Tech's Lost Chapter: An Oral History of Boston's Rise and Fall, Part One

In the popular telling, the dawn of personal computing begins in the summer of 1976, when Steve Wozniak showed off the Apple I at a meeting of the Homebrew Computer Club in Silicon Valley.

But decades of work preceded that moment, a series of hard-won innovations that shrunk processors and memory down to the point that a computer could fit into a box the size of a typewriter, and retail for less than $700.

Many of those seminal steps occurred not in the Silicon Valley garages of lore, but in the stuffier labs and offices of Greater Boston, stretching back at least to the Massachusetts Institute of Technology’s Whirlwind Project.

It began in 1944 as a wartime effort to build a flight simulator, but ended up moving the computing industry from electrostatic tubes to magnetic core memory and, eventually, from mainframes to minicomputers.

It gave rise to classic machines like the PDP-8, and corporate giants like Digital Equipment Corp., which for a moment in time rivaled IBM in size.

But DEC and others missed the industry shift that would place computers in nearly every American home, despite the region’s expertise, contributions and sizable head start. Or maybe, in part, because of it.

The collapse of tech manufacturing along the Route 128 corridor, which has been debated and studied in the ensuing decades, is a revealing case study of disruption, rapid technology cycles and what has come to be known as the Innovator’s Dilemma.

The region’s businesses clung so tightly to the revenue spun out by legacy products, and aging assumptions about what computers were and who would need them, that they failed to appreciate where the industry was moving.

It’s a tale little discussed today, a nearly lost chapter of tech history among the notoriously short memories of an industry obsessed with what’s new. But it still offers lessons for modern entrepreneurs, including a stark reminder that it is not merely cloddish corporations that can topple — but entire industries and regions.

In recent months, we asked the people who were actually there what happened, and why. What follows is the real story, as told in their own words.

Gordon Bell, researcher emeritus at Microsoft Research and an early employee at Digital Equipment Corp., which he joined in 1960:

I pretty much will trace everything back to MIT Whirlwind, a vacuum-tube computer that spawned the invention of core memory, which is the main memory of computing until probably 1980.

Vjeran Pavic for Re/code

That, in turn, spawned (MIT) Lincoln Lab. The TX-0 came out of that, and the circuitry and architecture of the TX-0 really spawned the basic minicomputer that was, in fact, traditional Boston.

Dan Bricklin, DEC employee from 1973-1976, who went on to invent the digital spreadsheet as a graduate student at Harvard Business School:

The [Whirlwind Project] was the precursor of so many machines we have today. People at DEC, Data General and others got their start there, they were on that project. Some of the people who ran Project MAC and Multics, the time-sharing system and stuff that became Unix, were on that project.

Bob Frankston, who worked on Project MAC as a MIT student, and later co-founded Software Arts with Bricklin:

As I entered the scene in the ’60s, there was sort of an explosion of ideas because we had all these new technologies to play with. There were a few places where this was happening. MIT happened to be — and it might just be my perspective — the main site.

You could see nascent ideas like computer networks and what you would now call the cloud. Jerry Saltzer wrote Typesetter and Runoff, and Runoff led directly to HTML. Some of the people there working on the CTSS (Compatible Time-Sharing System) project built one of the first email programs.

And we were doing it all ourselves.

Bell: DEC formed in 1957 to sell transistorized circuits to make computers and digital test equipment. But in 1960 it came out with its first computer using those circuits, the PDP-1, which was really the forerunner of the PDP-8, which I say is the classic minicomputer.

That was sort of the beginning of computing, and from that a number of other companies later formed, including Data General.

In 1966, we have the minicomputer boom, because integrated circuits came out, and anybody who wanted to make a computer could take integrated circuits, slap them on a board and they’ve got a computer company.

Frankston: If you look at a lot of the early market here, building time-sharing and all those systems, it was not to create just one solution. It was to create enabling technology. It was giving people the tools to build the tools.

In New York, computers were still about business. And in Boston, the reason I stayed was there were all these experiments, ideas and science. We were exploring what we could do with computers.

Ina Fried for Re/code

Ray Ozzie, longtime Microsoft executive, founder of Talko and creator of what would become Lotus Notes. He first moved to Boston to join Data General in 1979:

It was kind of what you picture the Bay Area being like today with a lot of young people, single, coming from major colleges all over the United States. If there was a difference in the environment, however, it’s that New England companies tend to be frugal. So there weren’t many of the kinds of things you see in the Bay Area right now, with provided meals and that kind of thing.

Work and play were very blurred, and everyone felt like they were in the center of the universe. It was well-known as being a tech hotbed. There were universities all over the city, and great music, great food, a great place to be — and still is.

