Chapter 1 - Introduction
1.6 A Brief Overview of Computer Communications 1968-1988
Now to briefly overview the three waves of computer communication from 1968-1988 with the names of the corporations playing prominent roles in this history:
Data Communications emerged between 1967-1971 in response to the growth in use of mainframe computers and the desire for remote peripheral connections, regulatory changes, growth in the number of timesharing and service bureau firms, advances in semiconductor technology, availability of venture capital and a hot IPO market. Over one hundred firms announced modem or multiplexer products. In 1968, Codex Corporation introduced the world’s first 9600 bit per second modem and American Data Systems (to become Micom in the future), the first time-division multiplexer. Other competitors were: Milgo, Infotron, General DataComm, Timeplex, Paradyne, Vadic and Universal Data Systems. The Data Communication firms were shielded from new competition through most of the 1970’s when experts forecasted sales of modems and multiplexers to peak at an uninteresting $150 million. In 1980, sales of Data Communication products exceeded $1 billion.
Years before the emergence of Data Communications, however, computer scientists sponsored by the Advanced Research Projects Agency (ARPA) proved the circuit-based telephone system woefully lacking for interconnecting computers. In essence, they demonstrated that the paradigm of communications inherent to the Data Communication firms, (i.e. computer communications using the circuit switching architecture of the telephone network), was inadequate for computer networks. In 1968, ARPA funded the Arpanet computer communications network. It was the first network designed for computers rather than for humans. The radical step was to build a message-based, not circuit-based, communications network. In 1972, a public demonstration of Arpanet proved to the world that a new paradigm, “packet switching,” worked.
Yet even before the public demonstration ARPA scientists knew Arpanet would work and that the military would want to interconnect other types of networks, such as radio-based or satellite networks, to the Arpanet. But the Network Control Program networking protocol of Arpanet assumed that all computers existed on the Arpanet. How to interconnect networks became a nearly decade-long effort culminating in three important, highly similar, yet incompatible, public internetworking protocols: XNS, TCP/IP, and OSI. (Xerox Corporation developed XNS and made some of the protocol public. There were also other entirely proprietary protocols, e.g. IBM’s SNA, DEC’s DECNET, etc.)
During the 1970s, the proof that packet switching worked also inspired the creation of many kinds of networks. Three networks targeted to the office and factory and evolved from packet switching will be of the most interest: Ethernet, token ring and token bus. At first, these local area networks (LANs) did not use the telephone network; they interconnected the computers in a building or campus using coaxial cable. This distinction would cause the Data Communication firms to ignore LANs until they either had to enter the market through acquisition or found it too late to adapt to this new technology.
As the 1970s came to a close, the explosion in the use of minicomputers caused corporate buyers to seek products enabling single terminals to connect to multiple computers, or computers to other computers. In 1979, eager entrepreneurs launched three important LAN start-ups: 3Com, Ungermann-Bass and Sytek. 3Com and Ungermann-Bass elected to offer Ethernet products. Sytek chose to pursue broadband-based products. The two leading Data Communication firms, Codex and Micom, also encountered this growing customer requirement, and responded with products rooted in their multiplexer technologies, a circuit-switching model, rather than the packet-switching paradigm inherent in the leading LAN architectures. While data PBXs cost a lot less at first, they lacked the advantages of LANs. The competitive phase of Networking was particularly intense as over two hundred firms announced products. Other prominent firms to this history are: Interlan, Bridge Communications, Concord Data Systems, Proteon, Excelan and Communication Machinery Corporation.
Fortunately, users, especially government agencies, anticipated the uncertainties surrounding LAN technologies and internetworking protocols. In the late 1970’s, a number of standards-making efforts were launched. The two that will be observed most closely are IEEE 802, to determine LAN standards, and ISO/OSI, to determine networking and internetworking protocols as well as LAN standards. IEEE 802 issued its Ethernet standard in 1983, and, in subsequent years, token bus and token ring standards. The OSI standards, especially those embracing LANs, took longer to negotiate. Vendors and customers gravitated to the only standard that was fully public, TCP/IP.
During the years standards were being negotiated, the battle for the Office of the Future contributed to the competitive chaos that engulfed Networking. Experts envisioned all information technologies one day being digital, and a time rapidly approaching when voice, data and video information would be carried on the same communication networks. The question became which black box would do the switching: the traditional PBX, the new digital PBXs, or LANs? Everyone had an opinion for no product as of then had shown it could handle all three data types.
Then in August 1981, IBM introduced its personal computer. The Networking market would soon be changed forever. While it would take a few years for the profoundly different needs of personal computers and computer terminals to match general customer requirements, when they did, LAN sales soared and data PBX sales collapsed. Not all LAN firms saw the changes being wrought by the personal computer, or reacted quickly enough. Corporations went on a tear installing networks, increasingly personal computer networks, and by 1985, Networking sales totaled $1 billion, up from a mere $62 million in 1982.
By 1985-1986, corporations were creating not only LAN networks; they were building their own enterprise-wide voice networks, or wide area networks (WANs). WANs were made feasible by the dropping prices of high-capacity, digital T-1 circuits and newly available T-1 multiplexers. Once again new start-ups, such as Network Equipment Technologies, forced the action. Existing Data Communication firms, like Codex and Micom, struggled to respond. Before long, corporations wanted to transmit their LAN traffic over their WANs. A new breed of Internetworking firms, like cisco Systems and Wellfleet, emerged. The routers they developed introduced a new architecture and new operating systems designed for packet switching. Data Communication, and most of the Networking firms were caught flat-footed or were so consumed with other opportunities, or problems, they lost out on what would in time be the largest computer communication market: Internetworking.
In 1988, two public demonstrations proved that computers from many vendors could function seamlessly as one large network. At the Enterprise Networking Event the OSI networking protocols were demonstrated and at Interop, TCP/IP. What happened to the firms of interest - Codex, Micom, 3Com, Ungermann-Bass, NET and Wellfleet - completes this history, that is, with a few surprises. (You probably can see some already.)
Computer communications was essential to the growing use of computers, first by letting remote locations access Host computers, and then by giving users easy access to multiple computers, first locally on a network, and then through internetworking to any network anywhere. Enterprise-wide, global information systems could not exist without computer communications. The modern global Information Economy could not have emerged without computer communications.
This history attempts to reconstruct the sheer joy, never-ending struggles, glorious victories, disheartening failures, utter madness and inner workings of entrepreneurial capitalism. For the emergence of computer communications from 1968 to 1988 represents an example of entrepreneurial capitalism at its best: when visions were transformed into the infrastructure interconnecting the world’s computers. It is a history of numerous technological innovations becoming economic growth through firms selling products: products totaling $5 billion of sales in 1988. I hope you enjoy reading a small fraction of the riveting stories of the individuals, firms and organizations that made it happen.