Tracy Kidder’s The Soul of a New Machine was one of the great nonfiction hits of the early 1980s, a gripping blow-by-blow account of how Data General developed the Eclipse minicomputer. Many people originally viewed it as a description of how high-tech projects ought to work.
Later it came to be seen as more of a cautionary tale. The Eclipse was never a great success in the market and the engineers who created it were burned out by the project. Most ended up getting divorced and leaving the company. And Data General itself seems to have vanished from the face of the Earth. I have no idea what happened to it, but I haven’t heard any mention of it for many years.
Scott Rosenberg has set out to duplicate Kidder’s success by writing a book called Dreaming in Code: Two Dozen Programmers, Three Years, 4,732 Bugs, and One Quest for Transcendent Software. This was supposed to be a blow-by-blow description of how Mitch Kapor, the creator of Lotus 1-2-3, led a team of developers who who created a world-changing new software product: an open source, web-based Personal Information Manager called Chandler.
Unfortunately it turned out to be the story of how Mitch Kapor and his team spent millions of dollars and many years of effort to create a product that doesn’t do much. Of course this could be interesting as a cautionary tale it itself.
Joel Spolsky has written a review which is mostly a rant about what was wrong with the project from the start. Mainly he sees it as a design probem:
Near as I can tell, Chandlerâ€™s original vision was pretty much just to be â€œrevolutionary.â€ Well, I donâ€™t know about you, but I canâ€™t code â€œrevolutionary.â€ I need more details to write code. Whenever the spec describes the product in terms of adjectives (â€œit will be extremely coolâ€) rather than specifics (â€œit will have brushed-aluminum title bars and all the icons will be reflected a little bit, as if placed on a grand pianoâ€) you know youâ€™re in trouble.