YEAR
THREADS
At age 19, Blaise Pascal (France) constructs the first mechanical
calculator and offers it for sale. The machine is capable of adding and
subtracting.
[Oxford Reference English Dictionary (1996):
under "Pascal" and "Appendix 2 - Chronology of
Scientific Developments"]
Gottfried Leibniz (Germany) designs a machine for multiplication and
division.
[Oxford Reference English Dictionary (1996):
under "Leibniz" and "Appendix 2 - Chronology of Scientific Developments"]
1642: Pascal
Charles Babbage, inventor of machine computing, designs a mechanical computer, or "difference engine", assisted by Byron's daughter Ada Lovelace. The machine could perform calculations and print the results, but neither this machine nor his later "analytical engine" are constructed in his lifetime, due to practical and financial difficulties ("though construction of one began in London in the 1990s [sic]"). 1834:1
The difference engine can handle variables in equations, and give thousands of results after the first value is set. 1834:2
1674: Leibniz
Samuel Morse invents Morse code. 1835:1
William Cooke and Charles Wheatstone patent the electric telegraph, which is used between two railway stations in London. 1837:1
George Boole (UK) invents Boolean algebra. 1849:1
Boole's Mathematical Analysis of Logic, published in 1847, was based on the 13th century logic diagrams of Ramon Lull. John Venn would later develop the Venn diagram by extension. 1849:2
Boolean algebra is the foundation of binary logic essential to computing.
1834: Babbage
The English Channel gets the first underwater cable. 1851:1
1837: Telegraph
G. Schentz (Sweden) constructs a calculating engine and exhibits it at the Paris Exhibition. 1855:1
1849: Boole
Alexander Graham Bell invents the telephone. The first telephone has only one transducer for listening and speaking, before Edison adds another one. 1876:1
1837: Telegraph
Herman Hollerith invents a "tabulating machine using punchcards for computation". The machine is "an important precursor of the electronic computer", and was used to tabulate results of the US census. 1889:1
1889: Hollerith
1855: Schentz
Following J.J. Thompson's 1897 discovery of the Edison effect (flow of electrons from a filament to a positively charged electrode), the American engineer Lee de Forest builds the first thermionic triode valve. The device can compare two electronic inputs and produce a logical output. This is the first applications of Boolean algebra. 1906:1
The valve would later be replaced by the transistor.
1849: Boole
Tabulating Machine Company of the US becomes International Business Machines. 1911:1
1896: Hollerith
Wire-tap law
The US government introduces the first wire-tap law. Law enforcement agencies use it for counter-espionage, but soon it is used to fight crime. Tapped conversations can be admitted as evidence. 1918:1
1835: Morse
John Logie Baird introduces "the first usable television with a picture of 30 vertical lines, its image changing 12.5 times per second". The cathode ray tube is a development of the thermionic valve. 1926:1
1906: Thermionic valve
London and New York get the first transatlantic phone link. 1927:1
1876: Telephone
The US government defends the use of wire-tapping in Olmstead vs. US, arguing that the Fourth Amendment only protects material things; non-material communications can thus be legally tapped by the government. 1928:1
1918: Wire-tap law
Franklin D. Roosevelt is US President from 1933-45. During World War II, he "played an important part in the coordination of the Allied war effort". He also co-authored the Atlantic Charter with Churchill in 1941. 1933:1
US Congress enacts the Federal Communications Act prohibiting the interception of wire and radio transmissions without the sender's knowledge. The Supreme Court supports the statute and rules that wire-tapped information cannot be admitted as evidence. The Justice Department resists the decision. When war threatens, Roosevelt permits wire-tapping for "national security purposes".
After the war, the courts make a gradual swing back to protection of privacy.
1928: 4th Amendment
Alan Turing develops the concept of a theoretical computing machine, "a key step to the development of the first computer". 1937:1
1911: IBM
(WWII: 1939-45)
Turing conceives of the code-breaking machine Colossus. 1943:1
Presumably, this is in "England's Government Code and Cypher School". 1943:2
1934: FCA
Supported by the US Navy, Howard Aiken from Harvard builds the Mark I, "a 51-foot-long, 8-foot-high switchboard capable of mathematic calculations" without human intervention. 1943:3
The most advanced computer is the ASCC Mark 1, "developed at Harvard University with backing from IBM". It is 51 feet long, weighs 5 tons and consists of 750,000 parts. IBM chairman Thomas Watson is quoted as saying: "I think there is a world market for maybe 5 computers." 1943:4
1937: Turing
Hitler uses his cutting-edge hardware, the Enigma (based on a discovery by the Polish), for encryption. The Germans are "adept at jumbling up their messages with complicated algorithms that can only be cracked by computers". 1943:4
1943: Turing's Colossus
P. Eckert and J. Maunchly design the US Army's ENIAC (Electronic Numerical Integrator And Calculator) at the University of Pennsylvania. 1945:1
1943: Turing's Colossus
Harry Truman assumes US Presidency upon Roosevelt's death. He holds office from 1945-53. He immediately uses the atomic bomb against Japan to end World War II. His Marshall Plan gives aid to "war-shattered European countries" and contributes to the establishment of NATO. He will later involve America in the Korean War. 1945:2
1933: FDR
Donald Watts Davies joins a team led by Alan Turing at the National Physical Laboratory to build the fastest digital computer in England at the time, the Pilot ACE. 1947:1
1943: Mark I
William Shockley and others at Bell Labs invent the transistor, which will eventually replace the thermionic valve. 1947:2
1906: Thermionic valve
President Truman expresses the Truman Doctrine, the principle extending US aid to nations under Soviet or Communist threat. It is taken by Communists as an open declaration of the Cold War. 1947:3
1943: Hitler
According to the OED, the first computer, the Manchester Mark I, is installed at Manchester University, UK. 1948:1
1943: Mark I
Printed electronic circuits are developed. 1949:1
Transistors, resistors, capacitors and other electronic components can be linked closely together on a printed circuit board, rather than being wired together separately.
