While the internet is taking over the earth, Dr Vinton Gray Cerf - Vint, for short - is working to extend it to the rest of the solar system. And while Cerf is known for his sense of humour, this is not a joke. He's working with Nasa's Jet Propulsion Laboratory to make it a reality.
"We've designed a whole new set of protocols to augment the existing internet protocols, so we could effectively build a network of internets," says Cerf. "We put interplanetary gateways in orbit around the planets and satellites, and we run these new interplanetary networks through deep space. That's a very practical way of reducing costs in the interplanetary research effort.
"And the UK has a critical role in this," he adds. "It is going to be launching the STRV 1D research satellite in the middle of November, and it will be carrying our interplanetary protocols. We've just finished implementing the prototypes, and they work!
"Then there's a moon launch that is possible at the end of 2001, with a lander on it, and that's going to carry the interplanetary internet protocols. Then we are scheduled to be on two Nasa missions to Mars, to land on the planet.
"After that we have missions to launch satellites around Mars: those will become the backbone of a two-planet internet. From there we expand outwards: as each mission goes to the outer planets, they'll carry a little bit of the internet with them. It's like living in a science fiction novel! There's a practical reason for doing all this, but it's also just plain fun."
Being the father of the internet, founding president of the Internet Society, recipient of the US National Medal of Technology and other awards, and senior vice president of internet architecture and technology for WorldCom, has never stopped Cerf having fun. For example, when taking part in a "shoot-out" with Bob Metcalfe, the inventor of Ethernet networking, at an industry trade show in 1998, he donned a cowboy hat and a couple of (holstered) six-guns.
During a heated argument at an Internet Engineering Task Force meeting about the future of the Internet Protocol (IP), the basis of internet communications, in 1992, Cerf stripped down to a T-shirt bearing the words "IP on Everything". He's serious about what the slogan means, but he still chuckles about the pun. "Now everyone wants to put whatever they have on top of the internet, I should have a new T-shirt made that says 'IP Under Everything'," he says.
Not even posing for the Guardian's photographer, for this article, dented his good humour. When he saw I was suffering some strain from holding Frank Baron's light, he suggested that if I did a good job I might be promoted to a hat stand. Anyone with a fraction of Cerf's achievements could easily have become self-important, but he has down-to-earth ways of defusing tension and generating camaraderie that must have been priceless in the net's early days.
Which is not to say the early days have gone, because Cerf is still building the future. In the 1990s, for example, he was promoting the expansion of the number of internet addresses [IP numbers] available. When most people still didn't know what the internet was, he wanted an internet address for every electron on the planet. Now people are finally catching on to the idea of the internet being everywhere, he's pushing into space.
This is not just to solve Nasa's communications problems, but to answer the question Cerf asked himself: "What should I be doing now, that we'll need 20 years from now? One answer is that maybe some day we will need a commercial backbone that is interplanetary in scope, and we'd better be ready, technologically," he says. "We don't have to build all of it right away, but I want to know how to build it."
As he points out, the internet also started as a long-term research project with practical near-term goals. It began in 1969 when Arpanet, the Advanced Research Projects Agency network, was set up by the US Defence Department as a way of sharing expensive computing facilities between universities. Cerf helped to develop it when he was a graduate student at the University of California at Los Angeles, and then at Stanford University, after he joined the faculty in 1972.
The Arpanet was innovative, being based on a packet-switching system conceived earlier by Paul Baran (ie, it works like the postal service, except the packets are computer data rather than physical letters and parcels). But not everyone could be on the Arpanet, and there was the bigger problem of connecting different networks together. The world needed an internetworking system, or internet. That's what Cerf and Arpa's Bob Kahn described in a paper published in 1974, and their TCP/IP (Transmission Control Protocol/Internet Protocol) system made the internet what it is today.
Cerf joined Arpa in 1976 to manage the various research projects involved in implementing TCP/IP, and in 1983, it finally became mandatory as a way of moving internet data. This made Cerf, by popular acclaim, "the Father of the Internet".
Cerf has sometimes resisted what he calls "this figurehead notion", and he is always ready to put forward the names of other people who helped create the internet. "Things with this scale don't happen unless thousands of people are involved, and something that makes the internet valuable is that it makes it easy for people to contribute," he says. "That collaborative aspect of the net is what contributes to its power."
On the other hand, Cerf has also used his position to influence users, researchers, companies and governments, and at MCI WorldCom, where he works, to build high-speed internet backbones. Things that could have sounded outrageous coming from others were taken seriously by the US Defence Department, and by politicians such as Al Gore, because Cerf said them.
"Yes, there's a bully pulpit element to all this," Cerf says, borrowing a phrase President Theodore Roosevelt used about the White House. "For almost 30 years I've done everything I could to promote the internet. I feel simultaneously privileged to have had so many opportunities to speak to so many people about the internet, but what I honestly hope, and truly believe is the case, is that people have decided for themselves. If the internet is widely accepted, it's simply the recognition of an idea whose time has come. For that you can't take credit."
The IT industries are so quick to align behind winners that those not involved usually remain blissfully ignorant of the bloody battles that precede any standardisation. However, the internet's acceptance has been a long time coming, and it has been a struggle.
Cerf nods. "It's true, many people resisted it," he says. "Its predecessor, the Arpanet, was considered a silly idea that wouldn't work, and it was ridiculed by people who grew up in the telephone tradition. Most of the computer science communities also rejected the idea of connecting up, and Arpa had to insist.
"When it came time to convert from the old Arpanet protocols to the new internet protocols, in January 1983, there was tremendous resistance. Finally, we had to force it on people by turning off the old protocols, so we jammed it down their throats. Then came the lengthy debate between the OSI and internet protocols, and again, that was a 10-year battle."
Far from promoting the internet, most governments - including the US, the UK and the EC - adopted OSI (Open Systems Interconnect) networking in the 1980s, partly to block the advance of IBM's SNA (Systems Network Architecture). The far simpler IP, which had taken root in academe and the research community, was rarely considered, and almost always rejected for commercial use.
The irony is that not only has IP triumphed over rival networking systems, it is also being used to carry voice traffic (Online, October 26) by the telephone companies who used to disparage it.
"It is a delicious irony, but we certainly didn't set out to upset the applecart," says Cerf. "Looking back, we were so thoroughly rejected by the telephony community, we were left to our own resources to build our own private network, unregulated and unmanaged: we were outside their purview completely. But we could see the potential for upset, because we were experimenting with voice on the net as far back as 1975, when the backbone was only [running at] 50kbits/ sec. We used to transmit voice using a form of coding that made everyone sound like a drunken Norwegian: it was a lovely lilting sound.
"As far back as 1979, we were doing studies analysing which would be most cost-effective: two separate networks for voice and data, a hybrid network, or one which was all packets. Even then, the analysis said that an all-packet network, if it were big enough, would be more cost-effective. And guess what: that's what's starting to emerge!"
It has taken the world 20 years to catch up with Cerf, and it may be more than 20 before the interplanetary internet makes much of an impact. But if you start getting email from Martians, you'll know who to blame.