Soon after July 1987 black-bearded, balding, middle-aged computer industry analyst and consultant Michael Killen, loving husband of Josephine and father of Michael David (Max), Karl, and Diana, sat at his six megabyte high-performance workstation in Palo Alto preparing to write a book about IBM’s SAA that would rival Tracy Kidder’s The soul of a new machine [1]. He felt Frame Maker’s text processing package quivering beneath his hands; he heard the soft swirl of his 330 megabytes of on-line disk storage; and he sensed the presence in the room of his Imagen laser printer with five megabytes of main memory and a twenty-megabyte hard disk. His fingers flew.

The result is an overdramatized, overheated, over-written, and over-enthusiastic mixture of history and speculation, excessively padded out with meaningless detail, written in the style parodied above, that purports to give the behind-the-scenes story of the people and events that led to IBM’s March 17, 1987 announcement that began, “Today, IBM announces IBM Systems Application Architecture, a collection of selected software interfaces, conventions and protocols that will be published in 1987. IBM Systems Application Architecture will be the framework for development of consistent applications across the future offerings of the major IBM computing environments--System/370, System/3X, and Personal Computer.”

The story is mainly confined to the month before the announcement, with frequent flashbacks to the youth and early careers of the principal characters and to the IBM 360 glory days. The emphasis is on named IBM managers. Killen casts Earl Wheeler, VP of Corporate Programming and the leader of the SAA effort, in a role corresponding to Kidder’s J. Thomas West, who led the Eagle team at Data General. SAA, why IBM needed it, and what it is are described in loving and effective detail. Killen treats IBM’s decision to do it as a bet-your-company event equal in significance to the 360 decision of 1964. His description of how IBM acts to create and get behind something as big and bold as the SAA shows a remarkably good grasp of what SAA is and a perceptive view of the superficialities of IBM’s corporate culture and the physical trappings of its powerful people. However, it shows very little comprehension of what goes on at and between the programmers’ workstations, where the nuts-and-bolts decisions about SAA had already been made long before the series of meetings he describes took place. The resulting story is full of diverting and diversionary detail and gives the impression that the most important things at IBM are the overhead projector foils and the precise wording of presentations to management committees.

I doubt if this is “the exceptional insight into IBM” or “the first book about IBM’s top executives working in their inner sanctums” that a jacket blurb claims, but it does give some idea of the Chinese fire drill that is necessary when a huge bureaucracy that is spread all over the United States and finely divided by function makes a big move that cuts across the boundaries of all the vested fiefdoms. A reader who is more critical of IBM and its works than the author can see from what he says why SAA is described in some quarters not as the greatest gift to the user since the punched card but instead as a huge but necessary software fix to cure the consequences of two decades of unregulated, uncontrolled, and unstandardized hardware and software proliferation in the IBM product line.

I cannot recommend the book as a good history. It is not clear what is fact and what is fiction, and the freewheeling expressions of the players’ thoughts, vital to the history of an event like this, appear to be based on little more than the author’s fervid imagination. On the other hand, as Killen says, this may be the only history of the making of SAA that we will ever get.

I can recommend the book to those who can’t read enough about IBM or to those who are willing to wade through a lot of peripheral writing to find a good intelligible description of SAA. The book is superior to Kidder’s in one important respect: it has an excellent index.