On the Shoulders of Giants

This is a retrospective of those historically responsible for digital communication. These are the giants of our industry, the pioneers, the men, women, and corporations who created fire; our super novas. The systems they created are called networks. We call these legends, "Networkers."

 

Based on development by Jack Harrington and team at the Air Force Cambridge Research Center (AFCRC) in 1949, MODEMs were used for computer-to-computer digital communications in 1953 by the SAGE (Semi-Automatic Ground Environment), combining radar systems, telephone lines and a 22,000 square ft. monster “computer” controlled by NORAD. This was later commercialized by Bell Telephone in 1958, and ultimately became known as T-1. [1]


The term “hyperlink” was conceived by Ted Nelson in 1959, 1964, or 1965, depending which source one is willing to believe, during his stint at Project Xanadu. Samuel Taylor Coleridge would have been proud.


In 1964 a commercial airlines system called SABRE was deployed (Semi-automatic Business Research Environment), connecting two IBM main-frames to 2,000 terminals in 65 cities by American Airlines. [2]

 

Packet switching was developed in the 1960's by Paul Baran and Donald Davies, with Davies introducing it to the National Physical Laboratory in the UK utilizing a 768K line speed. [3]

 

Other papers, memoranda, and theories were espoused throughout the sixties, but none so colorfully titled as J.C.R. Licklider's memo to his colleagues at BBN (Bolt Beranek and Newman, Inc), “Intergalactic Computer Network.” [4]


1968 saw the improvement of Frequency Division Multiplexing with Time Division Multiplexing. The net effect is 45 compared to 15 time shared users on one telephone line to a single computing device.

 

ARPAnet went online in 1969. Realistically, it was the first multi-user, multi-platform data network in the history of mankind. Britain was only weeks behind the curve with ALOHAnet. This was the internet in utero.


1969 also brought together the combined efforts of Bell Labs, MIT, and GE to build a main frame time sharing system called Multiplexed Information and Time Sharing System, or Multics, which also had another goal: simultaneous multi-user access to the system. Bell pulled out due to lack of progress, but Ken Thompson from Bell forged ahead and created a hierarchical file system, command line interpreter, program and device files on Bell's PDP-7 based on Multics, which in short time would become Unics or Uniplexed Information and Computing System. This was a single tasking OS that also benefited from the likes of Dennis Ritchie, Brian Kernighan, Doug Mcllroy, and Peter G. Neumann. When pressed, none of the team members could remember exactly how Unics became Unix, although legend has it that Steve Bourne later announced it was Neumman, and Kernighan maintained he was to blame, however it was never officially attributed. [5]


Although Douglas Engelbart began working at Stanford Research Institute (SRI) in 1957, it wasn't until 1969 that he unleashed “on-line” on the world. With ARPA funding, he established a new lab at SRI called Augmentation Research Center (ARC), where he and his team developed bitmap computer screens, hypertext, collaboration tools, a harbinger to what would become known later as a GUI and a curious little device he called a “mouse”. In 1969 for a thousand unwitting attendees at SRI, he leveled the field with the introduction of the “oNLineSystem,” or NLS, showcasing these developments live for ninety minutes. Until his death in 2013, and despite his other achievements, history would mostly recall Douglas Engelbart as the “Father of the mouse.” [6]

 

A non-descript box truck (fondly referred to as “the bread truck”) began to roam the San Francisco bay area in 1973. With ARPA money and SRI technology, this later became known as Packet Radio Network (PRNET), the first mobile, packet switched, wireless “phone.” The first datagram arrived on four wheels. [7]

 

1973 also found Vint Cerf [8] and Bob Kahn building the DOD TCP/IP suite on paper. 1977 realized the first transmission of the protocol around the globe. Cerf was instrumental in the formation of ICANN and the later commercialization of TCP/IP. Cerf went to high school in Van Nuys, California with John Postel, known for SMTP, as an administrator for IANA, but mostly as the first editor of the renowned Request for Comments (RFC). In 1998 Cerf eulogized Postel with RFC 2468. It was lovingly entitled, “I remember IANA,” and it began:

 

“A long time ago, in a network far, far away...” [9]


IBM announced SNA (Systems Network Architecture) in 1974, considered the first protocol stack for interconnecting mainframe computing resources. [10]

 

Based on an idea proposed by Robert Metcalfe in 1973 using ALOHAnet as inspiration and named “luminiferous ether”, Xerox began working on what was later known as Ethernet between 1973 and 1974, filed a patent in 1975, and finally deployed in 1976. Metcalfe left Xerox in 1979 to form 3Com, where he convinced Digital Equipment Corporation, Intel, and Xerox to promote an Ethernet standard, or DIX Ethernet based on 48 bit source and destination addresses with a 16 bit ether-type field. The IEEE adopted it as the 802 standard for Local Area Networks in 1980.[11]

