The building of a computer, told with suspense

One of the more challenging tasks of writers these days is to interpret the computer world for laymen who have only vague notions of what is going on there.

The usual approach is patiently didactic, relying heavily on analogies. It is assumed that most ordinary readers won't understand the electronics involved, so authors are careful not to spend too much time explaining how a computer works. Most writers choose to concentrate on the remarkable speed and complexity of computer performance in a variety of applications.

By contrast "The Soul of a New Machine" makes its appeal to the lay reader in the form of a near-novelistic suspense story drawn from the actual development of a new high-performance computer at Data General Corporation in Westborough, Mass.

It is easy to visualize the best-selling novel that might have been written from the materials author Tracy Kidder patiently gathered while hovering over Project Eagle. Kidder chose, however, to write multipurpose nonfictin: a painstakingly factual account of how a specific product was designed and built; an informal study of management and professional styles; a layman's guide to computers and programming.

To humanize his account, and to suggest the impact of personalities on such a seemingly cold p roject as the design of a machine, Kidder turns aside periodically for close-ups of key figures in his industrial drama. These vignettes are what give the book its novelistic flavor and hint the fictional road that was not taken.

Among these engaging protagonists are:

* The project manager, a highly motivated designer who, under the stress of unrealistic, self-imposed deadlines, could exhort his troops with evangelistic zeal one day and be heard muttering "computers are irrelevant" the next. Described as "so happy and funny and warm-hearted" at parties he threw, the manager could also intimidate most of his office staff and get commitments not far short of solemn oaths (understated as "signing up") to support Project Eagle through extraordinary effort.

* The computer architect who refused an important transfer to North Carolina, fearing he would constantly hear on the radio, not stock market, but livestock, reports. So he stayed on in Massachusetts and helped Project Eagle over a major hurdle, giving the home office an edge over the Southern branch, working on its own version of a 32-bit machine.

* The microcode specialist who supervised the Microkids (high grades, just out of college), daringly chosen to do work usually saved for veteran programmers.

* The secretary who left an insurance company to work for Data General after she saw an ad that started out "Are you bored" She declared, "It was speaking to me!" Later, under particular stress, she would speak of the project and the company as "like one of those terrible movies. I just have to see how it comes out."

The reader will be in a little doubt about whether Eagle (late marketed as the Eclipse MV/8000) will fly, but Tracy Kidder manages considerable suspense along the way. Although only a computer specialist will fully appreciate the achievement of this little band of obsessed engineers, even a curious outsider can get a feeling for the difficulties they faced and the ingenuity nand inspiration they brought to bear on a highly doubtful assignment.

The book also provides a window on corporate life at a kind of frontier level , where reputations and fortunes can be made or lost almost overnight. Computers designed at considerable human as well as monetary cost can have a shorter life than a good washing machine, because the designers of these electronic marvels are pushing, pushing, pushing all the time for yet another advantage in this potentially lucrative and highly competitive market.