Speaker by Various Artists

3

Vint Cerf: TCP/IP and me

by Cerf

The Fundy Post reminded me that I'd forgotten to write anything for the 25th anniversary of the TCP/IP cutover: day one of the proper internet, which was January 1. It's a bit late now, but I though it might be nice to haul out the interview I did with the so-called "father of the internet" Vint Cerf, shortly after the 20th anniversary in 2003.

It's fairly self-explanatory, but there is more background on what we're talking about here. You may feel free to share stories of how TCP/IP changed your life in the discussion for this post.

--

Even Vint Cerf can't quite understand why he didn't organise a party to mark the anniversary of the day he changed the world.

January 1 this year marked 20 years since the day of the TCP/IP cutover: "That was the day that the Internet was born in terms of becoming palpable, even though it had been an 18-year labour. And we didn't do anything for the anniversary. We all looked at each other and thought, well, that was silly …"

Cerf is, as the publicity for the recent Knowledge Wave conference promised, "the Father of the Internet". It's no mere flattery, and it's the reason young men seek his autograph as if he were some snowy-haired, even-tempered rockstar.

In 1983, Cerf and his friend and collaborator Bob Kahn were at the Advanced Research Projects Agency (Arpa), a military research group operated by the US Department of Defence. For more than nine years they had been in charge of the design of a new kind of network: one that had no single point of failure, that would allow physically different networks to connect - to internetwork.

Few inventions prosper the way theirs has. The TCP/IP protocols (see panel) they devised not only now join more than half a billion people on the Internet, they're basically the only game in data networking. TCP/IP, which nobody actually owns, has seen off the commercial offerings of large companies. It ploughed right on while the US government and the United Nations blessed a competing technology that never got off the ground. So, why did it win?

"Thank you for asking - because a lot of people think it just happened and that was it," smiles Cerf. "It wasn't a slamdunk."

First, Cerf and his friends had to convince the scientists. The Internet's predecessor, Arpanet, had spread to universities, usually when they took up research contracts with Arpa. The researchers liked it - so much so that as the first day of January approached, many of them saw no need to change to the new TCP/IP network. But Cerf held an advantage: the purse strings.

"We had a few people saying, the hell with you, and we said, fine, don't come around wanting research funding next year because you won't get any," says Cerf. "There was more than moral suasion involved."

The remaining TCP/IP refuseniks simply had the plug pulled on them that first day of 1983. There were other factors: Cerf and Kahn had decided "that we would put no restrictions or constraints on information about this protocol," it was public and open. They had smart people on their side, including the late Jon Postel, who having been dubbed "God", possibly outranks even Cerf in the firmament.

"Jon was just a phenomenal technologist. He had really good taste - good engineering judgement about the choices he made. And he held everyone to a very high standard. I wouldn't want to argue that I had any particular skill in that regard, but at least I was smart enough to know that I should call on the talents of some incredibly smart people."

This was Cerf's other great contribution - as not a joiner, but a former of committees. In 1979, he set up the network's first governing body: the Internet Configuration Control Board. That became the Internet Engineering Taskforce, which still gathers to debate and anoint new technologies. In 1992, he formed the Internet Society, becoming its first president and later chairman. He chairs the new Internet Societal Task Force.

Cerf was also in on the ground floor of Icann, the body set up to administer the Internet's names and addresses after a Clinton task force decided that government agencies should withdraw from administration of the now-global Internet in 1998. Icann has been enmeshed in controversy for five years, accused of being secretive, arbitrary and arrogant. The age of the benevolent dictators has passed, but it hasn't gone smoothly.

"Oh no!" Cerf agrees. "And I must say, some of your fellow countrymen have been quite vocal in their criticism of Icann. A lot of is probably deserved. When you start trying to do anything of the sort that Icann is doing, you're hit by a variety of forces and motivations."

Cerf's role at Icann (he became chairman last year) brought him into conflict with his friend John Gilmore, the libertarian "cyber-rights" campaigner and founder of the Electronic Freedom Frontier. Gilmore emailed Cerf a letter last year: "I have absolutely no idea what you are doing leading that megalomaniac, unaccountable, unresponsive, anti-expression, anti-public-interest organisation. Did they take your kids hostage?"

Cerf doesn't return the fire. He's a pragmatist, he says. If he and others hadn't stepped in and formed the new body, the Internet would have been vulnerable to "people who were thinking of monetising almost anything to do with the Internet."

Cerf also shies away from Gilmore's latest crusade: against the Bush government's rollback of electronic privacy, and in particular Admiral John Poindexter's Total Information Awareness project, which aims to capture, slice and dice almost every electronic trace Americans leave, as part of the fight against terrorism.

"Oh, I know John has gone berserk on this subject," says Cerf. "I think that's a huge overreaction. Remember that I used to work at Darpa (Arpa's current name) and I had a chance to talk to John Poindexter. What was going on there, honestly, was just a technology exploration: how could we take a lot of very diverse information and discover patterns in it?

"So even if the technology never got used for the purposes that have caused Gilmore and others to become very exercised, it could become very useful technology, in, for example in understanding how economies work, which is something we really don't understand. I'm a technologist at heart, and I'd hate to see the baby thrown out with the bathwater in this particular dispute."

The apparent lack of a reflexive concern for privacy and liberty from government might seem odd in an Internet luminary, but, remember, this is a man who did his best work for the US military. An establishment man. He does, however, ponder the social impact of the Internet and its many voices.

"That diversity of voices places a burden on you, because you have to be even more thoughtful about the significance of the input that you’re getting. Critical thinking becomes an even more important skill in this environment of so many sources of information. So one of the things I would look for in teaching children would be to do everything possible to help them understand that they really have to think about the information that they get, because they're going to get a lot of it and some of it's not going to be very good."

The technologist is, as ever, looking ahead. He recently began working at Nasa's jet propulsion laboratory on a new project, again funded by DARPA: Interplanetary Internet. It's tricky.

"The speed of light is part of the problem," he says. "It's too slow. It's so hard to realise that the speed of light is so slow that in the worst case, going from Earth to Mars takes 20 minutes. So all the intuitions that we had in building the network and all the protocols that go with it - you throw 'em out the window."

Russell Brown, 2003

3 responses to this post

Post your response…

Please sign in using your Public Address credentials…

Login

You may also create an account or retrieve your password.