C/IL 102
Notes on The Internet: Behind the Web videorecording from The History Channel

Acknowledgement: The Internet Pioneers web site was also used as a source of information.

Perhaps no invention has ever had as much impact in as a short a span of time as has the Internet, which has become, in only 40 years, a key component of human culture and civilization, in particular with respect to entertainment, communication, and economics. Its rapid growth is due, at least in part, to the fact that it really has no controlling authority.

In a vague sense, the origins of the Internet go back to the 1940's, during which time Vannevar Bush (who had earned a Ph.D. in engineering from MIT and was the Director of the Office of Scientific Research and Development) facilitated and institutionalizied a relationship between the federal government (of the United States), the American scientific community, and business. This set the stage for the later creation of the National Science Foundation (NSF) and Advanced Research Projects Agency (ARPA). The July 1945 issue of The Atlantic Monthly included As We May Think, a very influential essay by Bush that described the concept of a memex, a device for enhancing human memory by allowing document storage and retrieval based upon "associative" links, or what we know today, in the context of web pages on the World Wide Web, as hyperlinks (which allow a reader to jump instantly from one electronic document to another).

In a somewhat more concrete sense, the Internet can be traced back to 1957, when the surprise launch into outer space of the Soviet Union's Sputnik satellite spurred the United States to initiate several research & development projects intended to re-establish its superiority in science and technology. To provide a framework for some of these projects, in early 1958 U.S. President Dwight Eisenhower formed the aforementioned ARPA. (Later the word "Defense" was added, yielding what is today called DARPA.)

J.C.R. Licklider, a psychologist at MIT (but having a strong background in math, physics, and computing), is credited with putting forth a vision of human-computer interaction that suggested, among other things, the idea of computers being able to communicate with each other in order to facilitate the exchange of information among researchers. (In particular, as an employee of ARPA from 1962 to 1964, he was interested in ARPA research contractors being able to communicate with each other efficiently.) (The video referred to his vision as a galactic network.)

During the early 1960's, the theoretical foundations for computer networking were established by Leonard Kleinrock, Paul Baran, and Donald Davies, the last of whom is not mentioned in the videorecording.

Kleinrock's PhD research (1961-62, at MIT) applied queueing theory to communications networking. (A queue is nothing more than what in the U.S. is called a "waiting line", as in a grocery store or cafeteria. In Great Britain, one typically says "queue" rather than "line".) On the video, he identifies his two crucial ideas as being

  1. demand access: To optimize resource sharing in a network, a node (i.e., an entity on the network) shouldn't get what it needs until it needs it.
  2. distributed control: (as opposed to central control) Each node should contribute to the control function, as opposed to that responsibility being given to a single node.

In Kleinrock's approach, a message is transmitted through a network (beginning at the source node, ending at the destination node, and "hopping" through intermediate nodes along a path between them) in units referred to as packets. Each packet is sent through the network independently of others (even those that are part of the same message). This approach is called packet switching, which is significantly different from circuit switching (such as is traditionally used in the telephone network), in which a dedicated path (or "circuit") is established between the two communicating nodes and is maintained until the "conversation" has ended.

With the threat of nuclear war against the Soviet Union on everyone's mind, people recognized that it was important for the country's national defense to develop the technology to support a communications network that could function even if parts of it had been damaged or destroyed.

By 1962, enough research had been done to suggest that such a network could be built. However, it would take seven more years before that would be accomplished. (The videorecording suggests that the U.S. government's support for computer network research languished during the 1960's due to its heavy focus upon the Apollo space program.)

In 1966, Robert Taylor became director of ARPA's Information Processing Techniques Office (IPTO), succeeding Licklider. In order to meet the requests of research contractors for more computing resources and to decrease duplication of work among them, he decided that ARPA should link together (i.e., make capable of communicating with each other) the computers at ARPA-funded research institutions, which would allow greater sharing of resources and results. (Among the issues making this a complicated problem was that the computers intended to be linked were of different kinds; in effect, what was envisioned was a "heterogeneous" computer network.) A million dollars was allocated to this project, and Taylor recruited Larry Roberts to manage it.

In 1967, Wes Clark gave Roberts the idea to use Interface Message Processors (IMPs) as a central feature of the design. The idea was that each mainframe host computer would be connected to an IMP, which would act as a "gateway" to the network, handling all communication with other computers on the network. In 1968, Roberts sent out bids to 140 companies to build the IMPs. By the end of the year, the contract was awarded to BBN. In January of 1969, BBN began working on it, with Frank Heart (with the rather high-pitched voice on the video) as the project manager.

By October of 1969, IMPs had been delivered to UCLA and Stanford Research Institute and the machines there connected. By the end of the year, University of California Santa Barbara and University of Utah had also been connected to the network. By April of 1971, 18 computers were on the network.

In 1972, Ray Tomlinson developed —in his spare time, because no one was really asking for it— what the video refers to as the Internet's first "killer app", e-mail (although it wouldn't be called that for several years). In particular, he adopted the convention of using the @-sign in an e-mail address.

In the video, Bob Metcalfe (who is famous for inventing the Ethernet network technology) laughs about how three significant software applications were developed during the first year or two of the fledgling Internet (namely, remote login (telnet), file transfer, and e-mail), but not another one for about twenty years.

The early developers of the Internet created an academic-like culture of knowledge-sharing and cooperation. There was no "authority" who made design decisions unilaterally and imposed them on the other participants/contributors. Rather, ideas were presented and responses solicited, so that decisions ended up being made by consensus. The video says that the genius of its design was generality of purpose.

As an example of this, during the 1970's Vint Cerf and Bob Kahn developed and refined TCP/IP, which in 1983 was adopted as the universal protocol by which messages are transmitted over the Internet. Of significance about TCP/IP is that is made it possible to connect independent networks together to form an internet (i.e., a network of networks).

During the 1980's, use of ARPA's network (which had actually been split into two) remained restricted, mostly to academic and military institutions. It is not surprising, then, that commercial networks sprang up, including Prodigy, CompuServ, and AOL (America Online).

During the early 1990's, laws were passed that opened up the Internet to commercial use. At first, the impact was not all that great, largely because only skilled computer users could effectively find information on the network.

But that changed during the period 1990-93, due to two major developments. The first (in 1991) was the World Wide Web (WWW), which is basically the marriage of hypertext (as envisioned by Bush, see above) and the Internet. Tim Berners-Lee is credited with having "invented" the WWW, as he designed and implemented HTTP (Hypertext Transfer Protocol, the rules by which hypertext is transferred/exchanged) and HTML (Hypertext Markup Language, which is the "language" in which most web pages are written).

The second (in 1993) was Mosaic, a web browser far superior to any browser available before. In particular, it had a graphical user interface, made it much easier to follow hyperlinks, and supported the inclusion of images within web pages (as opposed to images having to be viewed as stand-alone documents). Mosaic was developed by Marc Andreesen at the University of Illinois and is credited with spurring a huge increase in popularity/participation in the WWW.

An equally significant increase in usage occurred in 1995, when the major online services (Compuserv, America Online, and Prodigy) connected to the Internet, thereby allowing their millions of customers to gain access to it.

In the 15+ years since then, the Internet has become ubiquitous in the lives of millions of people, especially the young. We use it as a source of news and entertainment; for communicating with friends, shopping, doing research, and paying bills.