Civilizing Internet

Internet could be the US public computer network

By David Churbuck

July 8, 1991

ANY TELEPHONE USER in the developed world can talk to any other. Not so the computer users of the world. The closest thing they have to a universal network is Internet, and it has but 3 million users in 33 countries.

Internet? Most people heard about it for the first time in 1988, when Robert Tappan Morris Jr., a collegiate computer hacker, clogged it with a devious "worm" program. The network he brought down was portrayed in the press as a dark, technie labyrinth of computer scientists who could understand computer commands like GREP! and could address messages to places like This image is not far off the mark. John Perry Barlow, a hackers' advocate (he is cofounder of the Electronic Frontier Foundation) and Grateful Dead lyricist, describes Internet's commands as a "savage user interface."

Internet has other problems. It is a communications weakling, capable now of moving only 1.5 million bits per second between 13 hub cities. That capability is expected to rise to 45 million bits by this fall, which sounds like a lot, but would accommodate only two color television signals. Also, Internet, a loose affiliation of private, academic and government-supported networks, is rather anarchic, without a single owner to police its lines and keep it running smoothly.

Yet Internet has enormous potential. It is the place where electronic mail was born, and from it have sprung most of the de facto computer networking standards used commercially and abroad. If Internet fulfilled the computer network ideal, computer users the world over could send messages, data files, pictures, sound, software and video clips to one another, secure in the knowledge that what they sent would arrive at the intended destinations.

If nothing else, a truly universal computer network would cut the telephone, fax and postal costs of its users. That's because computerized memos are electronically much leaner than voices or pictures. In theory, you should be able to send a 200-word memo across the country on a computer network for less than one cent. If you pick up the phone instead or use a 29-cent stamp, it's probably because your intended recipient isn't on any electronic mail system you can get to easily.

Internet, which traces its origins to a network set up in 1969 by the Defense Department, runs off $20 million a year in federal subsidies covering a high-speed cross-country link managed by the National Science Foundation (see chart). Supposedly, not anyone can tap in. Blatant commercial traffic, for instance, is banned, although defining what is and what is not commercial these days is not easy. But once you get on the system, Internet charges you nothing for the time you use. The only cost is the mental investment involved in navigating blindly through a network with no central administrator and several disjointed printed manuals that are not readily available.

Not for much longer. The National Science Foundation wants the regional networks that hang off of its backbone to start planning for a day when the government's full subsidy of the service will be replaced with a partial subsidy only. The magic word, according to Stephen Wolff, director of an NSF division devoted to networking and communications research, is privatization.

Already there is a tinge of privatization--and even commercialization, if you will. Where once the only way onto Internet was through a college account or via certain research institutions, today anyone with a computer and modem can get on the system by paying one of a half-dozen companies for an Internet access account. For instance, Software Tool & Die, in Brookline, Mass., charges $5 a month plus $2 an hour for modem access to the network. Other access sellers include UUnet, an offshoot of a seismic research laboratory in Reston, Va.; Cerfnet, spun off from a General Atomic contract to manage the San Diego Computer Center; and so on.

Demand is strong. Those college hackers who grew to love Internet may have moved on to the real world and don't want to give up their academic toy. But by far the most compelling reason to pay for access to Internet is its sheer size. There is a far higher chance of finding an electronic mailbox there than on any purely private network such as MCI Mail, AT&T EasyLink or CompuServe.

"The utility of a network increases as the number of users increases," says Wolff of the National Science Foundation. "The phone was no good if you could only call three people. This is a vision that captivates all of us who work with it."

The visionaries include Lotus Development founder Mitch Kapor, who foresees a world in which Internet, connecting millions of dissimilar computers, becomes the prototype of a national public network. But a lot has to be overcome before that nirvana is achieved. For starters, there is no directory. Telephone users dial directory assistance; on MCI Mail, a computer will attempt to find a match for any last name you type in. On Internet, you have to know the address of the recipient before you start.

A more troubling issue is who will manage the network. The present grassroots policing is full of flaws, particularly security flaws, as the Morris worm so graphically demonstrated to system operators.

"Internet has been like agar in a petri dish for networking technology," says Geoffrey Goodfellow, a longtime user of Internet and founder of Anterior Technology in Menlo Park, Calif. "I'm a product of that incubator, and I'm sure there will be a lot more commercial applications to come." Anterior Technology's latest venture is electronic-mail pagers, which use wide-area paging frequencies to enable users to read their electronic mail without ever tapping into the phone system.

The potential value in a universal computer network is evident. What remains to be seen is whether Internet can get its act together.

ILLUSTRATION: photograph - map CAPTION: Model for a universal network? (map)


Copyright Forbes Inc. 1991