1947: Transistor
At MIT's Lincoln Lab, the US Navy and Air Force support the Whirlwind machine, a system for Distant Early Warning (DEW) comprising a network of radars.
SAGE
The Whirlwind is succeeded by the SAGE (Semi-Automatic Ground Environment), a huge machine that can collect data from various radars, interpret "data relating to unidentified aircraft", and point missiles at incoming threats. SAGE is only semi-automatic in the sense that it requires a human operator. It is one of the first fully interactive real-time systems that can provide answers within a few seconds. Info flows through phone lines to the users. 1950s:1
1945: US Army's ENIAC
The SAGE inspires a few thinkers, including JCR Licklider at the MIT
Lincoln Lab, to see computing in a new light. Licklider thinks of it as
an example of human-machine symbiosis, "where the machine functions as a
problem-solving partner". Humans and machines are interdependent on each
other and form a single system.
1950s:2
JCR Licklider has a chance encounter Wes Clark at MIT's Lincoln Lab. Clark teaches Licklider how to programme the TX-2 (evolved from the TX-0), a machine which occupies a few rooms and has 64,000 bytes of memory. The TX-2 displays information on a video screen, making it one of the earliest machines for interactive graphic work. 1950s:3
Early '50s: Whirlwind
Licklider goes away and starts thinking about computers' potential to transform society. With computers, most citizens would be "informed about, and interested in, and involved in, the process of government". 1950s:4
Early 1950s: Man-machine symbiosis
There are less than a dozen electronic computers. They are so big that they fill entire "air-conditioned warehouses". 1950:1
1950s: TX-2
The designers of ENIAC come up with the Univac (Universal Automatic Computer), the first mass-produced computer.
Thermionic valves have practically been replaced.
1947: Transistor
Douglas Engelbart (who happened to be a radar operator during World War II) envisions interactive computing with keyboard and screen display (instead of on punchcards), as a way of managing an ever increasingly complex world of technology. 1950:3
1950s: Licklider
President Truman forms the NSA (National Security Agency) to protect US "executive and military communications".
The agency is shrouded in secrecy. Only a handful know of its existence. NSA practically stands for "No Such Agency". The NSA does not need warrants to tap into communications coming in and out of America. Its job is said to regularly spy on the American people.
1947: Truman Doctrine
IBM unveils the "Defence Calculator", later renamed the "701", capable of "2200 multiplications per second". The NSA uses it to break increasingly difficult encryption codes.
1943: Hitler's Enigma
IBM (instituted in 1911) starts building large electronic computers. IBM machines are not as good as Univacs (succeeding the ENIAC) but marketing strategy make its sales better than Univacs. 1953:1
1950: Univac
Dwight Eisenhower is elected US President. He holds office from 1953-61. 1953:2
1945: Truman
The first high level programming language, FORTRAN, is published by IBM. 1954:1
1953: IBM builds computers
October 1957
It is the International Geophysical Year. The Soviets launch the Sputnik. The space race begins.
There is widespread panic that the Sputnik proved Soviet capability to launch ICBMs. Eisenhower looks to the scientific community for advice. (He distrusts the military.)
"Eisenhower was the first president to host a White House dinner specifically to single out the scientific and engineering communities as guests of honour, just as the Kennedys would later play host to artists and musicians."
1947: Truman Doctrine
The Soviets launch Sputnik II, carrying the first space traveller, the dog Laika. Eisenhower picks James R. Killian Jr. as the nation's science advisor. There is great urgency to pump money into R&D. 1957:2
Sputnik I
Eisenhower forms the ARPA (Advanced Research Projects Agency) to stem rivalries in the military for R&D funding. Budget is approved for $2 billion. Roy Johnson, ARPA's first director, defines ARPA's mission in military terms — to counter the perceived Soviet threat. His vision: "global surveillance satellites, space defence interceptor vehicles, strategic orbital weapon systems, stationary communication satellites, manned space stations and a moon base". 1958:1
1957: Space Race
NASA is formed. Space projects and missile programmes are transferred from ARPA to NASA or passed back to the military. ARPA's budget is slashed to $150 million.
Roy Johnson resigns, leaving instructions to consider 4 choices for the future of ARPA:
- Abolish ARPA
- Expand ARPA
- No change
- Redefine ARPA's mission
Above: ARPA budget
ARPA reshapes itself by detaching itself from the Pentagon and focusing on long-term research efforts, in contrast to the Defence Department's short-term goals. ARPA would fund the really advanced "far-out" research. Most importantly, ARPA begins to tap the universities where the best scientific talents are located. ARPA becomes a "high-risk, high-gain" research sponsor.
ARPA begins to have a distinctive style, and its small size allows "the personality of its director to permeate the whole organisation".
