Friday, June 7, 2554

Errol Morris: my brother invented E-Mail with Tom Van Vleck? (Part 4)

Translate Request has too much data
Parameter name: request
Translate Request has too much data
Parameter name: request

This is part four of a five-part series. Parts 1 | 2 | 3 4.

THE FIG TREE AND THE WASP

From “Commercial Gardening Volume III”, John Weathers (editor), 1913

It could be a 20th century version of Genesis. In the beginning there was Robert Fano, Fernando Corbat? and J. C. R. Licklider. There were others, of course. Many others. John McCarthy and Edward Fredkin at a Cambridge consulting firm, BBN (Bolt Beranek & Newman) — Licklider was at MIT, then BBN, then at ARPA, and ultimately back at MIT — but MIT was the hub. A group of professors and engineering students wanted to change how computers were used and as a result changed the world. Licklider, by all accounts, was the visionary, the force behind it all. He was not only the oracle of the revolution, but he managed to put himself in a position in the government (at the newly created ARPA) where he could channel financing to projects of interest to him. And he wrote a number of legendary articles, where he foresaw the World Wide Web and personal computers. [23] He saw computers not just as adjunct to problem-solving, but — he took the term explicitly from biology — in a symbiotic relationship with man. His paper “Man-Computer Symbiosis” (from 1960) opens with what might be considered a metaphor — except that it isn’t one, it is an illustration of what the relationship between man and computer can become.

The fig tree is pollinated only by the insect Blastophaga grossorum. The larva of the insect lives in the ovary of the fig tree, and there it get its food. The tree and the insect are thus heavily interdependent: the tree cannot reproduce without the insect; the insect cannot eat without the tree; together, they constitute not only a viable but a productive and thriving partnership…

Man-computer symbiosis is a subclass of man-machine systems. There are many man-machine systems. At present, however, there are no man-computer symbioses… The hope is that, in not too many years, human brains and computing machines will be coupled together very tightly, and that the resulting partnership will think as no human brain has ever thought and process data in a way not approached by the information-handling machines we know today. [24]

Many of these early pioneers came from different fields. Licklider had come from psychology; Corbat? from physics; Fano from electrical engineering. [25]

I contacted Dr. Corbat?.

ERROL MORRIS: Tom Van Vleck urged me to talk to you about my brother and his contribution to e-mail.

FERNANDO CORBAT?: I don’t remember the technical details of who did what and so forth. But Noel was right in the middle of things — working on the inner machinery of the system — and was one of the really crack system programmers. He knew how to make things work, how to put them together. And could write really well. It was a big project. And a very tough one, because we were breaking new ground in all directions. But the core of it was the young people who believed in the idea and made it work.

Professor Fernando Corbat?, c. 1965MIT MuseumProfessor Fernando Corbat?, c. 1965MIT Museum

ERROL MORRIS: What was the idea?

FERNANDO CORBAT?: Back in the early ‘60s, computers were getting bigger. And were expensive. So people resorted to a scheme called batch processing. It was like taking your clothes to the laundromat. You’d take your job in, and leave it in the input bins. The staff people would prerecord it onto these magnetic tapes. The magnetic tapes would be run by the computer. And then, the output would be printed. This cycle would take at best, several hours, or at worst, 24 hours. And it was maddening, because when you’re working on a complicated program, you can make a trivial slip-up — you left out a comma or something — and the program would crash. It was maddening. People are not perfect. You would try very hard to be careful, but you didn’t always make it. You’d design a program. You’d program it. And then you’d have to debug it and get it to work right. A process that could take, literally, a week, weeks, months —

ERROL MORRIS: Yes. I remember my brother lugging around reams of paper, all of these computer print-outs. It seemed like a nightmare.

75 ThumbnailTry writing some code and then send an e-mail from 1965.

FERNANDO CORBAT?: Yes. People began to advocate a different tactic, which came to be called time-sharing. Take advantage of the speed of the computer and have people at typewriter-like terminals. In principle, it seemed like a good idea. It certainly seemed feasible. But no manufacturer knew how to do it. And the vendors were not terribly interested, because it was like suggesting to an automobile manufacturer that they go into the airplane business. It just was a new game. A group of us began to create experimental versions of time-sharing, to see if it was feasible. I was lucky enough to be in a position to try to do this at MIT. And we basically created the “Compatible Time Sharing System,” nicknamed CTSS from the initials, that worked on the large mainframes that IBM was producing. First it was going to be just a demo. And then, it kept escalating. Time-sharing caught the attention of a few visionary people, like Licklider, then at BBN, who picked up the mantle. He went to Washington to become part of one of the funding agencies, namely ARPA. ARPA has changed names back and forth from DARPA to ARPA. But it’s always the same thing.

