Chatting with Ben Segal - Podcast Episode 11 - The Rocky Beginnings of TCP/IP in Europe


Episode 11 - The Rocky Beginnings of TCP/IP in Europe

Ben: [00:00:00] For instance, when the Cray came, when the Cray came in '86, '87, we had to sign this very rigorous license with the Americans, with both the Department of Commerce and the Department of Defense, that we would secure this machine. Okay, so we did use three-level security. The lowest level was: I put the Cray on a secure Ethernet with IP filtering, which I bought from Cisco, which had 20 people working for them at that time. And I'd met the guy from Cisco at a USENIX exhibition in San Francisco..

David: [00:00:35] Cisco had 20 employees?

Ben: [00:00:36] Twenty employees in 1986 - '85 probably I met him then, '85 - and I saw .. the guy came from Stanford, Len Bosack he was called. And I was interested and I went to talk to him. "What are you selling?" "Ah, selling a router". "Oh yeah?" - because in those days IP routing was done in minicomputers, inside your own computer, OK.

David: [00:01:00] Right.

Ben: [00:01:00] Okay. You didn't have a dedicated router, it wasn't necessary. People, you know, they passed the packets. So, very interesting - and he gave me the brochure about it. And this was called "cisco", (San Fran-cisco). And I took that brochure home and I remembered. And it had IP filtering. And at home we were thinking about the Cray and I thought surely, we need - all we do is put this on a filtered Ethernet. And so, you know, only certain people can get into the network. And it was very nicely done: you configured the thing, you know, it could .. you could go out to anywhere and only certain address groups and certain ports, you know, could come in, and so on. So this was wonderful. And I bought the first two Cisco boxes probably in Europe.

David: [00:01:54] Wow.

Ben: [00:01:55] Again, I had a problem with CERN, because CERN was thinking about routing and they had decided they were going to buy their routers from a company called Bridge. I think it was Bridge - anyway it wasn't Cisco. Cisco was American, it was too small, you hadn't heard of them. So .. but I wanted that because it had the filtering. So I managed to buy them in the end.

Ben: [00:02:15] Another story which goes from that is that some years later, a guy comes to see me from Amsterdam and says: "Ben, I'm looking for a router that has X.25 capacity on it so that we can route across X.25". And I pulled my Cisco catalogue off the shelf and said: "Look, they have X.25" - because the US military actually needed X.25

    even in the States there was discussion going on about whether we should use the Internet. So this guy then went back: he was running the main gateway of USENET in Europe, in Amsterdam, and his idea had been subversively to convert from uucp protocol .. You know what was uucp?

David: [00:03:06] Unix to Unix copy.

Ben: [00:03:08] .. to IP. His friends, he had about five or six important friends who were running the nodes in each country of USENET. And all he had to do was give them a Cisco box each, they could interconnect across X.25, which was allowed, you know, so that they could use that, they wouldn't have to use dial-up any more. And away they went. And they did that without telling the PTT's who were ferociously against it. It turned out the Germans, who were the strongest against it, were not actually monitoring what was going over X.25. Because the Germans had threatened: "If you use TCP/IP we'll cut you off" etc.. It was really, it was hardball politics. This was '88 or so. And so I introduced him to the Cisco salesman who rushed off to Holland and sold them the boxes. And that's how the Internet started in Europe without people noticing.

David: [00:04:11] Essentially, you were .. if you ask for forgiveness, not permission, essentially..

Ben: [00:04:16] Absolutely!

David: [00:04:17] .. you would have never been able to get the permission, so you just went away with it and hoped for the best.

Ben: [00:04:21] So that's what he did. Also, it was the Unix club remember also?

David: [00:04:25] Yeah.

Ben: [00:04:25] Because Unix was important because Unix was there, it was a good technology and Unix had - Berkeley Unix - had TCP/IP in it.

David: [00:04:34] Yeah. Yeah. Yeah.

Ben: [00:04:35] So this is what helped. It was a sort of symbiosis. Whereas Unix at CERN was still a curiosity. OK, we'd put it on the Cray. That was a momentous decision, that again. That was my boss, Les Robertson, who'd pushed that.

Other news