6 Introductum Wide Web are prominent examples of informally created applications that became popular, not as the result of some central agency's mar­ Heat and ,-,UIH..4< keting plan, but through the spontaneous decisions of thousands of a.ndMeanings of Hacket(Switching independent users. In reconstructing the history of the Internet, I have been struck time and again by. the unexpected twists and turns its development has taken. Often a well-laid plan was abandoned after a short time and replaced by a new approach from an unexpected quarter..Rapid advances, such as the introduction of personal computers and the invention of local-area networks, continually threatened to make existing network technologies obsolete. In addition, responsibility for operating the Internet changed hands several times over the course Of all the ARPANET's technical innovations, perhaps the most cele­ of its first thirty years or so. How, in the face of all this change and brated was packet switching. Packet switching was an experimental, uncertainty, did the system survive and even flourish? I believe that even controversial method for transmitting data across a network. Its the key to the Internet's success was a commitment to flexibility and proponents claimed that it would increase the efficiency, reliability, and diversity, both in technical design and in organizational culture. No speed of data communications, butit was also quite complex to imple­ one could predict the specific changes that would revolutionize the ment, and some communications experts argued that the technique computing and communications industries at the end of the twentieth would never work. Indeed, one reason the ARPANET became the century. A network architecture designed to accommodate a variety of focus of so much attenrion-wirhin the computer science community computing technologies, combined with an informal and inclusive was that it represented the first large-scale demonstration of the feasi­ management style, gave the Internet system the ability to adapt to an bility of packet switching. 1 The successful use. of packet switching in un predictable environment. the ARPANET and in other early networks paved the way for the The Internet's identity as a communication medium was not inher­ technique's widespread adoption, and at the end of the twentieth ent in the technology; it was constructed rhrough a series of social century packet switching continued to be the dominant .networking choices. The ingenuity of the system's builders and the practices of its practice. It had moved from the margins to the center, from experi­ users have proved just as crucial as computers and telephone circuits mental tp"normaJ" technology." in defining the structure and purpose of the Internet. That is what Many computer professionals have seen packet switching as having the title of this book, Inventing the Internet, is meant to evoke: not an obvious technical advantages over alternative nlt;:thods for transmitting isolated act of invention, but rather the idea that the meaning of the data, and they have tended to treat its widespread adoption as a Internet had to be.invented-and constantly reinvented-at the same natural result of these advantages. In fact, however, the success of time as the technology itself. I hope that this perspective will prove packet switching was not a sure thing, and 1'0\ many years there was useful to those of us, experts and users alike, who are even now no consensus on what its defining cbcaracteristics were, what advan­ engaged in reinventing the Internet. tages it offered, or how it should be implemented--in part because computer scientists evaluated it in ideological as well as technical terms. Before packet switching could achieve legitimacy in the eyes of data communications practitioners, its proponents had to prove that it would work by building demonstration networks. The wide disparity in the outcomes of these e<i\.rly experiraents with packet switching demonstrates that the concept different and that, far from being a a <H"pr,Ar 8 Chapter 1 Heat War 9 technology's winning out, the "success" of packet switching depended . U-2 plane was greatly on how it was interpreted. in the pubhc an . ~ I he USSR in 1960 the Berlin Wall went up m 190 , Packet switching was inv~ntecl independently by two C.ompmer down boverg~tethe Cuban Mis~ile Crisis. The shadow of nuclear researchers working in very different Contexts: Paul Baran at the Rand loomedrouover popular culture. .I'.he On the (Shute. Corporation in the United States and Donald Davies at the National R'Z-Safe (Burdick and Wheeler 1962)-both made mto Physical Laboratory in England. Baran was first to explore the idea, in tl at . 