Well, I didn’t know that I was supposed to say anything, so I won’t say much. But the way it all started was in 1964 I was at a conference in the homestead in Virginia. And Lick and I were sitting around talking with Corbató, who was running the time‐sharing system at MIT. And Lick was talking about what do we need next? And I was very interested in that. And so he was saying, “Well, we need a network to tie the machines together. We need some way to have them communicate.”
And I had seen many problems of this kind already by trying to share information that I’d put on a TX‐2, like the huge photo library that Minsky wanted down at MIT. But we were totally incompatible. And I had to build and buy an IBM tape drive and tie it to the TX‐2 to write a tape for him. And write the programs to try and figure out how to convert it to IBM tape. It was not fun. And it took a long time to get the data to him. But now that problem has totally changed. Now it happens instantly, all the time.
We have two big problems ahead of us is that I really worry about and want to just mention briefly. And that is, we know now a lot more about quality of service and that we could improve it dramatically on the Internet. And we should be doing that, but we’re not. We’re not really undertaking that issue very seriously, and we could. It will take some small changes to the network, but they’re going on all the time. And we’re making a huge change to v6 at the moment, so there’s lots of things are changing.
And we also need to pay a lot more attention to security, because I believe the networks will be doing the security work, not the computers. The computers are totally vulnerable and will always be that way. We’re losing the game there. They’re finding bugs faster than we fix them. Whereas in the network, we see all the traffic and all of bots and all the Denial of Service traffic, and everything is there. And a technique that was done at SRI proved to me that if you look at that communications in their bot hunter which they put on a computer to look at what you’re communicating, they can find the bot in your computer because of looking at its communications.
Well, in the network we could see that, too. We could find all the bots and all the bot communication, and we could find a denial of service way before it hit anybody. We could do a lot better than we’re doing, and we could add a lot more security to the network with some protocols that I’ve just put through the ITU and the TIA. So I just recommend that people start thinking about those two issues, because we don’t have long on the security issue before it blows up. Thank you.