How the World Got Wired

Chapter 4 — The First Message

On the night of October 29, 1969, a 21-year-old student at the University of California in Los Angeles sat down at a computer terminal and tried to log in to another computer about 350 miles away.

His name was Charley Kline. He was working late — it was around ten o’clock at night. The lab around him was mostly empty. He had a phone in his hand. On the other end of the phone was a man named Bill Duvall, who was sitting in front of a different computer, at a research institute called SRI, near Stanford University.

Kline’s job, that night, was very simple. He was going to type the letters L-O-G-I-N on his terminal. The letters would travel — for the first time in history — out of his computer, across phone lines, through a brand-new piece of equipment about the size of a refrigerator, and onto Duvall’s screen in Stanford. Duvall would read the letters out loud, over the phone, so Kline could be sure they had arrived.

Kline typed L. Duvall said: I see the L.

Kline typed O. Duvall said: I see the O.

Kline typed G. Duvall said nothing.

The entire system had crashed.

For a few minutes, on the night of October 29, 1969, the very first message ever sent on the network that would become the internet was the word lo.

It was, accidentally, a beautiful first word. Lo — as in behold, or look. The very old way English speakers used to begin a story: Lo, I will tell you what happened in this place.

Kline did not write anything dramatic in the lab logbook that night. He didn’t realize he had done anything historic. He scribbled, at 10:30 p.m., a short entry that said only: Talked to SRI Host to Host.

Then he and Duvall fixed the bug. About an hour later, Kline tried again. This time login went all the way through.

This chapter is about how he got to that desk.

Larry Roberts Has a Plan

At the end of Chapter 2, I told you that ARPA hired a man named Larry Roberts to build the network J.C.R. Licklider had imagined. At the end of Chapter 3, I told you that Roberts had heard about packet switching at a conference in Tennessee and gone home with copies of Paul Baran’s eleven-volume report.

It is now early 1968. Roberts is sitting in an office in Washington, D.C., trying to figure out how to actually do this thing.

He had a plan. The plan was: start small. Connect four computers, at four universities, in four different states. Make them able to send messages to each other. See what happens.

The four universities he picked were the University of California in Los Angeles (UCLA); the Stanford Research Institute (SRI), near San Francisco; the University of California in Santa Barbara; and the University of Utah. Three were in California. One was in the mountains. They had nothing in particular in common, except that ARPA was already paying for computer research at all four of them, and they were all willing to try something strange.

Roberts wrote up the plan and sent it out to companies, asking them to bid on building the equipment. This is how the American government often funds technology — it doesn’t build the equipment itself, it pays a private company to build it.

The big famous companies — IBM, AT&T, the giants — looked at Roberts’s request and either turned it down or proposed solutions that wouldn’t have worked. They didn’t really think it could be done. (Remember AT&T from the last chapter. The pattern repeats.)

The contract eventually went to a small company in Cambridge, Massachusetts, called BBN — Bolt, Beranek and Newman. BBN was not a famous company. Most people had never heard of it. But it had something the big companies didn’t have: a few brilliant young engineers, and the willingness to attempt something almost nobody believed could work.

All 24 chapters of How the World Got Wired · Library · All courses