 

It is vital to note that throughout the 1970's and into the early 1980's, the ARPAnet continued to expand by appropriating developments both inside and outside its sphere. Below is the landscape of ARPAnet, circa 1977. [12]

 

 

 

The first MUD (Multi-User Domain) was developed at the University of Essex by two students, Richard Bartle and Roy Trubshaw, in 1978 (Trubshaw called it “dungeon”, after the game Zork). In 1981 the university allowed their “dungeon” to be accessed by students after hours to game against and socialize with each other through its computer systems connected to the ARPAnet. [13]

 

The late 1970's realized the advent of home computers, PCs, with the possiblity to connect via modem to hosted services. One such service was named Micronet, and the parent corporation eventually decided to rebrand it to its own name: Compuserv. [14]


Virginia Strazisar was credited with creating the first TCP/IP router in 1979. She had worked on ARPAnet starting in 1969 and was also considered the pioneer behind ARPAnet's first gateway between mainframes. She was a key player in ARPA's packet switching “bread truck” wireless San Francisco experiment mentioned above. She was a contemporary of Vint Cerf, among others, and a heavyweight at BBN. Sadly the information on “Ginny” is unfairly limited. But I submit to you, ladies and gentleman, this: the first router was built by a girl. So there! [15]


The 1980's began rife with activity. In 1981 France Telecomm went live with Minitel, connecting all willing subscribers via terminals to its telecom network. Minitel is widely considered the first “web.” Bbs's (bulletin boards) began to crop up, office mavericks were connecting machines and peripherals using low level Ethernet and token-ring. File and print sharing were offerings by upstarts such as Novell Netware and Apple (Appletalk). Professional systems like Lexis-Nexis began to invade the space. ARPAnet was incorporating TCP/IP. DNS came alive in 1983, predating the dot com bubble by 17 years. Radia Perelman conceived STP. Symbolics.com became the first domain. The author of “Neuromancer,” William Gibson, invented the word “cyberspace." The OSI (Open Systems Interconnect) was established. IBM lashed out with its proprietary version of token ring, and Decnet fought back with 48 bit MAC addressed Ethernet. NSFnet went online in 1986, connecting the states and piggy backing off ARPAnet. In 1987 Cisco jumped into the fray with its first router, and there were already 20,000 connected hosts. World.std.com became the first dial-up provider in 1989. The internet was spanked to its first crying breath. [16]

 

In 1989 BGP was conceived on two napkins by Kirk Lougheed of Cisco and Yakov Rekhter of IBM over lunch at an IETF conference cafeteria. It was affectionately referred to as the Two Napkin Protocol, of course. [17]

 

 

 

 

From here, the tale is largely driven by the internet, software leviathans, hardware titans, a corporate insistence for a worldwide presence, and a human thirst to establish a global community. We know those stories.

 

This was meant to establish the gestation of digital communication and to recognize the pioneers those well known to history and those often overlooked who walk away with the arrows in their backs.

 

[1]https://en.wikipedia.org/wiki/Semi-Automatic_Ground_Environment

[2]https://en.wikipedia.org/wiki/Computer_network

[3]https://en.wikipedia.org/wiki/Donald_Davies

[4]https://en.wikipedia.org/wiki/J._C._R._Licklider

[1]https://en.wikipedia.org/wiki/Semi-Automatic_Ground_Environment

[2]https://en.wikipedia.org/wiki/Computer_network

[3]https://en.wikipedia.org/wiki/Donald_Davies

[4]https://en.wikipedia.org/wiki/J._C._R._Licklider

[5]https://en.wikipedia.org/wiki/History_of_Unix

[6]https://en.wikipedia.org/wiki/Douglas_Engelbart#SRI_and_the_Augmentation_Research_Center

[7]https://en.wikipedia.org/wiki/Packet_radio

[8]https://en.wikipedia.org/wiki/Vint_Cerf

[9]https://www.rfc-editor.org/rfc/rfc2468.txt

[10]https://en.wikipedia.org/wiki/IBM_Systems_Network_Architecture [11]https://en.wikipedia.org/wiki/Ethernet

[12]https://upload.wikimedia.org/wikipedia/commons/b/bf/Arpanet_logical_map,_march_1977.png

[13]https://en.wikipedia.org/wiki/MUD1

[14]https://en.wikipedia.org/wiki/CompuServe

[15]http://www.networksorcery.com/enp/ien/ien109.txt

[16Internet History Timeline: ARPANET to the World Wide Web

[17]http://www.computerhistory.org/atchm/the-two-napkin-protocol/