1958: ARPA formed
1949: Printed circuit
Between 1959 and 1964, US R&D spending rises from $5 billion annually to $13 billion annually. That's 3% of the GNP. 1959:1
1958: ARPA budget
Paul Baran, an engineer who has worked as a technician on the Univac, the first commercial computer, joins the computer science department at RAND Corp and starts thinking about the "survivability of communication systems under nuclear attack". He is the first to see that command and control problems could be solved by digital computer technology. 1959:2
What's command and control? Command means, when you say "Launch missiles", the missiles are launched. Control means when you don't say "Launch missiles", the missiles aren't launched.
1953: IBM
Baran visits the Pentagon. He sees that computer networks could be made more "robust and reliable" by introducing redundant links. Independently of Licklider and Engelbart, he imagines "the future of digital technology and the symbiosis between humans and machines". 1959:3
1950: Engelbart
Baran invents "hot potato routing" (packet-switching) and tries to persuade AT&T of its merits, but is unsuccessful. 1959:4
Ken Olsen leaves MIT on a venture to commercialize MIT's TX-2 computer. He founds Digital Equipment Corp to make and sell components, then builds the minicomputer, which interacts directly with the user.
At the same time, time-sharing — an alternative to batch processing — is catching on.
Batch processing relies on punchcards or magnetic tape for manual input. Queues are long and it was common to wait more than a day for results to compute.
Time-sharing on the other hand gives interactive access to many users via terminals. Users interact directly with the mainframe, with the illusion that they have the computer all to themselves, when in fact they have only a fraction of the computing power. However, the direct access of time-sharing eliminates the long wait in the case of batch processing.
The first e-mail programme called Mailbox is installed on a time-sharing computer at MIT as a convenient way of getting around odd working hours. People send mail to each other on the same machine. E-mail would evolve from an interesting toy to a useful tool. 1960s:1
Late '50s: Time-sharing
The US launches the satellites Echo and Courier for relaying telephone calls between America and Europe. 1960:1
1927: Transatlantic phone link
John F. Kennedy is elected US President. He is president from 1961 to 1963. 1961:1
1953: Eisenhower
ARPA's third director, Jack P. Ruina (the first scientist to head ARPA), introduces a "relaxed management style and decentralized structure". Ruina believes in "picking the best people and letting them pick the best technology".
ARPA's annual budget goes up to $250 million.
Ruina is himself a short-timer. He isn't bothered by high turnover and believes that ARPA would "benefit from frequent exposure to fresh views".
Projects undertaken by ARPA thus far: "[b]allistic missile defence ... nuclear test detection ... behavioral research and command and control".
1958: ARPA budget
A large computer catches Ruina's attention. The huge, expensive Q-32 is a hand-me-down from an axed Air Force project. Ruina wants to use it for command and control — it would provide "high speed, reliable information" for making military decisions. 1961:3
Ruina recruits JCR Licklider, an "eminent psychologist" with "broad interdisciplinary interests", to head a new behavioral sciences office.
After spending a lot of time with computers, Licklider has long been expressing the idea that computers could be used as more than just "adding machines". They could "have the potential to act as extensions of the whole human being, as tools that could amplify the range of human intelligence and expand the reach of our analytical powers".
Late '59-'60: Baran - command & control
IBM introduces the first word processor. 1964:1
1958: Integrated circuit
At a conference in Virginia, Larry Roberts, JCR Licklider and others stay up late "talking about the potential of computer networks". Roberts goes away with the revelation that "everything worth doing inside a computer had already been done", and decides to start working on communications between computers. 1964:2
The first minicomputer is produced in the US. 1965:1
Late '50s: Olson's DEC
Tom Marill, a psychologist who has been a student of Licklider, starts a small time-sharing company looking for some R&D work. He proposes to ARPA that he conduct a "networking experiment tying Lincoln's TX-2 to the SDC Q-32 in Santa Monica". This would be one of the first real experiments connecting disparate machines over long distances.
However, Marill's company is so small that ARPA recommends the project to be operated under MIT's Lincoln Lab. Lincoln Lab takes to the idea and puts Larry Roberts happily in charge.
Marill has seen a lot of computing research being duplicated and wants to tie up all the work being done in different places. He uses a crude 2000bps modem and comes up with what he calls a protocol for sending information back and forth between computers.
1950s: Licklider
Paul Baran has the full support of RAND and sends a "formal recommendation to the Air Force that a distributed switching network be built". The Air Force agrees. However, without the cooperation of AT&T or the newly formed DCA (Defense Communications Agency), Baran decides to wait until "a competent organisation came along". 1965:3
Late '59-'60: No takers for Baran
Just after Paul Baran's halt (previous paragraph), Donald Watts Davies, a physicist at the British NPL (National Physical Lab), writes the first of several notes about a computer network much like Baran's, and sends them out to some interested people. 1965:4
Late '59-'60: Baran
Davies (previous paragraph) gives a public lecture in London describing the notion of sending short blocks of data, called packets, through a digital store-and-forward network. A man from the Ministry of Defence tells Davies about Paul Baran's work.
With encouragement from the British telecommunications people, Davis applies for funding and implements an experimental packet-switching network at the NPL.
Davies is the one who came up with the term packet-switching. Although he is later embarrassed to find out that Baran had got there first, he consoled himself with the fact that at least he "got the name". Baran's version was called distributed adaptive message block switching.
Larry Roberts makes use of an ARPA meeting at Ann Arbor to propose the experiment of connecting all time-sharing computers to one another, over dial-up telephone lines.