ERROL MORRIS: They’re close.

FERNANDO CORBAT?: Yes, it is the same organization. They just keep periodically relabeling themselves. Sometimes “Defense” was the initial word and sometimes “Advanced Projects Research Agency.” Anyway, Licklider went down to DARPA [Defense Advanced Research Projects Agency] and proceeded to get in the position where he could go out and parcel out a lot of money, to try to encourage people to do good research. And, in particular, the one thing he was quite interested in was time-sharing and he came up to MIT and talked to a bunch of us. In those days, the MIT computation research was sprinkled in a lot of directions, a lot of different groups. And Bob Fano was one of the people that was involved in the discussions. He was a little bit like the elder statesman, even though he was just a few years older. But he took it upon himself to propose a new project, which came to be called Project MAC, which would collect together all these different groups who were interested in man-machine interaction.

ERROL MORRIS: What year was this?

FERNANDO CORBAT?: I believe it was ’63. I don’t have the dates firmly locked in my mind. But we had done a demo with four typewriters in November of 1961 and we kept improving it. Fano proposed this project, which we all kind of coalesced on, and by the end of 1962, he had the proposal written.

ERROL MORRIS: And the goal was —

FERNANDO CORBAT?: — to use CTSS as an initial platform and to explore the ramifications of man-machine interaction. It was also an opportunity to build a tool. [26]

Scientific American

ERROL MORRIS: From the beginning, did you have terminals hooked up to the mainframe? Was that something that happened early on?

FERNANDO CORBAT?: It happened fairly rapidly. During that summer of anticipation, we beefed up CTSS by acquiring an IBM terminal controller, which was a big box called a 7750 that allowed us to have a large number of terminals. Modems were just coming into play. We were able to attach modems to the 7750, and modems to the terminals, and interconnect the modems via phone lines. MIT, at that point, had its own switched telephone network and we were allowed to put our modems onto the MIT telephone network. So to connect a terminal, it was just a matter of dialing up from the terminal’s modem and being assigned to the next available modem at the 7750.

ERROL MORRIS: I spent a summer with my brother in 1969. And his apartment had a terminal.

FERNANDO CORBAT?: In ’69? Absolutely. We had a phone and a modem at each terminal. And we could dial up the computer. The speed wasn’t very high. The modems started out being 110 baud, which roughly is a dozen characters per second.

ERROL MORRIS: That’s incredibly slow —

FERNANDO CORBAT?: They were slow and as big as a shoebox. And it was really pretty awkward. We were lumbering along. But just being able to do it at all was a big step forward. The technology was just changing from vacuum tubes, the 709, to the transistor machine, the 7090. Project MAC was going to be several hundred people, and so we needed the space to start a project of that magnitude. Fano managed to work out the plan to get space over in Technology Square, which became Project MAC’s first home. Initially we had just two floors. One of the service bureaus had decided they’d over-expanded and wanted to give up the space they had in the building. We were able to get one floor, and then another floor opened up, and so forth.

ERROL MORRIS: Did many patents came out of this?

Cambridge Historical Commission

FERNANDO CORBAT?: We didn’t try to patent anything. The object was to influence the direction of the computer industry.

ERROL MORRIS: Which you did.

FERNANDO CORBAT?: — which we did. That was our goal. And to form a nucleus of computer science at MIT. Until then, computer specialists had been sprinkled around in various groups and places.

ERROL MORRIS: And still no computer science department —

FERNANDO CORBAT?: Most of the key leaders came out of the electrical engineering department. But electrical engineering at MIT was a peculiar department. It had a very, very broad view of what its charter was. It had everything, from biologists to material scientists — a rich tradition of being very spread out in its research directions. And so, most of the people that joined us, Project MAC, were part of electrical engineering, but not all. And Fano asked everyone to make their principal office in the same building where the computer was. This is turning into a longer story than I expected.

ERROL MORRIS: That’s fine.

