Tracy Kidder's classic portrait of (very bourgeois) geeks at work reminds us how fast technology advances. Barely a generation ago, designing a new computer necessitated writing an operating system too, leading to three linked main tasks: Creating the hardware, the operating system and the micro-code linking the two.

The development takes places as a skunk project at Data General, one of many defunct brand names (such as Commodore) that litter the path to the emergence of modern computing. In a major misunderstanding of their success factors, the management of Data General decided to transfer their research from Massachusetts to North Carolina, mainly to create a tax shelter. The physical distance between the engineers (in NC) and the management (in MA) led to a botched technology upgrade.

Enter our heroes. Tom West assembles a team of trusted survivors and rookies to develop a backward-compatible 16-bit minicomputer. Practicing the mushroom theory of Management ("keep them in the dark, feed them shit, and watch them grow"), he shelters his team from too much CEO interference but also does not offer much direction. The actual leadership and care comes from two project managers, one responsible for the hardware, the other for the software. Despite all expectations, the team beats both their NC rivals and finish their design.

The tale does not end well. The company does not adequately reward the survivors of what will later be labeled "death-march" project, thus alienating and losing their most productive members. The shell of the company lives on for some time, while the people move on to new projects and new companies, thus completing the circle of life of technology companies.

Armed with Brooks' classic article "The Mythical man-month" and the consultant's bible at the time "In Search of Excellence", Tracy Kidder remains a fresh account of innovation and project management. ( )

A technically and socially satisfying account of the development of a new computer in the late 1970's. Its glorification of the exploitative and ruthless management style of Data General is off-putting, but this might be a minor quibble on an otherwise fine book. ( )

Not a difficult read, and an interesting snapshot of a team of people at a technology company (Data General) in Massachusetts in the 70s as they work to develop a new minicomputer. The pace is good, and the portraits of the engineers are well-drawn. I appreciated Kidder's ability to explain the hardware in terms a lay person can understand. The book captures well the excitement that develops among a team of workers inspired by the challenge of inventing something new--and the fallout when they are forced, near the end, to document their individual contributions in order to decide on patents. The picture of Tom West is at the center of the book, early on, he seems to have a little of a sinister side, but he eventually emerges from the project more or less unscathed and redeemed. However, I do think that today there would probably be a diagnosis for him--more's the pity, he would have been on some kind of bipolar meds and "the machine" probably would never have been developed (and who knows what toll it took on his personal life). I don't know if later editions have a "Where are they now" section, but that would be interesting to read. ( )

I was drawn to this book in 2008 because of my two decades of experience with Data General (DG) computers, both the 16-bit Nova class and the 32-bit MV class of minicomputers. My experience included system administration, script writing, programming in timeshare BASIC, setting up smart PCs as workstations, and running a help desk. I was too busy working with DG systems in the 1980s to realize that our MV was the product of a history-making process.

In the mid 1970s, when Digital Equipment Corporation announced the VAX series, their first 32-bit minicomputers, DG responded with a crash course called the "Eagle Project." This project is the subject of Tracy Kidder's Pulitzer prize-winning book, The Soul of a New Machine (1981). Kidder's book single-handedly made Data General's MV line of minicomputers the best documented computer project in history.

Kidder's book reads more like a fast-paced novel -- with somewhat less sex and violence -- than like a pedantic history book.

My Favorite Chapter: My curiosity was piqued when I came to Chapter 6, "Midnight Programmer." Being one myself, programmers were people I could identify with. But as I turned the pages, the story line gradually began to lose its appeal as I read about "Microkids" who worked on "Microteams" at their crowded desks in "Micropits" at DG. After all, my work as a programmer was done as a loner, not as a team member. To be productive, I needed quiet and solitude. Then I read this sentence that instantly drew me back into the narrative: "Much of the engineering of computers takes place in silence, while engineers pace in hallways or sit alone and gaze at blank pages." "Yes," I said to myself, "that's the way I write new software -- like an engineer at DG designs new hardware."

My Favorite Quote: I almost came unhinged when I read, "Not Everything Worth Doing Is Worth Doing Well." This was a favorite saying of Tom West, the chief designer in charge of the Eagle Project, the principal character in Kidder's narrative. In saying this, West meant, if you can do a quick-and-dirty job, and it works, do it. As a manager, West was pushing to get things done on time and on cost. As a programmer for whom errors are intolerable, this is a piece of managerial advice that I could never, ever internalize or respect. This saying may have worked for The Microkids at DG, but not for programmers like me.

We stayed with our DG MV at my institution for about a decade. Our operating system was AOS/VS, the most commonly used DG software product. It included CLI (Command Line Interpreter) allowing for complex scripting, DUMP/LOAD, and other custom components. Our MV ran our tailor-made software on demand with very little down time.

Although we were happy with the reliability of our Data General systems at my institution, the time came when a major change in software vendors necessitated a corresponding change in hardware. We replaced our 32-bit Data General hardware with 64-bit Alpha hardware from Digital Equipment Corporation.

Looking back on the experience two decades later, I must say in retrospect that the DG MV series was too little too late. While Data General was investing its last dollar into a dying minicomputer market, the personal computer was rapidly on the rise. For example, the 32-bit MV/8000 went out of DG's door in 1980. Barely a year later, in 1981, the 16-bit PC/XT went out of IBM's door. It was just a matter of time until PCs overtook the minicomputer market with the arrival of 32-bit, then 64-bit, then dual-core, then quad-core PCs.

This book is still a good read more than twenty-five years after it was written.

Trivia: Tom West, the protagonist in our story, was Tracy Kidder's college roommate. ( )

An excellent expose on the wilder days of the development of mini computers - this one is the DG MV8000, the first Data General 32 bit machine. ( )