Programmers At Work

Dan Bricklin – 1986

Dan Bricklin

Philadelphia native Dan Bricklin was born on July 16, 1951. He graduated from MIT in 1973 with a bachelor’s degree in electrical engineering and computer science, then worked as a programmer for Digital Equipment Corporation and Fas Fax Corporation before entering graduate school in business at Harvard University. While at Harvard, he designed an electronic spreadsheet program, drawing upon the expertise and advice of his classmates and professors.

In 1978, while still at Harvard, he teamed up with Bob Frankston, an old classmate from MIT, to develop a workable version of the program. This became VisiCalc. They formed a company called Software Arts, which they incorporated in January of 1979, and in April of the same year they signed a contract with Personal Software, the company that was to market the VisiCalc program. (Personal Software eventually became VisiCorp.) The news of VisiCalc spread rapidly. By May 1981, VisiCalc sales had exceeded 100,000 units. In 1983, cumulative sales topped 500,000 units. The success of Software Arts continued until 1984 when it entered into an extended legal battle with VisiCorp over the rights to VisiCalc.

In May 1985, Dan Bricklin left Software Arts to be a consultant at Lotus for a short time. Bricklin has since founded Software Garden, his newest venture, which was officially incorporated in November 1985. His first product is named “Dan Bricklin’s Demo Program.”

I met with Dan Bricklin at his suburban Boston home. He works out of an extra bedroom, which serve as the office for his new, one-person company, Software Garden. With a new philosophy stemming from his experience with the rise and fall of Software Arts, Dan Bricklin is starting over. This time it is not with the innocence, unleashed enthusiasm, and incredible energy of his first venture, but rather with a certain degree of caution, insight, and control. As Bricklin says, his intent is not to run a ranch or tend a farm, but to cultivate a garden of software, just like a garden in the backyard; one that is sufficient to meet his needs and from which he can gain pleasure and satisfaction.

It is evident that Bricklin’s past experience with VisiCalc has weighed heavily upon him. He is a soft-spoken, easygoing, intelligent man. In a careful, reflective manner, Bricklin discussed his experience developing the VisiCalc spreadsheet program and running Software Arts, and shared his ideas of what he looks forward to achieving with his new program and his new company.

INTERVIEWER: During the time you were studying at MIT, how were you involved with computers?

BRICKLIN: In early 1970 I went off to MIT, and did computer-oriented work. I got a job at Project MAC, which is now known as the Laboratory for Computer Science. There I met Bob Frankston, David Reed, and other programmers who went on to do lots of good stuff.

I programmed there for a long time, all the way through undergraduate school. My first project was a calculator. It surprised me that the time-sharing system at MIT, Multics, did not have a command-line calculator where you could say, “Calc 2 plus 2,” and it says, “4,” or sine of X, or whatever, so I worked on one. An AI group upstairs was working on LISP. There was a wide variety of people at MIT, like Richard Stallman, the famous hacker. Also, MIT exposed me to some real, experienced professionals in many fields.

INTERVIEWER: What other projects did you work on at MIT?

BRICKLIN: I did a wide variety of things. I was a member of the team implementing APL, and by 1973 I was running the project that implemented APL. I was also involved in the LISP project.

In the fall of 1973 I took a job at Digital Equipment Corporation in Typeset-10, their computerized typesetting group. I had originally been offered a job in their languages group; my training was in languages. But during my interview at Digital, I ran into a person I had worked with at MIT, Michael Spier, who suggested I look at the typesetting group. My father was a printer and his father was a printer before him, so I found the typesetting group more intriguing than languages. Typesetting was more of a real-world application. We had video screens and computerized typesetters.

INTERVIEWER: What did you do for the Typeset-10 group?

BRICKLIN: The first thing I did was wire-service translation programs. The wires coming in would be translated into a language the computer could understand for its typesetting. I learned a lot about real-world systems on this job. If the program had a bug and caused printing to be late, the newspaper would be late, costing them a lot of money, and I’d hear about it. We were working to deadlines, under a lot of pressure.

The people at the newspaper just wanted to do their jobs. They were not interested in the technology. As long as the technology worked, they were satisfied and they didn’t think about it. So I learned about how non-technical people perceive technology.