FERNANDO CORBAT?: And we wanted to come up with hardware that was designed to be time-sharing from the word go, rather than designed for batch processing. So we shopped around. And we talked to about half a dozen different vendors. IBM thought they were the prime possibility, because we were using an IBM machine for CTSS. But we discovered afterwards that they had been working on the 360 line of computers — a very ambitious project in its own right — to produce a family of machines, small to large, which were all batch processors. Unfortunately, a lot of the architecture was against the grain of what we were trying to do with time-sharing. [27]

ERROL MORRIS: So you didn’t go with IBM —

FERNANDO CORBAT?: We formed a triumvirate. Bell Labs, MIT and the G.E. Computer Department agreed orally to work on this project. And each group had its own expectations of what they were going to get out of it. MIT was trying to influence the world. Bell Labs was trying to get a machine which was going to be a workhorse for their laboratories. And General Electric was hoping to get a brand new shiny product that they could sell to other customers. That was the genesis of Multics. We threw down the gauntlet and said, “This is what we’re going to do,” which was very unusual for software projects in those days.

ERROL MORRIS: Did you have any idea how difficult it would be to actually change how people used computers?

FERNANDO CORBAT?: No. We didn’t realize how difficult it was going to be. Once you start from scratch, you really have to do everything. And even though we had anywhere from 50 to 100 people at one time working on the project, it was still a tremendous effort to try to get everything right. It was a visionary goal that everyone who joined the project came to believe in. It wasn’t just another job. And that’s how we kept people like your brother interested as a career. He was part of a group of people that knew how to make things happen. It was an exciting time, and everybody was inspired because they felt they were working on something worthwhile, not just another programming job. Not just another product.

MIT Museum

ERROL MORRIS: But e-mail? Was it a big thing? Or was it just —

FERNANDO CORBAT?: E-mail just evolved. One of the first things you notice when you have everyone on the same machine is that it’s possible to share information back and forth. You can let others use your programs. You can send them data. You can send messages to one another. So early in the game with CTSS, we started up the ability to send messages to people that were logged in. I don’t remember exactly when we began to allow messages to be saved. But that was a natural consequence. And about the time that Multics was coming to life, the idea of networking also began to come to life.

ERROL MORRIS: Networking?! Did everything start at MIT?

FERNANDO CORBAT?: No, it started with Larry Roberts, then at ARPA, who proposed — with Licklider’s encouragement — what came to be called the ARPANET. [28] He went around to all of the sites that Licklider was funding and talked up the idea, “Wouldn’t it be great if we could all send programs back and forth to each other, and to even operate computers from one side of the country to another?”

ERROL MORRIS: But didn’t you have to deal with different kinds of computers?

FERNANDO CORBAT?: The ARPANET was complicated, because everybody in the country had different kinds of equipment. And so, part of creating the ARPANET was to come up with a plan to, homogeneously, allow people to talk to each other, even though they had different kinds of equipment. And Larry proposed a so-called interface message processor — an IMP — attached to what ever computer they were using. And the IMP smoothed out the fact that everybody had a different computer. [29] There were a lot of DEC computers, but there was a tremendous amount of variety. And that allowed the variety to be screened off, so people could communicate. Each of the sites that was interested got an IMP and programmed it to match the ARPANET, and then programmed the IMP to match whatever equipment they had. [30] At MIT we did that, too.

ERROL MORRIS: With so many innovations, what are you most proud of?

FERNANDO CORBAT?: Oh, the fact that it was a team effort. We were able demonstrate that interactive computing was not only a viable idea but a powerful one.

ERROL MORRIS: But when you first started —

MIT Museum

FERNANDO CORBAT?: It was very different. Batch programming software was not designed to be interactive. It was designed to dump out reams of paper, at best. It’s when you begin to get into the man-machine business, when we had first had time-sharing, that you designed different interfaces. You designed different ways to edit text. You designed different ways to write programs. It changed your strategy. All of that first began to show up on these mainframes that were being time-shared. And then, it migrated from the mainframes to the mini computers, which maybe had 10 or 15 people on them, and finally the personal computer, as the hardware kept allowing more new options.

ERROL MORRIS: Did you foresee any of this back in the early ‘60s?

FERNANDO CORBAT?: In a dim way. Not quite so dramatically. We didn’t quite know which shoe would fall next. For example, we knew, from the earliest days, that a graphical interface would be advantageous. But the economics were such you couldn’t afford it. It would cost $50,000 to create a terminal for just one person. It just was not feasible.

ERROL MORRIS: But didn’t you experiment with graphical interfaces, as well? I remember seeing one.

Scientific American

FERNANDO CORBAT?: Yes, we did have one attached to CTSS, which was called the Kluge, a very fancy graphical terminal. [31] [32]

But it wasn’t until the personal computer showed up that people began to really see the potential. There is a side bar on that. One of the reasons that IBM jumped into personal computers was that Apple II was so successful. But the reason the Apple II was so successful was not because it was such a great machine. It was because there was one program on it which suddenly made it of immense value to companies, namely VisiCalc, the first electronic spreadsheet. It was a runaway success. Apple doesn’t like to admit it, but that was the main reason why people were buying Apple IIs. And that’s what got IBM’s attention — suddenly the business world was finding something valuable that they weren’t building.

