navigationTo Federal Internet Guide Home Page

A Pentagon Plan Became the Internet

By Mike Mills
Washington Post Staff Writer
Tuesday, July 2, 1996; Page A06

The world is tied together by the Internet and the World Wide Web today in part because three decades ago, policymakers in the Pentagon got worried: Would the United States' military communications systems survive a nuclear attack or sabotage?

So experts were called together to come up with a system that would allay those fears. They began by junking then-prevailing design concepts of establishing a single route to get information from one point to another. Instead they conceived of many different routes between points, with information being broken up into small "packets" that would find their way to the destination independently, then be reassembled.

If one route was cut, a packet would automatically seek out and travel by another.

Vint Cerf, who as a leader of the original development team often is called the father of the Internet, compares it to sending a novel through the mail on hundreds of postcards. Each postcard is addressed, crammed with as many words as it can hold and numbered to aid reassembly at the other end.

In the same way, each Internet packet contains part of the message, is numbered to keep it in order, and bears an address as it sets out across the network. When the packet reaches an electronic fork in the road, "router" computers look at the address and send the packet toward its destination. If one route is closed or congested, it goes by another.

In 1969, the Pentagon tried out this approach in an experimental network called Arpanet. At first, it linked just a few computers at defense labs and universities. It turned out to be an economical, reliable system for moving information in large quantities.

By the mid-'80s, it had blossomed into a broad network for all stripes of educators, researchers and government workers. Many used it to fire electronic mail messages back and forth. Others sent complex documents of hundreds of pages, or put their own research findings "on the net" so that anyone could retrieve them for display on their own computers.

It was now called the Internet. Private users were getting on as well, through accounts at small companies that set up to sell links into the network. It spread overseas too, first to the industrial countries, then to most every country in the world.

No one owns or controls the Internet. It's more like a co-op. Again, a postal analogy is apt. Postal authorities in different countries agree on certain rules for exchanging letters. They will bear stamps of such and such value, the addresses will be in English, they will not be larger than a certain size, etc.

In the same way, the Internet is simply a collection of local computer networks that have agreed on how to exchange electronic information. Messages will be addressed in a certain way, so that everyone's computers will understand and deliver them. They will not be larger than a certain size, and so on. And, like the postal system, the Internet is only as fast as its slowest link.

The network is financed through a complex food chain of payments. Individual users pay money to their Internet access providers, which in turn pay money to regional networks to link to them. The regional networks then pay money to companies that operate "backbones," high-capacity circuits that cross countries and continents.

As the network has grown, so have problems of finding information on it. Search programs with geeky names like "archie" and "gopher" were developed.

But the network wasn't keeping up with the color and ease of use that was taking hold in stand-alone computing. While PCs were moving away from typed commands and going graphical, displaying bright colors, responding to the pressing of electronic buttons, the Internet was still trapped in the gray world of text.

Getting to different information resources on the Internet generally required users to type in complex commands, to connect laboriously to one place, then back out and begin a similarly tedious process to connect to another.

In the '90s, the Web changed that. It put a colorful, even sometimes psychedelic, cloak over the Internet. It allowed people to put "links" into documents, so that by clicking on a highlighted word, they could move directly to another "site" containing related information. More and more, packets moving across the network contained not text but photographs or sounds or colorful graphics or software.

Beyond smells, tastes and tactile sensations, there really is no form of information that the Web cannot carry. Today the Web is limited -- by capacity, by a shortage of software that takes full advantage of its potential and by the comparatively small number of people who have access. But one thing is certain: There is no telling where its growth will stop.

© Copyright 1996 The Washington Post Company

Back to the top

Go to Federal Page | Go to Government | Go to Politics Section

Home page Site Index Search Help! Home page Site Index Search Help!