Others say the more important moment arrived eight years later, when a repurposed delivery van equipped with a wireless transmitter sent a message from San Francisco to Norway and back to California by way of satellite.
The date was November 22, 1977, and no one seems to remember what message was sent — or even who was in the van. But they do remember how it was sent. This marked the first time the TCP/IP protocol — the same protocol that underpins today’s Internet — was used to send information across not one, not two, but three independent computer networks.
“It wasn’t just a transmission,” says Bob Kahn, one of the key figures behind that moment. “It was a whole system of network protocols being demonstrated over three different networks.”
You can certainly argue that the first Internet transmission happened much earlier. The world generally agrees it happened in 1969, when researchers at the University of California at Los Angeles sent the inaugural message across the ARPAnet, the government-funded network that eventually evolved into the Internet as we know it. But the Internet and the ARPAnet are two very different things. The Internet is fundamentally a collection of disparate networks — that’s why it’s called the Internet — and this required the introduction of TCP/IP, which came of age on that delivery van in 1977.
TCP/IP was designed by two men: Vint Cerf and Bob Kahn. Earlier this year, both were inducted into the Internet Society’s (ISOC) Internet Hall of Fame, alongside such pioneers as Sir Tim Berners-Lee, Van Jacobson, and Ray Tomlinson. This past April, we celebrated Cerf’s induction, but Kahn deserves just as much credit for the rise of TCP/IP.
Cerf and Kahn developed the protocol in the early 1970s. Kahn worked for ARPA — the U.S. Department of Defense agency that funded the ARPAnet — and Cerf was an assistant professor at Stanford University. Both had been involved with the ARPAnet since it’s earliest days — Cerf as a graduate student at UCLA, and Kahn as an engineer with BBN, the Boston-based outfit that built the network’s hardware.
Kahn left BBN in 1972 to join the government office that oversaw the ARPAnet, and one day in 1973, he appeared in Cerf’s lab at Stanford. “I have a problem,” he said.
Basically, Kahn needed a way of controlling the network from the computers that connected to it. But he and Cerf didn’t want any one machine to have more control than any other — and they wanted all sorts of machines on the network. “The problem is that if you are serious about using computers, you better be able to put them in mobile vehicles, ships at sea, and aircraft, as well as at fixed installations,” Cerf says.
Eventually, they realized the project needed a protocol that could work across the many disparate networks need to connect all those machines. And then they built one. Their TCP/IP paper was published in May of 1974.
But a good two years would pass before the protocol was officially used to connect multiple networks. On August 27, 1976, a delivery van that belonged to the Stanford Research Institute — one of the research organizations attached to the ARPAnet — was sitting at a former stage coach stop somewhere between San Francisco and Monterrey, California. Since the previous year, the van — typically referred to as a “bread truck” — had served as a mobile node on a packet radio network that covered the area, but that day, for the first time, it used TCP/IP to send packets across both the wireless network and the ARPAnet.
Even still, some believe the bigger event came more than a year later when the van sent a message across a third network as well. With two networks, you’ve merely built a bridge. With three, you have an Internet. “It was true inter-networking,” Cerf said in 2007 during an event celebrating the thirtieth anniversary of the transmission. “TCP was supposed to show how multiple networks could be interconnected, and I thought it would be more convincing if we could do three networks rather than just two.”
Driving down the road somewhere south of San Francisco, the van sent a message that traveled across not only the packet radio network and the ARPAnet but a satellite network that connected the ARPAnet to Europe. It hopped from California to Boston, on to Norway and Great Britain, and then back to California by way of a small town in West Virginia.
The message was unimportant. In all likelihood, those in the van were just remotely logging into an APRAnet machine at the University of South California. But the moment was important — at least in hindsight.
“The Internet was all about a set of protocols and procedures that would allow different components to be interconnected, so that we could we could connect one machine to another over different networks,” says Bob Kahn. “[The three-network transmission] was a demonstration that the technology really worked.”
Kahn doesn’t remember if he was in the van or not. At the time, it was just another experiment. But it pushed TCP/IP forward — and TCP/IP would soon give rise to the Internet. The ARPAnet officially adopted the protocol in 1983, and by then, it was already used across research networks in Asia.
In 1985, Kahn left the Department of Defense to found the not-for-profit Corporation for National Research Initiatives. But by the then, the ARPAnet was well on its way to bigger things.