Internet: History

In tracing the history of the Internet, it is useful to begin at its conceptual foundation. The Internet is an example of a type of network called a packet-switched network . These networks differ from telephone networks in a number of important ways. Technological differences aside, one significant difference between these networks is that packet-switched networks are designed to support a wide variety of applications, whereas the telephone network was designed to support one application (voice communications) optimally, though a few other applications are possible as well.

Intellectually, the origin of the Internet can be traced back to the early to mid-1960s, when Leonard Kleinrock, Joseph Licklider, Paul Baran, Lawrence Rogers, and others developed the ideas and theories underpinning these general purpose packet-switched networks. By 1967 some early experiments with using packet-switching technologies were taking place at the National Physical Laboratory in England. In 1969 the U.S. Defense Department's Advanced Research Projects Agency (ARPA) funded a larger scale network project. The initial network interconnected the University of California-Los Angeles, Stanford Research Institute, University of California-Santa Barbara, and the University of Utah. Researchers at these institutions began to develop the software needed to make the network operate, and, by the end of 1969, were able to send some data packets over the network. But the capabilities were very rudimentary, and much work remained to be done.

In the early 1970s, the network software, which caused the computers in the network nodes to perform basic packet-switching functions, was standardized into the Network Control Protocol (NCP) , and new sites were added. By 1971 there were 15 locations (nodes) on the network, serving 23 host computers. As the basic network software was being developed, so were the (initially rudimentary) applications that would use the network. One of the early applications was electronic mail; in fact, the use of the now standard "@" sign for e-mail was begun in 1972.

While the primary focus had remained on constructing packet-switched networks, Robert Kahn posed the "Internet problem"—namely how to get autonomous networks to exchange information—in 1972. The idea that a network could support (and even encourage) heterogeneity would ultimately be of great importance to the success of the Internet over technologies that were unable to provide this support easily.

The network continued to grow as well. By 1973, the first international site was introduced, which was to the University College in London viaNorway; in addition, the ARPANET supported approximately 2,000 users. The genesis of today's computer environment was also being developed at this time, with the basic theory of the Ethernet local area network (LAN) , which is today the dominant local networking system, and the modern computer workstation. The Alto workstation was developed at Xerox's Palo Alto Research Center (PARC) and had a graphical user interface (GUI) with icons and a mouse. While nobody could predict the extent to which these technologies would come to dominate computing, they, together with Kahn's statement of the Internet problem, would combine to form the Internet as it is known today.

By 1974, enough had been learned about techniques for implementing packet-switching technology that a second generation protocol and the associated network software could be proposed. This was called the Transmission Control Protocol (TCP) . This proposal included what users today understand as TCP and IP. The initial tests of TCP did not take place until 1975. By this time, commercial packet-switching services (though not based on NCP or TCP) had come into being, and the possibilities that these networks afforded began to be imagined. In fact, in 1976 Queen Elizabeth II sent an e-mail.

TCP continued to be developed during this time, and, in 1978, TCP and IP were divided into separate components so that their functions could be improved. Other notable events of the late 1970s included the development of the first Multi-User Domain (MUD), the proliferation of mailing lists, and the emergence of emoticons. By this time the utility of electronic mail was more widely recognized. To extend this capability beyond the domain of the ARPANET, new networks were formed. They included CSNET, BITNET (because it is time network), and FIDONET. Each of these networks used different network protocols, had different organizational forms, and reached different users. CSNET was targeted at university computer science departments and received funding from the National Science Foundation. BITNET was targeted at a more diverse academic audience, and was organized cooperatively so that each member paid for its connection to the nearest node, and agreed to transport others' traffic. FIDONET was built upon message forwarding over dialup telephone lines and was generally used by home computer users and hobbyists.

