A sketchy, speculative, and jargon-marred reconstruction of how IBM came to announce a corporate commitment to Systems Application Architecture. In a prefatory apologia, consultant Killen concedes he had to rely on back-channel sources rather than principals for information on the decision. Apparently hopeful of creating a crossover text like Tracy Kidder's The Soul of a New Machine, the author nonetheless spins his conjectural tale with considerable assurance, reproducing conversations and even personal deliberations that may or may not have taken place. More troublesome, though, is Killen's failure to provide context for the spurious detail he amasses. As a practical matter, SAA (which is not scheduled to reach market before next March at the earliest) represents an overdue effort by IBM to address the computability problems built into its data-processing systems--and those of the industry. In brief, programs prepared to the SAA standard should run on any of three computer families made by the company (and so-called plug-compatible units offered by copycat rivals), including PCs. Hardware as well as software difficulties are being tackled at the same time. For example, IBM engineers are working to reduce the substantive differences that characterize keyboard configurations. SAA's near-term and longer-run implications are obviously significant. Unfortunately, Killen provides almost no perspective on what the breakthrough project's outcome might mean to IBM or its competitors, suppliers, and customers. Nor, save in passing, does he have much to say about UNIX, an AT&T operating system that could give SAA a run for its money. Such information as the iffy narrative affords, moreover, is couched in high-tech bafflegab. To cite but one example: ""These interfaces will be supported across OS/2, the 3X's operating system, and the two most important 370 operating systems, MVS and VM."" A largely pointless exercise, then, that fulfills the prophecy of a computer industry acronym--GIGO: garbage in, garbage out.