He is greeted with little enthusiasm as the plan would be a drain on valuable computing resources, with the added burden on the host computer of having to act as communications router. Furthermore, a standard protocol is needed as each computer speaks a different language. No one is excited about the network. No one can see why anyone would want to exchange data with anyone else, when all they need is right in front of them.
At the end of the meeting, Wes Clark suggests that a small computer be inserted between each host and the network, to do the task of routing.
By coincidence, in England, this is exactly what Davies has separately concluded.
1965: Roberts in charge
The US Supreme Court respects the right to privacy and reverses the 1928 decision on the Fourth Amendment, extending it to cover personal communications, not just physical entities. 1967:2
1928: 4th Amendment
Larry Roberts and Wes Clark implement the IMP (Interface Message Processors).
An IMP is a separate computer coming between the host computer and the network. It would send and receive data, check for errors, retransmit in the case of errors, route data, and verify message delivery.
1967: Ann Arbor
Steve Crocker distributes the first RFC (Request For Comment) titled "Host Software", an open invitation for feedback on computer handshakes.
RFCs would become the "accepted way of recommending, reviewing and adopting new technical standards". It was "a simple mechanism for distributing documentation open to anyone", used for disseminating ideas and "spreading the network culture".
An ad hoc assemblage of young talents forms the Network Working Group (NWG), an informal group trying to come up with protocols. New standards would "often emerge by consensus". 1969:1
The NWG takes the "layered approach to
protocols" — a methodology starting from low level
to high level interface.
1969:2
Above: computer handshakes
Under time pressure, the NWG comes up with Telnet, a protocol that enables remote log-ins. 1969:3
A basic protocol will not appear until 1971 — this will be the Network Control Protocol (NCP). 1969:4
This first network consisting of 4 nodes (UCLA, SRI, Santa Barbara and Utah) is established. It uses dedicated links and suffers disruptions when students begin to play with the IMPs, turning it off and on, resetting it and reloading it. 1969:5
1965: Davies' packet network
The American government has realized the "strategic value" of the information held by the "handful of computers around the world". The notion of establishing links between these computers is supported by two reinforcing ideas:
- There is advantage in cooperation and sharing of
information between the defence-related projects
taking place. Also, "frequent, informal
correspondence" could result.
- The physical security of the data stored on these computers was threatened by the Cold War's possibility of a nuclear conflict.
1947: Truman Doctrine
The ALOHANET, designed by Norm Abramson, is constructed. It consists of radios broadcasting data back and forth among "7 computers stationed over 4 islands" in Hawaii, using small taxi-cab radios. 1969:7
ARPA HQ moves out from the Pentagon to a rented office building in Arlington, Virginia. ARPA budget is declining from its "historic peak" — "[T]he Vietnam War was consuming everything". 1969:8
However, the computing budget is not decreased. Larry Roberts works successfully to get support from the top and from "an additional dozen principal investors across the country to buy into the idea of the ARPA network". Meanwhile, Frank Heart's team continues to improve the hardware and software of the IMPs. 1969:9
1961: ARPA budget
1969: Sharing resources
The movie M*A*S*H, a black comedy about war, is released.
1967: The first IMP
UCLA is connected to BBN Cambridge, Massachusetts, via a new 50-kilobit line. This is the first transcontinental link which is "also an immediate boon to network maintenance and troubleshooting". 1970:2
The BBN team is able to do monitoring, maintenance and fix problems remotely, with the capability built into the IMPs' design. 1970:3
Next to go online are:
- MIT
- RAND
- System Development Corp
- Harvard
AT&T installs cross-country 50-kilobit links between BBN and RAND, and between MIT and Uni of Utah. These are followed by more 50-kilobit links. 1970:5
1970s: Network growth
IMPs have to report to the Network Control Centre at BBN every minute to confirm that they are alive. 1970:6
Above: Remote monitoring
Alex McKenzie, convinced that the network is now ready to upgrade from experimental to operational, is appointed by Frank Heart to take charge of the Network Control Centre; his opinion was that the network should be run like "an electrical utility".
The NCC is moved into a new BBN building designed like a fortress against possible attacks by anti-war protesters.
The IMPs begin to download new operating software via a distributive propagation in which "every IMP downloaded the software from a neighbour". An IMP can also restore its own operating system this way, in case of damage. 1970:8
Larry Roberts and Frank Heart's team discuss ways to enable terminals to connect directly with IMPs. This would open up the community to casual users, further realizing Licklider's vision. The terminals would connect via dial-up lines, and the device to host them would be called a Terminal IMP, or TIP. The first TIP can handle up to 64 terminals. 1970:9
1967: The first IMP
BBN begins to sit in with the NWG to work on the host-to-host protocol, FTP and Telnet. "It was TIPs and Telnet together that paved the way for rapid expansion of the network".
FTP would be "the first application to permit two machines to cooperate as peers instead of treating one as a terminal to the other".