Though Kahn had articulated the Internet problem in 1972, and work on TCP had begun in 1978, it was not until 1982 that the conversion from NCP to TCP/IP took place (the crossover took place on January 1, 1983), and the notion of the Internet was first defined as a set of interconnected networks. The conversion to TCP/IP was bolstered when the U.S. Department of Defense declared TCP/IP to be the standard for its computer networking applications. Other events of the early and mid 1980s include:

• Increased international expansion.
• The introduction of the Domain Name System (DNS), with the now familiar.com, .org, .edu names.
• Gateways between ARPANET, CSNET, and BITNET.
• The involvement of the National Science Foundation (NSF) in the funding of the Internet backbone segments (NSFNET) reaching between the five university supercomputer centers they funded (Princeton, Pittsburgh, San Diego, Cornell, and Urbana-Champaign). This would be the demise of CSNET and BITNET, because it became easy for networks to be directly connected.
• Network news was developed.

By the late 1980s, commercial interest in computer networking was growing. This was prohibited by the NSF's Acceptable Use Policy, so limited private networks (such as UUNET) began emerging. This was also the time of the first "bug," the Internet worm that disrupted many of the hosts attached to the network. The worm incident prompted the establishment of the Computer Emergency Response Team (CERT) at Carnegie Mellon University. By the end of the decade, there were more than 100,000 computers attached to the network.

It was also during this time that the utility of the Internet as an information resource began to emerge. Many researchers made their reports available via anonymous File Transfer Protocol (FTP). However, the problem of locating reports of interest brought the techniques of information storage and retrieval from the library and information science community to the Internet. The first tool for locating reports was Archie. Released in 1990, Archie was an index and search tool for anonymous FTP sites that researchers could use to locate information more efficiently. The next step was more interactive information content, which was embodied in Gopher for textual information, released in 1991; soon thereafter, an index of Gopher sites, called Veronica, was released in 1992. As useful as Gopher was, it was still limited. The hypertext-based WorldWideWeb (WWW) was initially released in 1991 by Tim Berners-Lee, which provided a framework for integrated information content. Despite this, no compelling interface for the web existed until Mosaic was released in 1993. Even without the web, the number of computers connected to the Internet increased by an order of magnitude (to 1 million) in only three years.

The promise of the Internet was not lost on commercial users. In the early 1990s, private network service providers (for example, PSINET) emerged to carry commercial traffic. These service providers created the Commercial Internet Exchange (CIX) to exchange traffic among themselves so that a user of any commercial network could contact a user of any other commercial network. The NSF came under increasing pressure to privatize the NSFNET. This was finally accomplished in 1994. The NSF continued to support next generation Internet research through projects such as the very high-speed Backbone Network Service (vBNS). With the removal of the restrictive Acceptable Use Policy of the NSF, commercial Internet ventures flourished with the introduction of streaming audio in 1995, Internet banking, and the like.

The growth explosion that followed privatization had a large impact on the broader telecommunications sector of the United States and world economies. Many traditional carriers were developing an Internet strategy, and the idea of "convergence" was central to these plans. In short, convergence refers to the notion that, as content (such as text, audio, image, and video) was digitized, any underlying network technology could be used for transport. This meant that IP carriers could easily invade the traditional "turf" of telephone companies, and other communication providers. This idea of convergence was one of the underlying forces behind the passage of the Telecommunications Act of 1996, which redefined the regulatory structure of the telecommunications industry. In essence this legislation attempted to create a policy convergence framework that could mirror the technological convergence that was going on in the industry. While notions of convergence are still relevant today, the hype and anticipation of the mid-1990s has cooled considerably.

The emergence of the Internet placed substantial pressure on existing legal structures as well. The Electronic Frontier Foundation (EFF) was founded in 1990 by Mitchell Kapor to explore these questions. As commercial interests became significant (following the Internet's privatization), issues such as trademarks, cryptography, copyright, and privacy became important legal as well as public policy issues. While there has been substantial evolution in this regard over the last decade, substantial open questions remain.

see also Bell Labs; Government Funding: Research; Internet: Backbone; Telecommunications; Telephony.

Martin Weiss

Internet Resources

Leiner, Barry M., Vinton G. Cerf, David D. Clark, Robert E. Kahn, Leonard Kleinrock, Daniel C. Lynch, Jon Postel, Lawrence G. Roberts, and Stephen Wolff. "A Brief History of the Internet." <http://www.isoc.org/internet/history/brief.shtml>

"Life on the Internet." <http://www.pbs.org/internet/>