I remember when one newspaper advertised on the radio that they were going to have the transcripts of the Watergate tapes available, and the Watergate tapes were coming over the wire in a certain format that the program didn’t know how to handle. To get it out in time, we had to modify the program on the fly. At the same time, we were reading the tapes coming off the wire for the very first time. I remember that very well. I was actually flying to another newspaper when that occurred and I was called back because of the problem. It was real exciting. All I had was a toothbrush in my pocket when I returned, because all my luggage had gone off to Canada.

The newspaper was a real-world system, with immense pressures, compared to being in a compiler development group or back at MIT, where people are very tolerant of problems and say, “We’ll fix it next month.”

Something else that made an impression on me was the realization that real users work differently than programmers. Whenever something would slow down, the people at the newspaper would arbitrarily give it a higher priority on the schedule, which would screw up the whole scheduling algorithm on the computer. And they wondered why things were “funny.”

For example, they used a Model 33 teletype for the system console, and once a little bit of paper got stuck in the optical-character reader that was reading the typed copy. So, because of this little piece of paper, there were just a few more errors per page that the typesetting program would find and it would say “bad format.” But all those messages were on the operator console, and the operator console got behind because it was a cheap, slow 33 teletype. The whole system was waiting for that machine to catch up.

So there they were, with this million-dollar computer system, and the entire newspaper ran late just because of this one little piece of paper, and because they had a slow 33 teletype. Then they went off and got themselves a thirty-character-per-second terminal a day or two later, and everything was fine.

INTERVIEWER: What came after the Typeset-10 project was finished?

BRICKLIN: The next thing I did was to get involved in the first word-processing effort at DEC. I had already done a typesetting terminal program for DEC, and then I went into word processing after having done some software work on another typesetting terminal. I got to microcode at a very low level, and I actually wrote the first pass of the microcode for a machine that had 512 bytes of memory. It was going to be burnt into ROM. Then I went into PDP-8s and word processing.

INTERVIEWER: What kind of features did this word processor have?

BRICKLIN: We had very strong limitations on our hardware. Basically, we had to use standard Digital hardware. We were able to get a minor modification to one of the Digital terminals to allow it to scroll down as well as up, but that was about it. We couldn’t use any special hardware for word processing like most other companies ended up doing. I was the project leader for a group that designed a word-processing system. It was a PDP-8 that had 16K, 12-bit words. Our product was very competitive with today’s WordStar in terms of features.

We had a good mailmerge, along with list processing and background printing. We could edit a document as large as a floppy because it was not an in-memory editor. We all designed it, and I wrote the actual specification for the word processor, as well as the code for the file system, the command system, and the background printer.

That experience taught me a lot about bringing a product to market. Designing computerized typesetters taught me about the great importance of using the screen effectively and minimizing the number of keystrokes, since many typesetters are paid by the keystroke. That knowledge later was important to me in designing word processors, and it was also valuable in my work with VisiCalc.

INTERVIEWER: Was the workplace at DEC as high-pressure for you as that at Typeset-10?

BRICKLIN: At Digital I was working usually from 11:00 in the morning until 1:00 at night. I wore torn blue jeans, beard out to here, and hair way down my back.

Digital moved up to New Hampshire and I didn’t feel like moving with them, so I started looking around and talking to headhunters. It became obvious to me that I should get an MBA so I would be more marketable. I was also conscious of the people I saw who were in dead-end jobs as programmers, where they were competing against the young whippersnappers like myself, who had good training, would take lower salaries, and work longer hours. I could see it was going to be very hard to stay on top. I saw programmers in their fifties having problems getting jobs. Plus, I had always wanted to go into business for myself, so I thought business school would give me the proper training.

I applied to business school at Harvard and MIT and was accepted at both schools. Initially I wanted to go to MIT because it was a shorter program and I thought I should just get in and out quickly, but in the end I chose Harvard because I felt there was so much to learn I had better spend two years. While waiting to go to business school, I took another job that proved to be quite interesting.

The job was head of programming in a very small company, which made electronic cash registers based on micros. It used Motorola 6800s. They were able to put 64K on one of their boards. It used a coax cable, of sorts, which ran between each of the cash registers. It was programmed in some variant of FORTH. I did a lot of work on the existing system, maintaining it, adding some new features, and upgrading it to the new hardware.

The user base consisted of teenagers working in fast food stores. If something broke, it would take a day for a service technician to show up, and you don’t want to stop a fast food place. It was a real sophisticated system. The home office would poll all their stores every night by computer and dump the information out electronically. They knew how many pickle slices were left on the shelf. When you said you wanted a Big Mac, it knew the recipe.