1969: NWG's Telnet
Intel introduces the microprocessor. 1971:1
The ARPANET is carrying an average of 675,000 packets a day, less than 2% of its capacity of 30 million packets a day. "[A] program could be moved around the network at such a speed as to approximate real time" 1971:2
1970: 1 node per month
DARPA doesn't want to run the ARPANET any more as it has completed its computer research mission, and running the network is draining its resources. Larry Roberts (who is in charge of the ARPANET project) wants to sell it to a private contractor, and asks AT&T to see if they are interested. AT&T could have monopolized the ARPANET, but they decide that packet-switching is incompatible with circuit-switching. 1971:3
Around 1959: AT&T not interested
A basic protocol, the Network Control Protocol (NCP) emerges. 1971:4
The first pocket calculator is introduced in the UK. 1972:1
1971: Microprocessor
Public outcry erupts over the US Army's "information-gathering" on suspected civilian troublemakers in the political unrest in the late '60s. The files are ordered to be destroyed immediately. 1972:2
Ray Tomlinson at BBN engages the first e-mail delivery between 2 machines.
His scheme consists of 2 separate programmes, SNDMSG and READMAIL. Although technically trivial, the experiment is a historical breakthrough. This is the guy responsible for the @ sign.
Tomlinson's programmes are released on the ARPANET in the same package as Abhay Bhushan's newly finalised file-transfer protocol (FTP).
ARPA demonstrates the network at the first International Conference on Computer Communication in Washington (ICCC). During a demonstration to AT&T officials, the computers crash, and the officials dismiss packet-switching as non-viable.
Nevertheless, the conference proves that packet-switching works at the national level. This is attested by overseas network people such as Donald Davies from UK and Louise Pouzin (see next paragraph) from France, who also attended the conference. There is a consensus that the next step will be an international network.
1971: AT&T doesn't give a dinkin' datagram
Packet-switching network projects in France and England are successful. Louis Pouzin builds Cyclades, "a French version of the ARPANET".
The leaders of several national networking projects form the International Network Working Group (INWG). INWG produces the CATENET (Concatenated Network), "a transparent interconnection of networks of disparate technologies and speeds".
1969: NWG
Larry Roberts outlines a scheme based on the ALOHANET, this time with mobile computers. The army is intrigued, and experiments with wireless, distributed packet networks. Years later, the program is phased out due to technical expenses. 1972:6
However, the wireless experiment leads to SATNET in the 1970s, a satellite network between America and Europe. SATNET does well, until transatlantic phone cables are upgraded from copper to fibre-optics. 1972:7
1960: Satellite phone
ARPA has been renamed DARPA (D for Defense) to reflect its commitment to defence research. 1972:8
Sale of the ARPANET to TELENET (a subsidiary of BBN to market a private packet-switching service) falls through due to a conflict of interest (Larry Roberts is leaving DARPA's Information and Techniques Processing Office to join TELENET and thus cannot "recommend the sale by the government"). ARPANET is temporarily transferred to the Defense Communications Agency (DCA). 1973:1
1971: DARPA doesn't want ARPANET
BBN refuses to release the IMP source code, effectively limiting IMP maintenance and repair to BBN personnel only. DARPA has to threaten to divert $6 million in contracts away from BBN. BBN relents and distributes the source code for a small fee. This represents one of the first cases of serious issues related to intellectual property. 1973:2
ARPA gives birth to the Internetting Project to link Larry Roberts' packet radio network to the ARPANET. 1973:3
1972: Packet radio network
Vint Cerf and Bob Kahn come up with the idea of gateway routing computers to negotiate between the various national networks. 1973:4
This leads to the 1974 paper, "A Protocol for Packet Network Intercommunication"
Meanwhile, traffic on the ARPANET has grown to "a daily average of 3.2 million packets". 1973:5
1971: Not enough traffic
Larry Roberts writes the first e-mail managing software called RD. People take to it instantly. Soon variations of RD proliferate. 1973:6
1972: E-mail between 2 machines
ARPANET News by SRI (Standford Research Institute) is born — a newsletter distributed on paper and electronic form.
A lobby group called USING, comprising net users, draws up "plans and recommendations for improving the delivery of computer services over the ARPANET". Nine months later, DARPA sends out a warning that they are overstepping boundaries, and the group dies off eventually.
Bob Metcalfe at Xerox PARC develops the Ethernet.
Metcalfe was inspired while working to improve the ALOHANET. The technology is developed for one of the first personal computers, the Alto from Xerox PARC, without the need for expensive IMPs. In an Ethernet, machines talk to each other like humans: when more than one tries to talk at the same time, someone will randomly go first, while the rest wait for it to finish.
1969: ALOHANET
Contracting for computer services on the ARPANET has become cost-saving due to economy of scale. Large databases are becoming popular. 1973:10
1973: Net traffic
Bob Kahn and Vint Cerf write a paper called "A Protocol for Packet Network Intercommunication" describing transmission-control protocol (TCP), essentially a method incorporating a digital envelope (a datagram) to send messages.
Whereas NCP (network control protocol) was designed for the ARPANET, TCP is a "more independent protocol" needed for linking ARPANET, SATNET and Larry Roberts' packet radio network. The paper also introduces the notion of gateways, which would read only the envelope so that only the receiving hosts would read the contents.