Bricklin: It was really cool times. It was really exciting. Computers, we were starting to share, we could share a computer among many people. We could use them for real things. I could use them to write my thesis with, my bachelor’s thesis, using Runoff, which is the precursor to all sorts of stuff.

Mitch Kapor, founder of Lotus Development Corp. and co-founder of the Electronic Frontier Foundation:

What Boston had was the minicomputer industry, which was huge. Digital Equipment Corp. and Data General and Wang and Prime and others. Tens of billions of dollars and hundreds of thousands of employees. It had this incredible tech scene.

Ozzie: DEC was focused looking up at IBM and the higher-end mainframes, while Data General was looking at trying to disrupt DEC from the low end.

We were the upstart disruptors, and it was a tremendously fun environment. Just everybody was young, everyone was just out of college, it was a tremendous environment.

It was a fast-moving era. We had this view that if you built this powerful integrated system focused on whatever market segment you were at, you could just do anything.

But in 1977, Apple Computer introduced the Apple II, an all-in-one product that average consumers could begin to use. It would become one of the first mass-produced computers, along with machines like the Tandy TRS-80 and the Commodore Pet. IBM followed up with its 5105 PC in August 1981. Boston’s minicomputer and microcomputer companies began to shift in this direction, but not fast enough to capture the emerging market.

Reed Sturtevant, managing director at Project 11 and the designer of Freelance Graphics, a presentation software package acquired by Lotus in 1986:

The computers progressively got smaller and smaller, to refrigerator-sized, washing-machine-sized. There was a ton of innovation around the different form factors.

But those companies didn’t realize that the PC itself could be a commodity — and that that was okay.

I’m not sure why I would say that happened. It felt like they were doing the right things and hedging the right bets. But for whatever reason, the incumbent hardware companies never really caught, in Boston, the PC wave.

Bricklin: At DEC, I remember asking, I think at a shareholders meeting: “What about small computers?” Because we had one that actually ran Basic, I think.

[DEC Chief Executive] Ken Olsen was talking about wheelbarrows versus pickup trucks. Classical “Innovator’s Dilemma.” He knew he was selling to engineers, and they wanted the bigger ones.

Kapor: Because Apple was founded in Silicon Valley, I could see and felt like the center of gravity in the tech world was shifting there — and Boston was not ultimately the best place to be.

Some of my most sobering moments I had in the early years of Lotus were seeing how those guys weren’t getting it, and were dooming themselves.

This is before Clay Christensen had written “Innovator’s Dilemma,” but we were experiencing it.

I remember a meeting I had with Ken Olsen, first time I met him, I got summoned out to the Mill building in Maynard, where Digital was founded.

I had this bizarre interaction with him. I thought I was on “Candid Camera” or something, because he was ranting on about what a crappy, cheap case IBM put on its PC, and it wasn’t industrial-strength. He had himself designed the case for the DEC personal computer, and he was banging on it and showing me how solid it was.

I was going, “What planet am I on? This has nothing to do [with it].”

But in his world, when computers were on factory floors and so on, they needed to be robust. That really mattered. This didn’t matter at all.

David Litwack, head of product development at Cullinet Software and CEO of PowerSoft:

The whole minicomputer industry was in the Boston area. But it turned out to be a very short-lived phenomenon.

The same thing happened as with the mainframe: Every manufacturer tried to protect their business by being proprietary.

From the middle ’70s to the late ’80s, you went from a world where everything was proprietary — the chipset, the instruction set, the hardware, the language — to where applications were no longer hardwired to the hardware.

Ozzie: The decline of the minicomputer hit the region. DEC, Data General, Prime were extremely affected, and they were fixtures in the area, especially DEC.

It wasn’t clear how it would all shake out at the time; people more or less thought this is just the natural cleansing. They thought, “well, it’s cyclical, it will come back again.”

Bell: A hundred companies formed and ultimately died during that two-decade era, when they were snuffed out by the microprocessor.

So why does something start, why does it die? The minicomputer is a wonderful, wonderful case study of technology, and how long it lasts, and why it’s superseded or not.

At its peak, Digital Equipment Corp. was the second-largest computer company after IBM, and second-biggest employer in Massachusetts, with a workforce of around 120,000.

But by the mid-1980s, the minicomputer sector began to shed tens of thousands of jobs. The state lost more than 30,000 high-tech manufacturing positions in five years. Compaq bought what was left of DEC in 1998, EMC acquired Data General and Wang filed for bankruptcy.

Read Part Two of this two-part story here.