I watched a real small company actually exist selling against NCR and places like that. It was a big change after Digital, which just had hit about a billion dollars in sales and had made the Fortune 500. I was primed for business school.

INTERVIEWER: Did you come up with the idea for VisiCalc in business school? What prompted it specifically?

BRICKLIN: In business school I used the DEC system when I needed to write a little BASIC program to help me with my homework. But it wasn’t fast enough. Even if it only took me 15 or 20 minutes to throw together a little program to do the analysis we wanted for a group project, that wasn’t fast enough, and the program usually had bugs.

That’s when I came up with the idea for VisiCalc, putting together the immediacy of word processing and the fluidity of the screen.

INTERVIEWER: How did you initially envision the design of VisiCalc?

BRICKLIN: The original design was futuristic. I would have my hand on my calculator and the calculator would have a ball on the bottom of it, like a mouse, so you could move it around to position the cursor on the screen. It had the number key pad right there so you wouldn’t even have to take your hand off in order to do a calculation. I wanted to implement a head-up display, as with a fighter plane where you see the numbers in front of you. I prototyped it in BASIC in the spring of 1978.

INTERVIEWER: When you got the idea, did you imagine it was destined to become as successful as it did?

BRICKLIN: The program went through a lot of evolution before it became VisiCalc. I described it to a friend of mine at Harvard, John Reese, and he was very encouraging. The next thing I thought of was to use a Z80-based machine or something like it, and a TV screen. I thought it would still be nice if it had a mouse. But when I started to prototype it on the Harvard machine, which had no mouse, I had to figure out how to address things. I was faced with the problem of “how do you take these and do that, or take these and add it to that,” because there wasn’t a mouse on the Harvard machine. So I came up with the row-and-column way of doing it.

Then I told my production professor. He was very encouraging. He said: “You know, the type of stuff you’re talking about, people do on blackboards now when they do production planning. And sometimes they have blackboards that stretch the length of two rooms. And they’ll sit there and do the week-to-week planning of this many sold, this many manufactured, this many left. Your program sounds neat.” I talked to another professor in accounting, Jim Cash, who encouraged me by saying, “Good human interface is what is really needed in commerical products. It’s a big problem in many system designs.” And I thought, this guy really knows what he’s talking about.

Then I went to see my finance professor, who was discouraging. He looked up from his printouts and he said: “Well, there already are financial forecasting systems and people won’t buy micros, for example, to do real estate. Go ask this guy, Fylstra. He’S one of my students. He just did a survey to see why they won’t buy it.” So I called Fylstra to see what he was doing and he said, “Hey, my fiancé and I are making software and selling it out of my apartment here. We’re going to be selling chess programs. If you’ve got anything of interest, why don’t you show it to me.” I didn’t talk to him again until fall.

That summer I made the decision, while riding along on a bicycle on Martha’s Vineyard, that when I graduated I was going to go into my own business and try to make this product work. If I had to, I’d sell it door-to-door. In the fall, I finally went over to see what Fylstra had in machine resources. He had an Apple II and a Radio Shack. He offered me the use of the Apple II if I wanted to use it to develop something.

So I spent a weekend writing a BASIC program that would do one screen’s worth of a spreadsheet, and came up with the row-column, A-B-C, 1-2-3 coordinates way of doing things. I still wanted to use a mouse, but I made it work from cursor-moving paddles instead. But the game paddles only worked one direction, horizontal or vertical. So you would hit the fire button to get it to switch between horizontal and vertical. It turned out that the paddles were too slow in BASIC, so I went to the arrow keys. And since the Apple II only has two arrow keys, I used the space bar instead of the firing button to switch it from horizontal to vertical. I don’t like shift keys. I like to minimize shifting. That’s why I use slash to start commands, because it’s unshifted and on the home keyboard, minimizing keystrokes.

So I had this demo that would move the cursor on the screen; you could type numbers and formulas in. You could even point within the formula and say: “1 plus 7 and 1 plus A1,” by pointing to A1, just like today’s spreadsheets. It was just not able to scroll around, and would only do one screen’s worth. It took about 20 seconds to recalculate, and it would make a sound on every cell so you could hear it calculating. It turned out half of the time went into making the sounds. But that demo gave a feel or what you could do.