TCP also makes network transmission more reliable because the sending host has to receive acknowledgment from the receiving host to verify the transmission. 1974:1
US Altair produces the first portable computer. 1975:1
1972: Pocket calculator
The DCA (Defense Communications Agency) takes over the running of ARPANET. The is because federal law requires that ARPANET cannot be sold to an external party before checking to see if the Defense Department wants it. The guys at BBN complain of the increased red tape. 1975:2
1973: DCA runs ARPANET
Yogen Dalal at Stanford takes Kahn and Cerf's 1974 paper and comes up with implementable specifications for TCP. 1975:3
1974: TCP paper
Allegations surface that the files that were supposed to have been destroyed by the US Army in 1973 had instead been archived on the ARPANET. The story is reported in "the most draconian, cloak-and-dagger terms". The public is outraged that something like ARPANET existed. DARPA is called to account. Old print-outs reveal that the Army was responsible, but now DARPA is perceived to be in bed with the Army's secret operations. 1975:4
1972: Files ordered to be destroyed
John Vittal writes the MSG mailing programme which has an ANSWER command, making it easy to reply to e-mails. People won't have to retype or mistype e-mail addresses. MSG is the original 'killer application' that made e-mail popular, and is still in use in the 1990s.
"More than just a great hack, MSG was the best proof to date that on the ARPANET rules might get made, but they certainly didn't prevail. Proclamations of officialness didn't further the Net nearly so much as throwing technology out on the Net to see what worked. And when something worked, it was adopted."
1973: RD
Whitfield Diffie and Martin Hellman at Stanford University invent public-key encryption, realising that it will empower the masses. They publish the paper "New Directions in Cryptography".
Diffie and Hellman's scheme does not require the key to first pass through a secure channel. Each party has a secret key. They exchange mathematical information about each other's key to create a session key which can be used to encrypt future messages. The scheme requires real-time communications and thus cannot be used for e-mail.
1957: 4th Amendment
A private American company, Quasar Industries, launches a robot selling for $4000 which is claimed to be able to do anything a domestic maid could do, and more, including "teach the kids French". Artificial intelligence researchers recognize the robot as a fraud and speak out openly on the ARPANET. The robot is in fact remote-controlled during public demonstrations.
Some users realize that the ARPANET is now being used for personal communication that contain potentially libellous speech. They begin to practise and encourage self-censorship, or using disclaimers, for fear of attracting government supervision of ARPANET.
"What emerged from the debate was strong evidence that the networking community felt a deep stake in the creation of the Net, ARPA funding or no ARPA funding, and was trying jealously to guard its right to determine its future. In a realm where, in a sense, personal identity is defined entirely by the words people choose, free speech seemed second only to concern for the survival of the realm itself."
NSA approves and publishes a weakened version of DES (Data Encryption Standard), a secret key encryption scheme requiring the key to first pass through a secure channel. The DES is not allowed to be exported outside the US. 1977:2
1952: Truman forms NSA; NSA's 701
Meanwhile, three young professors at MIT receive the Diffie-Hellman paper, and implement a workable public-key encryption system. The system is called RSA, named after Ronald Rivest, Adi Shamir and Len Adelman.
Scientific American offers a reward of $100 for anyone who could decrypt a sentence code in the form of a 129-digit number encrypted with RSA. It won't be until 17 years later in 1994 that the code is cracked.
1976: Diffie & Hellman
Ward Christensen writes the programme "MODEM" and releases it into the public domain. MODEM allows two microcomputers to exchange files with each other over a phone line, using modems with acoustic couplers. 1977:4
Mail headers are getting out of hand. Some headers were as long as 35 lines. Ken Pogran, John Vittal, Dave Crocker and Austin Anderson announce RFC 724 (RFC = Request For Comment), a proclamation of a new e-mail standard.
It is titled "A Proposed Official Standard for the Format of ARPA Network Messages". The standard contains specifications pertaining to syntax, semantics and "lexical formalities".
Responses to RFC 724 are not enthusiastic. Jon Postel criticizes the standard for its claim to being "official". Alex McKenzie at BBN says that the ARPANET does not work on official standards, but perfects itself by trial and cooperation in an process of "step-by-step evolution". RFCs should thus play only the role of documenting such steps.
1969: the first RFC
Cerf, Kahn and others link up 3 networks using TCP: packet radio, ARPANET and SATNET. Messages travel 94,000 miles from San Francisco to London to California "without dropping a single bit". 1977:6
1974: the TCP paper
RFC 724 is re-written and published as RFC 733, now intended "strictly as a definition". Mail programs start to conform to the new guidelines, but within a year, conflicts pick up again. The problem is that the standard is incompatible with John Vittal's popular MSG mailing program. 1977:7
1975: MSG
The first magnetic tape is introduced by US Oyz. 1978:1
Inspired by Xerox PARC's PARC Universal Packet (PUP), Cerf, Postel and Danny Cohen come up with Internet Protocol (IP) by isolating the part of TCP which deals with routing packets. IP will be in charge of routing the packets, while TCP will take care of the packeting, error control, re-transmission and reassembly. TCP/IP enables fast and inexpensive gateways to be built. 1978:2
1977: TCP works
Ward Christensen and Randy Suess create the first BBS, Computer Bulletin Board System (CBBS) in Chicago, and tell people how to start their own BBSes. 1978:3
CommuniTree BBS
In Santa Cruz, California, the CommuniTree BBS goes online with a vision to build a spiritual community. Eventually, the BBS gets choked to death by antisocial teenaged students, causing A.R. Stone to conclude: "Thus, in practice, surveillance and control proved necessary adjuncts to maintaining order in the virtual community." 1978:4
A BBS distributes messages over phone lines between users on personal computers equipped with modems. The communication is independent of the Internet. BBS software is inexpensive, and anyone with a personal computer, a modem and a phone line can start one.
1973: Xerox's Alto, the 1st personal computer
British Telecom launches the first videotext information system. 1979:1
1975: Portable computer
NSF receives a proposal to build the CSNET (Computer Science Research Network) is submitted by a group of universities, initiated by Larry Landweber at the University of Wisconsin.
