A Language Owned by No One — The Idea That Built the Internet

The reason the internet is the ‘internet’ is not one magnificent computer, nor a single thick cable. It is an idea. A common language, owned by no one, through which different parties may converse as equals. How was this language born, and why did it spread out across national borders and beyond corporations alike? We trace the decisive invention by which the four-node net of the previous chapter was transformed into a ‘net of nets’ spanning the whole earth.

The Hard Problem of Spanning Nets

In the 1970s, networks other than ARPANET were springing up across the world. Packet radio nets that flung data through the air, satellite nets that relayed by way of orbit, the local nets of companies and universities. Yet each had its own specifications, and none could make itself understood to the others. It was rather like a forest of tribes, each speaking a different tongue.

The men who took up this wall were Vint Cerf of Stanford University and Bob Kahn of ARPA. What the two set out to do from 1973 was to conceal the differences among individual networks and drape a common layer over the top of them all. In May 1974, they published the fruit of that work as a paper. Different nets would be relayed by a ‘gateway’ (the later router) and treated as a single virtual network — this design philosophy of ‘internetworking’ was itself the very root of the word internet. On 22 November 1977, they succeeded in demonstrating a communication that spanned three different nets at once: ARPANET, packet radio, and satellite. The net of nets was no longer a dream.

The Net May Be ‘Dumb’ — The End-to-End Idea

Buried within Cerf and Kahn’s design was a philosophy that, looking back, was revolutionary. The end-to-end principle.

The thinking runs like this. Do not make the network itself do complex work. The net’s role is merely to carry packets. Nothing more. Whether the data arrived intact, whether the order is correct — all such ‘clever’ processing is left entirely to the edges of the net, that is, to the hosts that do the communicating. The net keeps to being a faithful but simple courier. The prototype of this idea had already been implemented earlier in CYCLADES, the net led by Davies and by France’s Louis Pouzin.

Why does this matter? If you keep the net simple, then no matter who builds what new service on top of it, there is no need to rebuild the net itself. Email, the Web, video, social media — the net carries them while knowing nothing about them at all. That is precisely why new inventions arose one after another, requiring no one’s permission. Not concentrating the ‘cleverness’ at the center is what made the internet endlessly extensible. And this mechanism, at first a single whole, was split in 1978 into IP (the Internet Protocol), which does the carrying, and TCP (the Transmission Control Protocol), which bears responsibility for reliability — completing the ‘TCP/IP’ that endures to this day.

RFC — A Custom Owned by No One

There is one more culture that decided the character of the internet. RFC (Request for Comments) — a plain request for opinions.

In 1969, Steve Crocker, a graduate student at UCLA, wrote the first memo setting down a networking arrangement as RFC 1. Ever since, the technical standards of the internet have developed through the accumulation of these open memos that anyone may read and write. Not command, but proposal. Not authority, but consensus. The name of Jon Postel, who served as the first editor, is still spoken of as a symbol of this culture.

And what is astonishing is the very nature of the IETF (the Internet Engineering Task Force), which still administers these standards today. The IETF has no legal personhood. There is no governing board of directors. There are no membership fees, and not even any formal ‘members.’ There are only volunteers gathered from across the world, mailing lists, and discussion aimed at consensus. The foundation of the internet is upheld not by any nation nor any corporation, but by this open, ownerless custom.

But this common language, owned by no one, still belonged to a handful of engineers and researchers. The one who opened it to ordinary people the world over was a single British scientist, through a peculiar decision of his. He invented a ‘way to link documents together,’ and then — he gave it away, for free, to the whole world. The story continues on to the day the Web opened the world.