1 1960s-presented accounts of nuclear war around 1960; Davies came up with his own version ofpacket switching m()Vl1eS m ie eary '.. United States a few years later and subsequently learned of Baran's prior work. its aftermath. And in 1964, mOVIe theaters ' ., StlYLtl.,ge- presented a brilliant black comedy of Cold War paranOla, Dr. Davies was instrumental in passing on the knowledge ofpacket switch­ ing that he and Baran had developed to Lawrence Roberts, who was (Kubrick 1963). 1 hili Dr though humorous, highlighted.the Vou nera 1 ~ty . in charge of creating the ARPANET. This chain of invention and St-rangeZave '. 1 r,Unitedf States', communications channe1s to d'lSI ,,uptlon by a SOVIet dissemination has become a standard element of origin stories about ::~ck which might make them unavailable just when th~y were the Internet; indeed, it is easy to get the impression that packet switching simply took a detour through the United Kingdom before needed most. In the movie, a psychotic Air F.orce c..omman~el nk~me~ h.0I t · motion by 111VO mg a re-emerging, unchanged, in the United States to fulfill its destiny as J k D Ripper sets a nuclear ocaus in •.. .. the underlying technology of the ARPANETl ac . of mutual assured destruction called "Plan R." ThIS plan- strategy .. ent president's authority to declare However, while Baran's and Davies's versions of packet switching which allows Ripper to crreumv . '1" had some basic technical similarities, their conceptions ofwhat defined war-is specifically designed to compensate for a wa.rtime f~l Uler~~ conllnan ,.control. , and communications. In the mOVIe, an Air Fo packet switching and ofwhat it was good for were very different. Much d general explains: of this difference was due to the strong political pressures that were brought to bear on computing research in tlwJJnited Kingdom and . Ian in which a lower-echelon commander may Plan R IS an emerl-?eflcy war p . k k-ifthe normal chain ofcommand in the United States. Large computer projects in both countries were order nucle.ar retal;uon a~~r ~~~:aw::t:~ discourage the Russkies from any develQped in a context of government funding and control, and has been disrupte .. .. e .. W ihi t n as part ofa general sneak national leaders Saw computers as a strategictechnQIQgy for achieving h that they could knock out as mg 0 .. d I attackope and escape retaI'ia tion because oflack of proper command an contro. important political goals. But inthe very different policy contexts of the United States and the United Kingdom, packet switching took on PI . R allows Ripper to launch a "retaliatory" attack even th.oug~1 n~ different meanings for Baran, Davies, and Roberts. Packet switching firstan strike has actually OCCUlTed. In reality,(as t hems fil ' disclaimer. was never adopted on the basis ofpurely technical criteria, but always ) th U S Air Force never·ha.d any such strat.. egy. Even before Dr. states, e . d'ffl rent solu- because it fit into a broader socio-technical understanding ofhow data Z d the Air Force was explonng....a very 1. e . te n Strange ave opene , networks could and should be used. t the threat of a first strike: building a commumcauons sys 1 tron 0 . d h" er command that would be able to surviy,e a~ attack an . so t at prop. 133 has Networking Dr. Strangelove: The Cold War ROQts ofPacket Switching an dcentro 1" could be maintained. As Edwards (1996, p... ) in the United States d C ld War defense analysts saw robust cornmurucauons documente, 0 f "Fl ible k a necessity in any nuclear'.con rontauon: exioie- networ s as As the 1960s opened, relations between the United States and the t .aregy required that political leaders contmue to commu- response s 1 . Th ,fore preserv- Union of Soviet Socialist Republics were distinctly chilly. The USSR nicate during an escalating nuclear exchange. el e. ' . had launched its Sputnik satellite in 1957, setting off alarm in the mg,centra 1comman d and control-political., ... leadership,·'hbuthi halso t United States over a "science gap" and prompting a surge of govern­ . ' .. cedata and communications Iinks-s-achieved t e Ig es reconnaissan , . ., ... .•. ,,' ,. was ment investment in science and technology. A series ofevents kept the l1lilitaryyrioTity." The need for survivable erally recognized the Co/4War 11 10 Chapter 1 White Heat was a researcher at the Air Force's premier "think tank," the Rand Baran Corporation. for a new COllllllu,nilcatiOllS Founded by the Air Force in 1946 as an outgrowth of operations sUI'viv'abllltY with high capacity (ibid., pp. research efforts initiated during World War II, Rand (originally sioned a system would allow personnel to on VOICe RAND, derived from "research and development") was a nonprofit conversations or to use teletype, or low~speed computer corporation dedicated to research on military strategy and technology. terminalsunder wartime conditions. to this new was Rand was primarily funded by contracts from the Air Force, though a technique that Baran (1960, p.