I showed it to some classmates, like John Reese, who pointed out that when you’re making a reference to another cell, you shouldn’t have to instruct it to use that cell. I was thinking about parsing it; if you say “1 plus” and then hit an arrow key, you must mean one plus that cell. So I went through that to try to get the minimum number of keystrokes to do everything. This became the interface that we know today as the standard, simple-spreadsheet, VisiCalc type of interface.

I showed it to another professor, Barbara Jackson. Her comment was: “Look, if you want to get a chairman of the board of a company to do something, it’s got to be really simple. This is getting close, but it’s not there yet.” So that encouraged me to make it simpler and simpler.

A year or two later when the product was on the market, I brought her in to see it and told her, “Since you’re on the Calculator Committee at Harvard Business School, you’ve got to realize this program is important because you have students who are using it now. You’ve got to prepare for this.” And in fact, that year the business professors used it to write answer sheets for their exams, freaking out all the other professors because they could do their work so fast. Of course, now Harvard requires you to buy a PC before you can go to their business school.

Once we got that human interface done, it was time to actually program it. I had been thinking of the data structure and doing a lot of figuring on how to make it as small and compact as possible, because we wanted it to run on a small machine. Most Apple IIs were 16K in those days, and floppy disks were not common.

INTERVIEWER: Was this when Frankston, you, and Fylstra decided to become partners?

BRICKLIN: Basically, yes. We shook hands over it. Frankston and I would make the program, Fylstra and his company (Personal Software) would sell it. We’d do it for the Apple first because Fylstra happened to have an Apple free to do development on and felt it was the best machine to start with. We also had tools available to work on the Apple.

INTERVIEWER: How did you actually go about designing VisiCalc?

BRICKLIN: I had designed the internals and much of the data structure and layout. To me, the most important part of a program is laying out the data structure. You also have to know what the human interface is going to be like. So I had a data structure that was compact enough to get a good amount of data, and it would be fast to access.

We also had to decide where we were going to base our operation. I was still in business school and Bob was consulting at Interactive Data Corporation. We finally decided to rent time from a large time-sharing system. Luckily they didn’t bill for months at a time so we could live off the float.

Bob would write the code at night when time sharing was cheap. He’d get up around 3:00 in the afternoon, when I would come back from school. We’d go over what he did to the program; I’d test the program, figure out how to do new features, interview accountants, and do other startup procedures. Then I would be there sometimes as late as 11:00 at night. The rates would go down at 6:00 p.m. and again at 11:00 p.m., and the machine would become fast by 1:00 a.m. or 2:00 a.m. He’d work until morning, when he’d go to sleep. That’s how we got the product written.

INTERVIEWER: Were you afraid another company would come up with a similar idea? What was your motivation for working so hard?

BRICKLIN: Once you get an idea that’s so obvious, you want to get it done. We were real afraid that Texas Instruments would find out about it and put it in their new computer. Apple and Atari had been shown the product under nondisclosure. Atari was very excited to get a product, but they didn’t have a machine yet. We got an early prototype which looked like a Cromemco machine, but they had pulled the board and put in an Atari board instead. It was a real prototype.

Apple was not that encouraging about the product. I was not there for the meeting; Fylstra, our publisher, took them a copy of what was to become VisiCalc. As programmers, we thought we’d be done in four weeks. After about two or three weeks, Bob had enough of it going that it would scroll around and would add and subtract and recalculate. So Fylstra took that out for them to see. I kept on enhancing my BASIC prototype and trying out new things. Many ideas had to be dropped because we only had 16K of memory.

We got good response from the West Coast. I brought in some other professors. Another production professor, who was very encouraging, said, “Boy, you know, it’s really neat to do all those calculations in almost no time. That would have taken hours to do it by hand. I’ll have to update the examples in my book.”

So I wrote an early manual and Bob would try to implement what I wrote. We were a good team because whenever Bob would slack off, I would say “Bob, you’ve got to make this thing faster, it doesn’t feel right,” or I’d say, “you’ve got to put this feature in.” And whenever I’d say, “Well, I think we should cut a corner here,” he’d say, “No, no, no, let’s put it in.”

He ended up using a variation on my design for the cell storage. I did a lot of the design work, but the internal program structure was all his. And that’s how we ended up with the VisiCalc program.

INTERVIEWER: what was the initial reaction to the finished project?

BRICKLIN: It got good reviews from a few people. Most of the magazines ignored it. It wasn’t written up in some of the magazines for about a year. BYTE magazine had a little editorial about it and that was all. The guy who wrote the editorial was best man at Fylstra’s wedding, Carl Helmer, and he already knew about it. And even that piece was written for engineers and talked about sine and cosine, which we hadn’t implemented, but because it mentioned that, we had to implement it. I wasted my summer implementing sine and cosine, and other features like that.