CSNET will be independent of the ARPANET (due to defence security restrictions), and will be cheaper to build, using slower links and no redundancy. ARPANET has been expensive to subscribe, and computer science research is falling behind due to lack of computer networking. Universities need a network to stay competitive and to train the next generation of computer scientists.
NSF rejects the proposal but remains enthusiastic.
1974: NSF network idea
Keith Peterson and Ward Christensen release XMODEM, a new file transfer protocol with error-correction, based on MODEM. 1979:3
1977: MODEM
Intel develops the RISC (Reduced Instruction Set Computer) chip. 1980s:1
1971: Microprocessor
A couple hundred networks are wired up. 1980s:3
1973-75: ARPANET expands
Landweber's committee comes up with a new proposal for CSNET, with a 3-tier structure (ARPANET, TELENET and an e-mail only network called PHONENET) to provide cheap access for everyone. The tiers are connected by gateways. The National Science Board (NSF's umbrella body) agrees to provide $5 million for CSNET. 1980:1
1973: TELENET
USENET is started by graduate students as a link between University of North Carolina and Duke University. It eventually blossoms into a distributed news network using UUCP. 1980:2
IBM introduces the desktop microcomputer, the IBM-PC. 1981:1
1971: Microprocessor
Bill Joy at Berkeley writes a version of UNIX with TCP/IP. 1981:2
Bill Joy from Berkeley joins the new Sun (Stanford University Network) Microsystems as a UNIX expert. The first Sun workstations come with Joy's UNIX with TCP/IP for free. This would be an important factor in the entrenchment of TCP/IP. 1982:1
1981: Berkeley UNIX
So far, e-mail has been riding on FTP. Jon Postel decides to build a separate mechanism for mail transfer, calling it SMTP (Simple Mail Transfer Protocol). It has new features and will be implemented together with the impending switch to TCP/IP. 1982:2
1975: MSG
On January 1st, ARPANET makes its official transition to TCP/IP. The network can now branch anywhere, and network data transfer is a piece of cake.
For security reasons, the Defense Communications Agency splits the ARPANET into:
- MILNET for sites carrying military information
- ARPANET for the research community
The two networks are connected by a gateway, so users can't tell the difference.
1978: TCP/IP
Esprit (European Strategic Programme for Research and Information Technology) is established by the EEC to coordinate the European Information Technology industry. 1983:2
Jon Postel, Paul Mockapetris (ISI) and Craig Partridge (BBN) publish two RFCs describing the Domain Name System (DNS) using tree-branching structure and specific-to-general addressing. "Eventually, a committee agreed on seven 'top-level' domains: edu, com, gov, mil, net, org and int." They stand for:
- edu = university
- com = company
- gov= government
- mil = military
- org = non-profit organization
- net = network service provider
- int = international treaty entity
1977: Standardising e-mail headers
1980: 3-tier structure
Standard modem speed is 300 bps, slower than the average human reading speed, and modems are still quite expensive — around US$500. The BBS is still an exclusive technical hobby.
Tom Jennings, a visitor of CBBS, creates Fido BBS. Flames are common, but Jennings leaves it to users to deal with them. His philosophy: "Thou shalt not offend; thou shalt not be easily offended." Fido has been described as having a "street" atmosphere, and will later expand into Fidonet.
1978: CBBS & CommuniTree
Apple introduces the Macintosh with mouse and window interface.
CD-ROM
The CD-ROM is introduced as data-storage device. 1984:1
"Cyberspace"
William Gibson publishes Neuromancer, in which he coins the word cyberspace.
1981: IBM PC
Tom Jennings helps one of his users create a version of his Fido BBS software for another computer platform. The system files can now be downloaded by users, and as a result, Fido BBSes proliferate. 1984:2
1983: the 1st Fido
Inmos (UK) manufactures the transputer, a microprocessor with integral memory designed for parallel processing.
Fibre-optics
In the US, fibre optics are first used to link mainframe computers.
1972: fibre optic links
With the availabilty of affordable 1200 bps modems, it is no longer prohibitively expensive to echo e-mails over long-distances late at night when phone rates are lowest. Tom Jennings assigns each Fido BBS with a unique node number, and Fido BBSes across America begin to observe the "National Fido Hour" from 1 to 2 am nightly when the BBSes call each other up to echo e-mails.
At around this time, the introduction of Echomail in Fidonet makes it possible for conferences between many users, instead of simply one to one.
Fidonet continues to grow.
1984: Fidos proliferate
5 supercomputer centres are scattered throughout the US, and the NSF agrees to build a backbone to link them together. NSF offers free access to the backbone network (called NSFNET) if geographical regions build for themselves community networks. As a result, community networks begin to sprout, including:
- NYSERNET: New York State Educational Research Network
- CERFnet: California Educational Research Network (in honour of Vint Cerf)
Networks can now choose between connecting with ARPANET or NSFNET. NSFNET soon becomes more popular as it is faster and easier to connect with. 1985:4
Nearly all computer science departments and many private computer research sites in the US are connected to CSNET. The CSNET success parallels other networks:
- BITNET (Because It's Time Network), an IBM network open to all
- UUCP at Bell Labs for file transfer and remote execution
- USENET
- NASA's SPAN (Space Physics Analysis Network).
