The Internet

The Internet is simultaneously a global broadcasting medium, a mechanism for spreading information, and a space in which individuals and their computers collaborate and interact¹. It began with ARPANET, a single pioneering packet-switching network, and grew into a worldwide system of many independent networks joined by open-architecture networking¹. At the technical level it is a network of networks: the Internet Protocol was designed for interconnected systems of packet-switched computer communication networks, carrying blocks of data called datagrams between hosts identified by fixed-length addresses².

The network also has a formal definition¹. On 24 October 1995 the Federal Networking Council of the United States unanimously adopted a resolution defining the term¹. Under that resolution the Internet is the global information system joined by a globally unique address space built on the Internet Protocol, capable of carrying TCP/IP communications, and offering high-level services layered on top of that infrastructure¹.

Crucially, the Internet was designed as a general-purpose infrastructure rather than for one application¹. Its generality is what made later inventions possible: electronic mail arrived on the ARPANET in 1972 and remained the largest network application for over a decade¹, and the World Wide Web emerged on top of the same foundation years afterwards¹.

The idea of networked computers predates the hardware, and it began with two strands of work¹. In August 1962 J.C.R. Licklider of MIT wrote the first recorded memos describing his 'Galactic Network' concept — a globally interconnected set of computers through which anyone could quickly reach data and programs from any site, in spirit very much like today's Internet¹. He became the first head of DARPA's computer research program in October 1962¹. The theory was being worked out at the same time: Leonard Kleinrock at MIT published the first paper on packet switching theory in July 1961 and the first book on the subject in 1964¹.

In 1965, working with Thomas Merrill, Lawrence Roberts connected the TX-2 computer in Massachusetts to the Q-32 in California over a low-speed dial-up telephone line, creating the first — however small — wide-area computer network ever built¹. Paul Baran's group at RAND wrote a paper on packet switching networks for secure voice in 1964¹. The work at MIT (1961–1967), RAND (1962–1965) and the National Physical Laboratory in Britain (1964–1967) proceeded in parallel without the researchers knowing about each other's work¹.

The word 'packet' itself was adopted from the work of Donald Davies at the NPL, and the proposed ARPANET line speed was raised from 2.4 kbps to 50 kbps¹. In late 1966 Roberts moved to DARPA to develop the computer network concept and published his plan for the 'ARPANET' in 1967¹. That plan set out the packet-switched network that DARPA and the funded research community would refine into specifications¹.

In August 1968 DARPA released a request for quotations for the ARPANET's packet switches, the Interface Message Processors (IMPs)¹. The contract was won in December 1968 by a group headed by Frank Heart at Bolt Beranek and Newman (BBN)¹. In September 1969 BBN installed the first IMP at UCLA and connected the first host computer¹. By the end of 1969 four host computers formed the initial ARPANET¹.

The first real test came on 29 October 1969, when a team led by UCLA computer scientist Leonard Kleinrock tried to send a message from a computer at UCLA to one at the Stanford Research Institute⁷. Graduate student Charley Kline typed the first two letters of the word 'login' before the network crashed — hence the message 'Lo'⁷. Two computers, several hundred miles apart, had communicated, a moment many now consider the beginning of the Internet⁷. Switched on in late October 1969, the ARPANET was the first large-scale, general-purpose computer network to connect different kinds of computers together⁴. Kleinrock later recalled typing the first two letters, 'Lo' for 'Login', from the UCLA computer room before the crash⁸.

In 1969 Steve Crocker, then at UCLA, established the Request for Comments (RFC) series as an informal, fast way to share ideas among network researchers¹. In December 1970 the Network Working Group under Crocker finished the initial host-to-host protocol, the Network Control Protocol (NCP)¹. In October 1972 Bob Kahn organized a large, very successful demonstration of the ARPANET at the International Computer Communication Conference — the first public demonstration of the new network technology¹.

Electronic mail arrived the same year: in March 1972 Ray Tomlinson at BBN wrote the basic send-and-read software¹, and for his networked email protocol on the ARPANET he chose the now-iconic '@' sign⁴. Email took off as the largest network application for over a decade¹. By 1973 well over 50% of traffic on the research-oriented network was email⁴.

The step from one network to a network of networks came from Bob Kahn, who introduced the idea of open-architecture networking shortly after arriving at DARPA in 1972, under a program then called 'Internetting'¹. In the spring of 1973 he asked Vint Cerf, then at Stanford, to work with him on the detailed design of the protocol¹. The first written version of their approach was distributed as INWG#39 at a special meeting of the International Network Working Group at Sussex University in September 1973¹. In Kahn's early design, black boxes — later called gateways and routers — would connect the networks while retaining no information about the individual flows of packets passing through them¹.

The refined Cerf–Kahn paper, 'A protocol for packet network intercommunication', was published in May 1974 in IEEE Transactions on Communications¹. RFC 675, the specification of the Internet Transmission Control Program by Vinton Cerf, Yogen Dalal and Carl Sunshine, followed in December 1974⁹. The original TCP was later reorganized into two protocols — a simple IP for addressing and forwarding individual packets, and TCP for flow control and recovery from lost packets — with the User Datagram Protocol (UDP) added for applications that did not want TCP's services¹. In 1977 Cerf and Kahn linked the ARPANET, the Packet Radio Network and the Satellite Network and proved the efficacy of TCP/IP in a round-the-world transmission from a moving vehicle, the SRI Packet Radio Research van⁴. The Internet Protocol specification, RFC 791, was published in September 1981, defining a protocol for interconnected systems of packet-switched networks that carries blocks of data called datagrams between hosts identified by fixed-length addresses².