We bought our own time-sharing system on borrowed money. I sold a house and used the money towards the down payment. Bob used his savings and borrowed from his family. We made the down payment on a minicomputer and moved into commercial quarters.

INTERVIEWER: How did you decide to market VisiCalc?

BRICKLIN: We had no idea how well the product would sell. We knew it was very good and everybody should have it, but it was unclear at that time whether everybody would jump on the bandwagon for electronic spreadsheets, since they had to buy this funny machine, and word processors had failed to take off.

INTERVIEWER: So once you set up your business, what did you have to do? Did you go around the country demonstrating it?

BRICKLIN: Our publisher, Personal Software, did a lot of that. We demonstrated it at the National Computer Conference, which is where it was publicly announced. We also demonstrated it at the West Coast Computer Fair and various shows.

Then we went back to do more work to put it on more and more machines. We put it on the TRS-80, which was a very important machine; it was the number two computer from our viewpoint. The programmer who did that was Seth Steinberg, who joined us. Steve Lawrence was our first employee. He did some of the programming to finish up VisiCalc, in the numeric area and some other functions. Then he set out to do the Atari and the Commodore PET versions.

For the Z80 version, Seth decided to do a literal line-by-line translation of the 6502 code into the Z80 code. And since we had a nice time-sharing system, he could invent all the tools he wanted. So he wrote a printing program that would list side by side the 6502 code and the Z80 code. We found out years later that bugs in the 6502 code were also in the Z80 code and vice versa. It was that good a translation. That code lasted for a long time. It became the IBM version, with some mechanical translation and some minor hand translation.

We were able to get on the IBM PC quickly. When IBM announced their PC, they announced it with VisiCalc. And they would not announce it unless VisiCalc was running to a certain percentage, enough to ship; they’re very fussy. In order to do that, we couldn’t use the standard conversion programs because we had our own assembler. One of our programmers, David Levin, modified the assembler to do the translation from the Z80 code into 8086 code, or 8088 code, and flagged the problems. It had to be pretty smart about doing the conversion; it was a lot smarter than most of the other converters around. Because of that, we were able to get up on the IBM quickly; shipping a few weeks after IBM shipped their PC, so we sold an awful lot of copies of the new product.

Bricklin’s Original command diagram for VisiCalc from November, 1978. The diagram was written on a piece of spreadsheet paper (hence the grid in the program) that Bricklin used in business school.

INTERVIEWER: How did you get the arrangement with IBM?

BRICKLIN: IBM came to us first because the copyright was ours. Our contract was that they should deal with Personal Software, so we sent them off to Personal Software. We actually had a three-way arrangement with IBM. When IBM called, we met with them the next day. We told them about our contract, showed them the parts of our contract that we were allowed to show them, and let them see how they would have to deal with us. We were very up-front about that.

Now remember, we had been dealing with a lot of different manufacturers, but we tried real hard with some who became successful and some who didn’t. You never know which ones are going to make it.

INTERVIEWER: Did you sense at that point that IBM was going to make it?

BRICKLIN: It took a while before we knew what the machine was. It was obvious that the IBM machine was a great design. But the general idea of using an 8088, making it with a full keyboard, an 80 x 25 screen, and all that stuff, was something a lot of other people were doing at that time, including the best hardware designers around. It was clear it was a good design; that’s why we felt we were going to make it.

We had a lot of contractual problems with Personal Software. For example, we had to put VisiCalc on whatever machines they wanted. They wanted to get out of our contract, but didn’t want to pay enough to get out of it. And so we had to be able to put our product on as many machines as possible.

INTERVIEWER: Let me ask you about the rise and fall of Software Arts. Did you ever think you could go all the way down?

BRICKLIN: Even though I was all the way up, I lived in the same type of house I would have lived in if I worked at Digital, or anywhere else. I always kept myself down there in terms of spending so that I could afford to do what I’m doing now. I never wanted to depend on the company.

That’s why we were trying to sell the company at our peak, but we ended up with a lawsuit at that point, and with a lawsuit, you can’t sell a company very well. We had a company that was valued very highly. On paper, Bob and I were pretty rich, but unable to do anything about it, though we tried. By having visibility in our industry, I gained some “equity” that I took away with me. It made it possible for me to be an entrepreneur again.

