Fernando Corbató: Time-Sharing Pioneer, Part 2

Abstract

Fernando Corbató: Time-Sharing Pioneer, Part 2 Dag Spicer (bio) Fernando Corbató is likely best known for his work developing time-sharing operating systems. Early in his career, Corbató had a pioneering role in the development of both the Compatible Time-Sharing System (CTSS) and Multics at the Massachusetts Institute of Technology. The second in a two-part series, the following interview is based on an oral history conducted by Steven Webber for the Computer History Museum in February 2006. (For a full transcript, see the CHM archive: http://archive.computerhistory.org/resources/text/Oral_History/Corbato_Fernando/Corbato_Fernando_1.oral_history.2006.102658041.pdf.) Steven Webber: Let’s get back to the summer of 1963. You mentioned the Project MAC summer program. What’s the actual timeframe on Project MAC? Fernando Corbató: J.C.R. Licklider, who had once been a professor at MIT, was at Lincoln Laboratories and he got his hands on some computers there. He was fascinated with the programming experience and ended up going to BBN (Bolt, Beranek and Newman), where he wanted to build a time-sharing system. A lot of efforts were ongoing to create time-sharing systems then, including the Dartmouth time-sharing system. The BBN system had a four-person team: John McCarthy was one of them, primarily as an adviser, and Ed Fredkin and Shelly Boilen, along with Licklider. They tried to create a simple time-sharing system based on the DEC PDP-1 computer, which was the first of the minicomputers. Today we’d call it a personal computer, but it was still a relatively big box. They got that system going in approximately the summer of 1962. For a long time, they bragged that they were first until I pointed out to them that we had demonstrated ours in 1961, but these are small details. McCarthy was promoting time-sharing along with Licklider at BBN, and Licklider then got recruited by ARPA to embark on a program encouraging “man-machine interaction.” The story I heard was that the Cuban Missile Crisis in 1961 had shocked the Pentagon because they discovered that a lot of their information systems were practically unusable. They couldn’t get information fast enough, and they were crashing. It was just a mess. They were so shocked that they had such poor man-machine interaction with their computers, their communication systems, their databases, and the like that they wanted to start research to improve on this and get better tools for the military’s purposes. So after Lick got recruited he began to travel around trying to encourage people to embark on time-sharing, and he came to MIT to see if anybody was interested. Webber: Was this after you had already demonstrated CTSS? Corbató: I’m not sure. We were certainly about to demonstrate, but the key thing is that Bob Fano, myself, Doug Ross, and about a half dozen or so of us who were participants in the computing activities at MIT were at this meeting with Licklider. Licklider started to ask us what we wanted, what we would think about doing, but it was like a catfight. Everybody had his or her own notion or plan or scheme. Licklider took it in stride, but we were dismayed by our behavior in front of this potential sponsor who wanted to give us money to do something. We were so disorganized. It was disgraceful. Fano came back from that meeting and recognized that there was a lack of leadership at MIT. He thought about it and subsequently had a long train ride with Lick coming back from a meeting in Hot Springs, Virginia. I think Fano would claim that he was persuaded to go ahead when he came back from that meeting. He decided that he would start to organize a venture that came to be called Project MAC. That formative stage was in the fall of 1962. Bob made the decision that we would use CTSS as a prototype for Project MAC to exhibit time-sharing, and we would do that by getting a carbon copy of the machine at the Computation Center. I think they placed the order just around the turn of the year, 1962...