All these networks communicate with TCP/IP and as a result come to be collectively called the Internet.
(Presumabely, all these networks have gateways to the NSF backbone.)
1983: CSNET grows
Other countries have their own networks too:
- Several academic service networks in Europe
- CDNet in Canada
- The JANET (Joint Academic Network) in the UK 1986:2
Eventually these networks build gateways to the US Internet, and the Internet comes to mean the international network. National boundaries begin to dissolve. 1986:3
1980s: Hundreds of networks
The NSF (National Science Foundation) takes over Internet responsibilities from DARPA. 1986:4
IBM labs in Zurich discover high-temperature superconductors with potential for superconducting computers in the future. 1986:5
1985: Transputer
Representatives from major networks meet on the West Coast for a grand summit meeting, and agree to use DNS. 1986:6
1983: DNS proposal
Dan Lynch starts the Interop trade show to promote TCP/IP. It is attended by hardcore networking people for the first couple of years. 1986:6
At around this time, Tim Pozar at SRI (Stanford Research Institute) begins working on a scheme to distribute Fidonet e-mail globally over the Internet. In other words, the Internet acts as a link between Fidonet nodes in different parts of the world. Ken Harrington at SRI provides administrative and finanial support. 1986:7
1985: Fidonet
The ISO (International Organization for Standardization) produces the OSI (Open Systems Interconnection) protocol. US and European government adopt it as the official standard, but it never catches on because TCP/IP was there first and, more importantly, has proven itself. 1988:1
TCP/IP is also very open: it can support all sorts of networks. Another reason for TCP/IP's success is the support by UNIX. 1988:2
1977: RFC 724
Dan Lynch's Interop trade show attracts business people including "Novell, Synoptics, and Network General". The Internet is starting to appeal to the business world.
Its success provided an object lesson in technology and how it advances. "Standards should be discovered, not decreed," said one computer scientist in the TCP/IP faction.
1986: The 1st Interop
Mark Pullen, a program manager at DARPA, retires the ARPANET, since the much faster NSFNET has taken over as the major backbone. ARPANET is 20 years old. 1989:2
1983: ARPANET and MILNET
1986: Global Internet
1990s: No. of networks
The National Science Foundation (NSF) lifts restrictions against commercial use of the Internet. Electronic commerce on the Net is now possible. Some of the early founders of the Net bemoan this, while others welcome it. 1991:1
1990s: Wired figures
Phil Zimmerman releases the PGP (Pretty Good Privacy) encrypting programme, essentially an amalgamation of the Diffie-Hellman algorithm, RSA and standard secret key encryption. The program finds it way to Usenet.
PGP incorporates a digital signature and can be used with
e-mail. PGP-encoded messages are virtually uncrackable by today's
computers. PGP has become the de facto standard for e-mail
encryption.
1991:2
Unable to get a licence for the RSA
algorithm, and with an impending ban on strong
cryptography, Zimmerman releases PGP 1.0, knowing that it
is illegal.
1991:3
RSA threatens to sue unless Zimmerman stops distributing PGP. Although he complies, the programme and its source code are already circulating on the Net. 1991:3
1976: Diffie and Hellman
1977: NSA's DES
The combined effort of supporters from the Netherlands, New Zealand, France and Spain results in PGP 2.0. NSA gets really pissed off by the fact that the technology is in foreign hands. 1991:5
Above: PGP 1.0
Apple releases the Powerbook laptop microcomputer, able to function as a desktop. 1991:6
The FBI drafts the Digital Telephony bill to counter the loss of eavesdropping capability as new technology arrives.
Essentially, the FBI wants wire-tapping capability to be built into communications infrastructure. But without support in Congress and with strong civil opposition, the bill dies.
1991: NSA on PGP
- Internet address registration ("thus acting as a gateway or potential chokepoint for determining exactly which sites are granted permission to join the high-speed network"), assigned to Network Solutions.
- Directory and database services ("keeping track of how to locate people and resources"), assigned to AT&T.
- Informational services provided to Net users ("modernization of tools for making use of the Internet"), assigned to General Atomics.
1991: NSF deregulates e-commerce
The US government represented by assistant attorney William Keane tries to decide whether Zimmerman should be charged with exporting munitions without a licence. 1993:2
1991: PGP
Digital Telephony is legislated. It requires phone companies to re-wire their networks so that the FBI can tap in. Civil rights supporters like Senator Leahy and the Electronic Freedom Foundation work to delete on-line information services from the legislation. Court orders are also required.
The civil rights supporters include AT&T, DEC, Lotus, Microsoft and Sun Microsystems.
1991: Digital Telephony bill
Meanwhile, the Clinton Administration’s proposed Clipper chip faces widespread opposition from academics and civil rights groups and dies eventually of technical flaws.
The Clipper chip is developed by the NSA and is an encryption scheme in which the government would hold a key in escrow to all encrypted data.
1991: Digital Telephony bill
A study by the Software Publisher’s Association reveals that although the US closely guards DES, there are 152 DES-based products being developed and distributed by 33 foreign countries, and the foreign versions are in fact superior to the American version.
1977: DES
Back in 1976, Scientific American offered a reward to decrypt an RSA code. An international team from 24 countries finally takes 8 months and 1600 workstations to crack the code. 1994:3
1977: the RSA challenge
1990: No. of users
January 1996
1993: The charge
For the latest development on cryptography issues: http://www.cdt.org
That's all I have for now, folks!