The experience with the company was worthwhile. I’ve turned down some pretty good job offers at various companies over the last year. So I know if I want to I can go to work somewhere else, but I’m an entrepreneur.

I’m very happy looking back. It would have been nice if we had stayed at our peak and become very rich, but we didn’t. I did better than if I had stayed at Digital–let’s put it that way.

I liked working at Software Arts. I was very happy there, except during the period of business problems. But even that was a learning experience. I learned a lot about the life cycle of a business. I’m also more confident starting a business again after having gone through the whole cycle. By getting to the point of actually auctioning our chairs, our terminals, our pictures off the wall, my desk, and the T-shirt right off my back, I now know that you can go all the way up and all the way down, but you still survive.

INTERVIEWER: Looking back on Software Arts, would there be anything you’d do differently today?

BRICKLIN: Well, I would have liked to have not been sued; that would have made things different.

INTERVIEWER: How did that come about?

BRICKLIN: We got a notice in the mail saying, “You’re sued.” We had been having disagreements with our publisher, VisiCorp (Personal Software), for years, but we had been negotiating settlements for over a year or two at that point. For some reason, they decided that they wanted to bring it to a head as a lawsuit. Maybe they decided that they had a strong enough case and would get the money from their product, VisionCalc. But they didn’t check to see if VisiOn was going to sell well first. They burned their bridges before they crossed them. We didn’t expect them to actually sue us.

There are things that I could have done differently if I knew what I know now. Some things could have been done differently if the market had been different. Would some large publisher, like McGraw-Hill or somebody, have been willing to go into the business? Could we have convinced Apple to carry it? Who knows? But the outcome wasn’t that bad, so why would I want to do it differently? I wouldn’t have gone through all the experiences I had otherwise. I’m still healthy and I’m not poor and I have the respect of my colleagues. Other than in terms of becoming rich and having an unblemished history for my company, I’ve succeeded in just about every measure. And most people have their ups and their downs: Look at what happened with Apple.

INTERVIEWER: If the opportunity presented itself, would you go ahead and do the same?

BRICKLIN: Only if I thought it was correct. I like doing what I’m doing now. I’ve managed a big software business; I know what it’s like. At this point, I can do what I want to do just as well with a small business. It’ll get too cumbersome at some point, but it’s a challenge. There are always different challenges; one was to build a big company. Now the challenge is to be successful in a very small company.

INTERVIEWER: Do you see Software Garden growing out of your experience with Software Arts?

BRICKLIN: Yes and no. At Software Arts, we got to try a lot of ideas: big company, great environment, an R&D organization. I decided that you either need a lot of money coming in from your product, or you’ve got to have very low overhead. And what I like to do is try different kinds of products. I’ve created a variety of products; I intend to do more in different areas. I intend to have a little garden here, not a farm where I’m going to emphasize one area, unless it turns out that one product is particularly lucrative.

I’m selling my products through the mail and through distributors who buy from me. I’ve already signed up one distributor and I’m working on a few more, selling the product inexpensively for $74.95, trying to get some publicity started.

INTERVIEWER: What’s the name of your new product?

BRICKLIN: Dan Bricklin’s Demo Program. Nice simple name, right? I decided I was going to call it Dan Bricklin something, because I know that’s necessary to help sell it; enough people have told me that. The name is descriptive of what it is; it’s for making demonstrations.

The packaging is real simple. See this Tyvek envelope? And the documentation will be typewritten, but nice. I decided to run the business out of my house, and for now, that’s the way it is. The disk comes stuck on a piece of cardboard, and in this envelope. That’s it.

I know how to do the real slick packaging, but what would I, as a user, do with my slick packaging? It takes up space on my shelf. The program sits in the computer. So I decided to go with simplicity. Why try to act like somebody I don’t have to be?

INTERVIEWER: Who do you see buying this program?

BRICKLIN: Any person who is involved in designing programs, anybody who is involved in the development process. Documentation people will use it to write demos and tutorials. Once you try this program a couple times, it will pay for itself in terms of time saved.

I’m also trying to look at the viability of different businesses today. If you have the lowest possible overhead, a good product, good public visibility, and you have low prices, how many units can you sell? Some people have done very well in that area. I’d like to see what’s there, because it will make a difference for what other things come out of the Garden.

INTERVIEWER: Do you think you’ll always be programming software?

