1.2. Genesis of the Internet

Amongst such hackers were the architects of the Internet.[1] This is not to say, however, that the Internet began as a hacker project. On the contrary; it grew out of the ARPANET, a network that began its development in 1969 by DARPA, the Defence Advanced Research Projects Agency, of the US Department of Defense. Nodes of this nascent network were progressively installed at university campuses around the United States, with the size of the network numbering four nodes (or “hosts”) in 1969, 13 by 1970, and 23 by 1971. In 1972, email was developed, and became the first “killer application” of the Internet; one that was more than just a cool hack, but was a unique practical application of the new network. During the following year the network expanded overseas, with new nodes in Hawaii, Norway and England.

Over the succeeding years ARPANET expanded further by interconnecting with other wide area computer networks such as Usenet, BITNET and CSNET, a network of the National Science Foundation (NSF). It 1983 the DARPA split the military nodes of ARPANET into a separate network known as MILNET, with the balance eventually connecting into a new network NSFNET, established in 1986 by the NSF, who essentially took the place of DARPA. The Australian Academic Research Network AARNET connected with NSFNET in 1989, by which time there were over 100 000 hosts on the network. In March 1991 NSFNET and its connected networks, together known as the Internet, were opened by the NSF to commercial usage for the first time. The following year there were over a million hosts on the Internet. Today there over 350 million.

One of the innovations of the ARPANET was that its switching technology—that is, the way in which communications were directed from sender to recipient—utilised “packets” of data, rather than a dedicated circuit established between sender and recipient, as for example in the case of the telephone network.

Packet switching had several advantages, including greater efficiency, and greater robustness in the event of a network failure, as packets could take several alternate routes to the same destination. The architecture of a packet-switched network is also less hierarchical than that of a circuit-switched network, in that every node in the network is a peer to every other node, with potentially as critical a role in passing a packet towards its destination. By the same token, a packet-switched network is also intrinsically less secure than one that is circuit-switched.

The network protocols that respectively control the division of information into packets and their transmission from sender to recipient are known as TCP (Transmission Control Protocol) and IP (Internet Protocol). The TCP/IP protocol pair has comprised the basic network communications standard for the Internet since 1983, and laid the foundation for various other network protocols that were to follow.

Another important foundation of today’s network services on the Internet is the Domain Name System (DNS), which was introduced in 1984. The DNS enabled Internet servers to be accessed by means of easily-memorable names rather than numbers, and for the names to be stored in a distributed database to which all Internet hosts had access. The names were arranged in reverse-hierarchical order separated by dots, usually with the name of the server first (such as “remus” or “www”), followed by the name of the institution (such as “rutgers” or “ibm”), followed by the institution’s type (such as “edu” for an educational institution and “com” for a company).

A further important foundation of today’s Internet is BGP, the Border Gateway Protocol. Until BGP was introduced in 1994, NFSNET provided a backbone or central network that linked the various smaller networks of the Internet together. Although IP packets might have traversed several networks en route from source to destination, they always did so via NFSNET at at least one point. BGP rendered this redundant and allowed Internet routing (that is, the process by which IP packets are directed from sender to recipient across potentially numerous autonomous networks) to be decentralised. This allowed for the decommissioning of NFSNET in 1995. From that point, realising the vision of the Internet pioneers from DARPA, whilst there were a number of important constituent backbone networks, there was no single “core” of the networks of the Internet.

On top of the flexible TCP/IP network protocols, utilising the DNS for addressing and BGP for routing between autonomous systems, numerous Internet services were, and indeed continue to be, added. These so-called application layer protocols (in contrast to the lower-level transport and network layer protocols TCP/IP)[2] include the Internet email protocol called SMTP, a file transfer protocol FTP, the protocol that supports the World Wide Web, known as HTTP, and hundreds of others. The most widely used of these services have become Internet standards, and may be identified as such by the use of the number of the “Request for Comment” document (or “RFC”) assigned when their specification was first proposed to the Internet community.

The detail of the RFC process will be discussed in more depth in the following chapter, but key is the fact that a document proposed for an RFC may in principle be drafted by anyone, regardless of their affiliation. In this there are echoes of another precept of the Hacker Ethic that one author has identified; “Hackers should be judged by their hacking, not bogus criteria such as degrees, age, race, or position.”[3] The acceptance of a new Internet standard proposed in an RFC is predicated upon the achievement of consensus that it should be so accepted, from members of the Internet community who participate (freely) in an unincorporated forum known as the Internet Engineering Task Force (IETF).

Notes

[1]

References for this section are Zakon, Robert H, Hobbes’ Internet Timeline (2006), Moschovitis, Christos J P, History of the Internet: A Chronology, 1843 to the Present (1999) , Leiner, Barry M, Cerf, Vinton G, Clark, David D, Kahn, Robert E, Kleinrock, Leonard, Lynch, Daniel C, Postel, Jon, Roberts, Larry G, & Wolff, Stephen, A Brief History of the Internet (2003), IETF, FYI on “What is the Internet?" (1993) and IETF, The Internet Activities Board (1990).

[2]

This terminology derives from the OSI networking model discussed at Section 2.2.1.3, under which seven layers of a network were defined—though TCP/IP networking has no equivalent to two of these seven theoretical layers (the session and presentation layers).

[3]

Levy, Steven, Hackers: Heroes of the Computer Revolution (2001), 43.