It used to be a surprise when one learned how old and venerable UNIX is. No more: UNIX seems to have found its place--not as prominent as its advocates hoped, but a secure place nonetheless--in a desktop environment dominated by MS-DOS, Windows, and soon Windows NT. Salus has done a great service by interviewing many of the players who gave us the several flavors of UNIX and presenting a lively narrative history. (In addition to his own interviews, Salus has gathered significant interviews from other researchers, especially Michael Mahoney of Princeton University.)

There have been other histories of UNIX, some written by its original creators and published in technical journals, others introductory chapters to well-known UNIX textbooks. But these too often present only one corner of the picture. Salus is especially strong in his chronicle of how UNIX moved--and changed as it moved--from AT&T in New Jersey to Berkeley, and then to Sun Microsystems by way of Bill Joy. Among the results of that cross-country journey was the introduction of TCP/IP into UNIX and therefore its close identification with the Internet. Few other histories include this aspect of the story, without which one simply does not understand UNIX. Salus also tells the fascinating but ultimately painful story of how Digital, on whose machines UNIX was developed and with whose VAX Berkeley UNIX was most associated, squandered and surrendered its advantage to Sun and the other workstation manufacturers.

My only criticism of the book is that the author often lets the person he interviews “speak for himself,” forgetting that sometimes the reader needs a little more context, or that what is being reported may be a topic hotly disputed by others. Most of the programmers seem to be wonderfully self-effacing, which is a refreshing contrast to the big egos one encounters elsewhere. But that leads to statements along the lines of “I asked him what that gizmo was and he said it was a terminal. So I sat down, ordered a pizza, and that night I wrote UNIX version 7.2, with windows, a database, and a text editor thrown in for fun.” Perhaps the creator has difficulty articulating the creative process, but it is not too much to ask that the author make an attempt.

The story of UNIX is full of ironies: from modest beginnings as something for Thompson and Ritchie’s own use, it found its way into general use for workstations, sprouting a number of incompatible, baroque, and feature-laden versions. It spread because AT&T, a regulated monopoly at the time, gave it away; but after divestiture, AT&T failed to profit from its creation. Though UNIX was developed on Digital computers, Digital also failed to capitalize on its lead. Meanwhile, MS-DOS, an inferior operating system, became the real standard for desktop computing after 1981. This book ultimately does not explain why all that happened, but it goes as far as any account published so far.