BRICKLIN: I guess so; I seem to have a knack for it. I don’t know what it’s going to be like when the technology changes. Maybe my skills are more appropriate in character-based screens where I’ve had most of my experience, as opposed to graphics, where I don’t have as much experience. Maybe technology will pass me by, maybe not.

INTERVIEWER: Do you think there will continue to be these small types of software operations? Or will the software industry be dominated by a handful of huge companies?

BRICKLIN: No. People are writing their own programs. Anybody who uses a spreadsheet is writing their own programs; it’s just that the language is different now.

Bob Frankston likes to tell a story about the telephone company. Back in the twenties they said that telephones were growing so fast that by 1950 everybody in the country would have to be an operator. By 1950 people said, “Ha, they were wrong.” Well, it turned out they were right, because everybody was an operator; they just had dial telephones. The technology had made it easy enough to be an operator.

Well, that’s the same thing with programming. We’re just making the users do more and more of the programming themselves, but they don’t know it. Using different style sheets with Microsoft Word is doing programming; using spreadsheets is doing programming.

There is an inherent cottage-industry component to programming. We constantly improve the tools that are available, the operating systems, environments, languages, and all. Consequently, an individual can always do a lot more than he used to.

It’s like writing. An encyclopedia requires a large staff; an individual can’t do it all. Some operating systems nowadays are encyclopedic. But there’s also always the short story, the novel, and the poem. Those are useful in their own way, though they may seem simple and trivial to do.

Any large company could have a million programmers working on some idea to produce a better one, but they don’t because it doesn’t work that way. The economics aren’t there. Sometimes the idea behind a program is one small creative effort. Just like in a short story, one little twist in the plot is the whole idea behind it. The same thing is true of the programs.

In terms of marketing, you do need a large marketing organization. But there’s a variety of ways to do that, just like with books. You can self-publish or you can go to a publisher: With movies, you have the distribution company that takes care of it.

If you look at the biggest sellers in the software industry, in general they were written by very few people. 1-2-3 was in essence written and designed by two people, one programmer and one partial programmer; VisiCalc, the same way. dBASE, I think, was originally written by pretty much one person. Paradox was written by two people. That’s the way things are done.

INTERVIEWER: Do you think software will break into entirely new areas, or do you think it will remain in the same areas, such as spreadsheets, word processors, and so on?

BRICKLIN: All sorts of products are developing now. People are saying, “What about networking, why can’t we connect this stuff together?” What about the publication systems, like the inexpensive ones available on the Mac? And the really great ones that are available on the bigger machines, like Apollo and Sun? A few years ago, no one thought that in-house publishing was going to be a major use of computers.

People want things to move too fast. How long did it take for the mouse to be in commercial use? Was it fifteen years, or more? Englebart came up with it back in the sixties or early seventies. And compilers–FORTRAN was developed back in the fifties. We take those things for granted, but they were done in spans of five and ten years. Software has major advances every half decade or decade. Well, why should that change too much?

Once you have certain hardware, you have certain capabilities that you didn’t have before. And frequently it turns out that you can do something without specific hardware. For example, certain systems on the IBM are done because the Macintosh had them and people want them. The new hardware, like the Mac, helped make it clear the systems are useful. Consequently, it is worth the effort to figure out how to do it on the old hardware.

The Apple II is another great example of that. When it started out, people thought they had pushed it as far as they could. But they are constantly finding more things the Apple II can do. It’s like a test pilot pushing a new plane beyond its expected capabilities. A lot of people crash and burn; they did on the Macintosh.

INTERVIEWER: What do you see as the failures of that machine?

BRICKLIN: The Macintosh has sirens that work against you, calling you to the rocks, such as using fonts. To put in different fonts slows down a word processor quite a bit. You end up with an unzippy word processor, which is not what you want. People like something fast. Once you put in different size fonts, that’ll make this slower and that slower, although the end result will be pretty. But you end up with a product that doesn’t feel right because you listened to the sirens. Eventually you learn not to do that.

INTERVIEWER: How important are speed, user friendliness, or other similar considerations?

BRICKLIN: For some products, that’s the whole product. That’s what makes it what it is.

Suppose you had Microsoft Word running on a Hercules card. That would be really slow for certain things, slower than it could have been if they hadn’t put so many enhancements into it. Then Word would be unusable to some people. Even though functionally it would be the same, the number of people who might want to use it would drop.

