Net Visionaries Seek New Vistas
By Mark K. Anderson


2:00 a.m. Aug. 12, 2002 PDT

http://www.wired.com/news/print/0,1294,54416,00.html


In some fields, getting a blank stare means you're on the right track.

When they invented the Internet, blank stares were commonplace. And now Vinton Cerf (co-designer of the TCP/IP protocol upon which the Internet is built) and Tim Berners-Lee (inventor of the Worldwide Web) are each generating the now-familiar quizzical expressions as they promote their latest vision of tomorrow's computer networks.

One is currently mapping out plans for an Interplanetary Internet; the other is drawing upon the Web's potential as a worldwide data bank for intelligent agents.

Both men were honorees at this year's Telluride Tech Festival in Colorado, which ended Sunday. Both are innovators whose contributions to society will be greater still if their current visions pan out.

Given their track record, they probably will. Telluride, home to the world's first commercial alternating current (AC) power plant in 1891, has a tradition of hearing out such visionaries as Berners-Lee and Cerf -- a tradition the festival's organizers proudly proclaim.

Cerf, who spelled out the latest developments in a new-generation Internet that would stretch deep into the solar system, said that an interplanetary network would not just be relevant for space exploration.

Consider the time lag problem, for instance. Transmissions from Mars can take as long as 20 minutes to reach Earth. And if the antenna is on the far side of Mars, it can take hours or days before the planet -- or the orbiter -- brings the antenna around into a line of sight with the Earth.

In such an environment, the Internet as it's now conceived would shut down. Servers would time out long before any packets of data could be exchanged.

"The system has to be capable of dealing with delay and frequent disconnection," Cerf said. "As a result, we've now concluded that the interplanetary network is an example of a much more general concept we call delay-tolerant networks.

"Some of the ideas we're pursuing will have utility on Earth in the mobile environment, where connectivity is often episodic and the data rates are often low and sometimes asymmetric."

Trent Hein, CEO of the network consulting company Applied Trust, said he's confident that Cerf's vision will once again become universally implemented reality.

"It's evolving a lot like the way that the early Internet did," he said. "In the '70s, the Internet existed -- but not a lot of people used it or knew about it. They were using extremely slow hardware and only had a few nodes. But you just have to start with ideas and tweak them and then grow it into something big."

Cerf said that the Interplanetary Protocols -- "very much like e-mail, a store-and-forward design" -- just completed their third iteration last month. (The current Internet protocols had four iterations before they were standardized in 1978.)

The design team is now angling to use next year's launch of two Mars rovers as a testbed for the Interplanetary Internet. And somewhere between 2005 and 2007, a Mars observing satellite is scheduled to launch -- which Cerf hopes will also be a new node in what they call the "InterPlaNet."

John Perry Barlow, vice chairman of the Electronic Frontier Foundation, said the conceptual boundaries the Interplanetary Internet breaks are important too.

"Probably the best thing is it makes you think big," Barlow said. "I think it's pretty unlikely that we're going to be getting e-mail from Mars in the near future."

Back on Earth, Berners-Lee has no less ambitious designs for the future of the Worldwide -- and someday, solar system-wide -- Web.

His vision of what he calls the Semantic Web is of an Internet full of data meaningful to computers as well as to humans.

So instead of Vinny's Florists having a Web page that simply lists the store's hours, its contact info and the floral arrangements it sells -- comprehensible only to the humans who visit it -- Vinny's Semantic website would also contain metadata that search engines could understand specifying that the business is a florist and its hours are X, its location is Y and the products it sells are Z.

This data would allow precise searches fulfilling precise needs, which are now impossible.

"You'll tell a search engine, 'Find me someplace where the weather is currently rainy and it's within a hundred miles of such and such a city,'" Berners-Lee told the Boston Globe in June.

"Today a search engine can go and find you pages with those words on them, but with the Semantic Web, it will come back and say, 'Look, I found this place and I can prove to you why I know that it's raining and why I know it's within a hundred miles of this place.' So you'll be dealing with much firmer information."

Barlow said Berners-Lee's Semantic Web makes an appealing next-generation vision.

"Part of what Tim is trying to do is open the Internet up to different forms of human communication that are much less constrained," he said.

Barlow also lauded both Cerf and Berners-Lee for developing an Internet whose social value endures, even as its commercial value ebbs and flows.

"We're really lucky that the folks that happened to be in that highly pivotal position were also socially very aware and men of good conscience," Barlow said. "Vint, for instance, is one of the sweetest human beings I've ever met. His essential good-heartedness has been embedded in the culture of the Internet."