The ARPANET switched from NCP to TCP/IP on 1 January 1983 in a 'flag-day' transition that required all hosts to convert simultaneously; it went surprisingly smoothly and produced buttons reading 'I survived the TCP/IP transition'¹. TCP/IP had been adopted as a U.S. defense standard in 1980, and the transition allowed the network to be split into MILNET for military use and ARPANET for research¹. The incorporation of TCP/IP into the Unix BSD system releases proved a critical element in spreading the protocols through the research community¹.

The Domain Name System (DNS) was invented by Paul Mockapetris of USC/ISI as a scalable distributed mechanism for resolving hierarchical host names into Internet addresses¹. It was introduced by RFC 882, published in November 1983¹⁰. From 1986 to 1995 NSFNET carried the network forward with $200 million of funding, and by the time the ARPANET was finally decommissioned in 1990, TCP/IP had supplanted or marginalized most other wide-area network protocols worldwide¹.

Alongside the protocols, the institutions of the modern Internet took shape¹¹. The Internet Engineering Task Force (IETF), founded in 1986, is the premier standards development organization for the Internet: it makes voluntary standards but in no way controls or patrols the network¹¹. There is no membership in the IETF — anyone can participate — and over 7,000 people actively take part in its work in any one year¹¹. Its technical documentation is published as RFCs, which describe the Internet's technical foundations such as addressing, routing and transport technologies¹¹. The Internet Society was formed in 1991, and ICANN was founded in 1998 out of a U.S. Government commitment to transfer the policy and technical management of the DNS to a non-profit corporation with global participation¹⁵; the IANA Stewardship Transition was completed in October 2016⁵.

The hypertext idea had a long prehistory: in 1945 Vannevar Bush wrote an article describing the Memex, a photo-electrical-mechanical device for memory extension that could make and follow links between documents on microfiche¹². In the 1960s Doug Engelbart prototyped the 'oNLine System' (NLS) for hypertext browsing, editing and email, inventing the mouse for that purpose¹². At CERN itself, Tim Berners-Lee created Enquire, a networked hypertext system used for project management, in 1980⁴. He wrote the first proposal for the World Wide Web in March 1989 and a second in May 1990, and together with Belgian systems engineer Robert Cailliau formalised it as a management proposal in November 1990⁶. Neither proposal was approved, but Berners-Lee proceeded anyway, and by Christmas 1990 he had prototyped 'WorldWideWeb' in three months on a NeXT computer — featuring a server, HTML, URLs and the first browser⁴.

By the end of 1990 the first Web server and browser were up and running at CERN, with the server code developed on a NeXT computer⁶. To keep it from being switched off, the machine bore a hand-written label in red ink warning that it was a server and must not be powered down⁶. info.cern.ch was the address of the world's first website and Web server, and the first Web page address was http://info.cern.ch/hypertext/WWW/TheProject.html⁶. That first page introduced the project as an initiative for retrieving information across a wide area, built on hypermedia and aiming to give universal access to a large universe of documents¹³. Because few users had NeXT computers, a simpler line-mode browser that could run on any system was written by Nicola Pellow during her student work placement at CERN⁶.

Berners-Lee released the WWW software in 1991: the line-mode browser, Web server software and a library for developers⁶. It became available to colleagues using CERN computers in March 1991, and on 6 August 1991 the files were posted to the alt.hypertext newsgroup, marking the Web's public introduction as interest spread around the world⁶¹². Thanks to Paul Kunz and Louise Addis, the first Web server in the United States came online in December 1991 at the Stanford Linear Accelerator Center (SLAC) in California⁶.

Early in 1993 the National Center for Supercomputing Applications (NCSA) at the University of Illinois released the first version of its Mosaic browser for the X Window System, with PC and Macintosh versions following shortly afterwards⁶. Written by student Marc Andreessen and UNIX expert Eric Bina, Mosaic was the first browser supported by a major institution and started the Web on the road from research project to blockbuster success⁴. On 30 April 1993 CERN put the World Wide Web software in the public domain, declaring that the technology would be freely usable by anyone with no fees payable to CERN — actions that allowed the web to flourish³¹². By late 1993 there were over 500 known Web servers, and the WWW accounted for 1% of internet traffic⁶.

1994 was the 'Year of the Web': the First International World Wide Web conference, initiated by Robert Cailliau, was held at CERN in May and attended by 380 users and developers, and was hailed as the 'Woodstock of the Web'⁶. By the end of 1994 the Web had 10,000 servers — 2,000 of them commercial — and 10 million users, with traffic equivalent to shipping the entire collected works of Shakespeare every second⁶. That year Berners-Lee left CERN to join MIT and founded the World Wide Web Consortium (W3C), established on 1 October 1994⁶¹².

ITU estimates that approximately 6 billion people — 74 per cent of the world's population — were using the Internet in 2025¹⁴. That is up from 60 per cent of the world population in 2020, with an estimated 1.3 billion people coming online during that period¹⁴. Still, 2.2 billion people remain offline¹⁴.

The network today still runs on protocol foundations laid decades earlier: the Internet Protocol of RFC 791 (IPv4) was specified in September 1981², and its successor, IPv6 — also known as IP Next Generation — was specified by RFC 2460, published in December 1998 by S. Deering and R. Hinden¹⁵. The RFCs remain the Internet's technical foundations, describing addressing, routing and transport technologies, and the IETF that publishes them remains open to anyone who wants to participate¹¹.