I think the feel of the product is very important. Lotus showed that. There was a product called Context MBA that was just as capable and had a lot of features, but it was so slow that many people never even considered using it, so it never caught on at all. What Lotus went after was speed.

On VisiCalc we got the same speed as the original Apple II. I said, “Bob, that’s not fast enough, we want it faster. It’s got to scroll as fast as the repeat key on your machine.” I didn’t know different machines repeated at different rates. Ours was about a medium rate. We tried to get that same feel in the IBM and it turned out to have a nice, crisp feel.

INTERVIEWER: Do you consider programming a skill or a science?

BRICKLIN: Well, parts of it are science and parts of it are skill. There are certain things described in a science that can be useful, like in power design and things like that. There is a craft component in which apprenticeship is the way to go. You learn skills by apprenticeship. And with some things, practice just makes perfect. So it’s more like a lot of the arts or crafts. It’s not like being an actuary; it’s more of a standard type of craft.

You can see that in the feel of certain programs of developers; it shows up when you look at their code. People who have formal training often have an advantage over people who don’t. Knowing what people have worked out academically is often very helpful. You know, some people have green thumbs and some people don’t, but it helps if you learn a bit about the area first.

INTERVIEWER: When you program, do you basically use the same method or approach every time?

BRICKLIN: No. It depends on what I’m doing. I mean, I try to be professional; you don’t have to be. But if you’re producing products for commercial use and you want to get things out, you can’t be very religious about your work, such as, “I only program in high-level languages,” or “real programmers use only assemblers.” You simply do whatever is appropriate for the particular situation.

INTERVIEWER: What role do you see software or computers playing five or ten years from now?

BRICKLIN: I think we’ll see more forms of computers than we do now. One form is a box that sits on a desk, the personal computer of today.

I think the personal computer of tomorrow is going to be different than the desk computer, which now is like a telephone that sits on your desk. Desk computers didn’t catch on in the home because the PC, as we know it today, is to the home as the mainframe is to a business–you have to go to the shrine, or connect up to it. A lot of people don’t want to connect to the mainframe, send cards in, and stuff like that. People aren’t going to want to go into the computer room in the den to check their recipe file. That’s ridiculous.

The personal computer of the future should be more like a notebook. I carry my notebook around and why shouldn’t it be a computer? Well, that’s different than the PC as we know it. Computer technology is going to be used for all sorts of new areas like that.

Let’s say years from now voice recognition has been worked out pretty well. You have a little tape recorder over there that understands commands, so why push a button? You just say, “Play me the last five minutes of the tape.” It’s not going to be connected to a PC on your desk to make it do that. It’s going to be built in.

The computer you have on the desk is going to constantly improve. The price is a big factor. Take a copy machine; you can get new ones for about the same price as you used to pay for the old ones, but the capabilities have gone up. I think we are going to see personal computers that are going to have quite a few different embodiments.

Portia Isaacson used to say that she thought robots were going to be important because your computer should follow you around. One way to have a computer follow you around is to miniaturize it. Why go to all the trouble to put legs on it if we can miniaturize it to the point that we can carry it on our bodies. We’re getting to the point soon where we can get a lot of computing power in a very small space.

Bricklin’s reference card for an early version of VisiCalc (called Calculedger or CL at the time), which he wrote on a rented IBM Selectric typewriter to use in a demonstration at Apple. The Appendix contains additional mementos from the early days of VisiCalc.


  1. […] dug back into my PAW archives and read over Dan Bricklin’s interview from 1986 and posted it today. You can find the link in the column to the right. If you go to the end, you […]

    Pingback by Dan Bricklin- 2010 « Programmers At Work — June 17, 2010 @ 4:23 pm | Reply

  2. […] de 1986 en que Susan Lammers recopiló entrevistas con los míticos genios de la época: desde Dan Bricklin (la hoja de cálculo VisiCalc) a John Warnock (el lenguaje PostScript) a Toru Iwatani (creador del […]

    Pingback by Programadores en acción | SUNETFON — February 8, 2014 @ 10:23 am | Reply

  3. […] de 1986 en que Susan Lammers recopiló entrevistas con los míticos genios de la época: desde Dan Bricklin (la hoja de cálculo VisiCalc) a John Warnock (el lenguaje PostScript) a Toru Iwatani (creador del […]

    Pingback by Programadores en acción | Misiongeek — February 8, 2014 @ 11:00 am | Reply

RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Create a free website or blog at

%d bloggers like this: