comment from Tom Truscott, Sept 25, 1990, Usenet History Archive: "The very first news groups were "NET." and local groups such as "dept". Later Horton et al. oversaw the lower-casing of NET. Only when ucbvax joined the net did "fa" appear. Indeed I was unaware of the Arpanet mailing lists such as human-nets until ucbvax enlightened us." 20. Steve Bellovin, Oct 10, 1990 - Usenet History Mailing List. Also - from Lauren Weinstein, Nov. 23, 1992: "Greetings. It's all too easy to forget, even for those of us who were there all along, how "small" it all started. When I was at UCLA-ATS (ARPAnet site 1) in the early 1970s, even small mailing lists could cause concern. I still distinctly remember the concerns regarding network loading from Geoff Goodfellow's NETWORK-HACKERS mailing list (this was in the days when "hacker" didn't have the negative meaning it has picked up since then) as the list passed *100* addresses. A list about wine (WINE-TASTERS, I believe it was called) which was mentioned in "Datamation" magazine caused memos to be sent out from the powers-that-be about "official use" of the net. There was also a lot of hand-wringing about the 255 site limit (that is, a limit on the number of IMPs [Interface Message Processors -ed]) in the network topology under NCP [Network Control Program -ed]. It's quite remarkable how much we accomplished on what by today's standards were slow machines with "tiny" amounts of memory, running with a 56 Kbit network backbone!" 21. Henry Spencer - Usenet History Archives "history" file. 22. From Amanda Walker, Tue, Oct. 16, 09:11 PDT, 1990, Usenet History Archives: "Indeed. I suspect that there are any number of examples of this, but the most egregious in my experience was at CWRU. The ECMP department had a VAX 11/780 on Usenet ("cwruecmp"), and the campus computer center had a DEC-20 in the room next door. The machines were separated by a grand total of about 30 feet and a piece of wallboard, but the computer center was not at all interested in "catering" to "those CS types" by stringing an RS-232 line between them. So, it was possible to send mail between them, but only by sending via a route resembling: crwuecmp => decvax => ucbvax (UUCP) ucbvax => columbia (CU20A, I think) (Arpanet) columbia => cmu-cs-c => cwru20 (CCnet) Yup, that's three networks, and two coasts just to get through a piece of sheetrock :-). Took about a week, too." 23. Alexander McKenzie, "Indeed, during a typical measurement period in June 1988, over 50% of the active Arpanet hosts were gateways, and they accounted for over 80% of the traffic." pg. 369 24. At AT&T, the computers "research", then "allegra", then "ihnp4" served as major mail and/or news distribution sites. At DEC - "decvax" gradually increased its role (e.g., "decvax" in New Hampshire would call long distance to San Diego across the country.) 25. Andrew Tabenbaum is quoted as saying something similar to "Never underestimate the bandwidth of a station wagon full of 9 track tape (or magnetic tape)." 26. "Flame Wars" (highly emotional attacks) can become annoying. There are ebbs and flows of interesting posts. Even though Usenet is addicting, it can also be overwhelming. 27. Ithiel de Sola Pool, Technologies Without Boundaries, Cambridge 1990, pg. 56. 28. From the Usenet History Archives. 29. From the "posthist" file from Usenet History Archives. BIBLIOGRAPHY Special Thanks to Bruce Jones for establishing and archiving the Usenet History Archives. Also thanks to the Pioneers for getting Usenet News off to the right start. Usenet History Archives are accessible via anonymous FTP at weber.ucsd.edu in the directory /pub/usenet.hist Bellovin, Steve M. and Mark Horton, "USENET - A Distributed Decentralized News System," an unpublished manuscript, 1985. Heart, F., A. McKenzie, J. McQuillan, and D. Walden, Arpanet Completion Report, Washington, 1978. Kleinrock, Leonard, "On Communications and Networks," in IEEE Transactions on Computers, vol C-25, No. 12, December, 1976, pg. 1326-1335. McKenzie, Alexander and David C. Walden, "Arpanet, the Defense Data Network, and Internet," in the Froehlich/Kent Encyclopedia, New York, 1991, vol 1, pg. 341-376. Ritchie, D.M., "The UNIX System: The Evolution of the UNIX Time-sharing System," Bell Systems Technical Journal, vol. 63, No. 8 (October 1984), pg. 1577-1593. Roberts, Lawrence G, "The Arpanet and Computer Networks," in A History of Personal Workstations, ed. Adele Goldberg, N.Y., 1988.