ERROL MORRIS: History is odd. It’s almost impossible to guess what will be saved and what will fall by the wayside. I’m indebted to Tom Van Vleck and glad that he has become an historian. Without him, I would never know about my brother’s work. I told him he should publish a book. And he said, “What would I want to do that for?” [laughter] He said, “If I put it online, I can constantly update it and correct it and improve it. If I write a book, it’s just a book.”

FERNANDO CORBAT?: Yes, yes. That’s the modern publishing conundrum, isn’t it?

ERROL MORRIS: I keep thinking of the crowd at my brother’s funeral. How did they all know? Do you remember how you found out?

FERNANDO CORBAT?: I don’t remember, but I remember it was a shock. The Multics group, by and large, was young. And death was not a common phenomenon. Very few people have died that worked on the project. That will change with time, of course.

ERROL MORRIS: Yes. Time catches up with all of us, eventually.

FERNANDO CORBAT?: But this project inspired people. They felt they were working on something that would last. It was a chance to make a mark on the industry and to change the way things happened.

ERROL MORRIS: To change the world.

FERNANDO CORBAT?: Yes. We were the forerunners of the way things went.

*******

I then learned that Robert Fano, Corbat?’s roommate and officemate, was still alive. And delighted to talk about the past.

Try writing some code and then send an e-mail from 1965. Go to the interactive feature ?

Follow Errol Morris at twitter.com/errolmorris or join him on Facebook.

*******

[23] Memorandum For Members and Affiliates of the Intergalactic Computer Network

[24] Man-Computer Symbiosis

[25] Oral history with Robert Fano from the Charles Babbage Institute, University of Minnesota: “It was around 1960. Let me see if I can remember the exact date. There was a great deal of concern on the part of the administration about computer facilities at MIT. So they appointed a committee to make a recommendation consisting of Al Hill, Phil Morse, and myself. Various records also show Jerry Wiesner [Pres. Kennedy’s science adviser and, eventually, president of MIT] was a member. But I do not recall — he probably left for Washington. But he did not really play a part. We promptly appointed a technical committee, which consisted of the various computer types at MIT: Corby, and John McCarthy, and Minsky, and Doug Ross, and Jack Dennis were there. Herb Teager was chairman. Now by that time the idea of timesharing was already moving, and they proposed the development at MIT of a timesharing system.”

[26] In an interview years earlier, Corbat? said, “So that was mostly to convince the skeptics that it was not an impossible task, and also, to get people to get a feel for interactive computing. It was amazing to me, and it is still amazing, that people could not imagine what the psychological difference would be to have an interactive terminal. You can talk about it on a blackboard until you are blue in the face, and people would say, ‘Oh, yes, but why do you need that?’ You know, we used to try to think of all these analogies, like describing it in terms of the difference between mailing a letter to your mother and getting [her] on the telephone. To this day I can still remember people only realizing when they saw a real demo, say, ‘Hey, it talks back. Wow! You just type that and you got an answer.’”

[27] Fernando Corbat?, “IBM was trying to humor us. They viewed it as kind of an interesting experiment. Unbeknownst to us in this whole time frame, they were hard at work developing what came to be known as the 360 line of machines and they had it close to their vests. The people that had sold the management of IBM on the 360 had convinced them that it was a machine to solve all problems from here to forever after. That wasn’t true. Clearly, timesharing was not on their agenda and they viewed it as an aberration. They saw the goal as to have a line of machines which allowed them to sell much like the auto manufacturers. You would have a starter machine and then a larger machine and then a larger machine. They saw it only as a way to sell machines to customers. They did not think of it in terms of the way people would use computers.”

[28] In his ARPA Program Plan of 1968, Lawrence Roberts wrote, “Just as time-shared computer systems have permitted groups of hundreds of individual users to share hardware and software resources with one another, networks connecting dozens of such systems will permit resource sharing between thousands of users.”

[29] IMPs were the forerunners of today’s routers.

[30] Interface Message Processor (IMP) 1969

[31] “Cyrus Levinthal, the Kluge and the origins of interactive molecular graphics” (pdf)

[32] Early Interactive Molecular Graphics at MIT


View the original article here

No comments:

Post a Comment