Over the years, many other legends have sprung up about Unix; and the history of its growth and development has been obscured. Salus's book, published in 1994, the 25th anniversary of Unix, is a successful attempt to clear away the rubble and find out what really happened. A look at the history of Unix is important, for both Linux and free software in general have their roots here.
Salus himself gives a succinct definition of the aims and content of his book when he writes in the preface, "Every book I have seen on modern operating systems or Unix has an obligatory two or three pages on the history of the system. By 1992, I had read about two dozen of these. None was completely accurate. Some were hilariously in error."
There is a wealth of classic quotes. Here are just a few, to give you the flavour:
Kirk McCusick on getting 32V running on the VAX: "This was fairly straightforward provided that you had exactly the same hardware as the release tape. There was no notion of auto-configuration. You had to have exactly what they had, you couldn't have anything extra, and you couldn't have anything missing. There has to be precisely two disks and they had to be RP06s." (p.156)
Brian Redman at USENIX 1979, p.193: "some guy got up and they booed him off the stage because he was a marketing consultant or something. Gee, I thought, these guys are serious".
Sunil Das: "Technically, Unix is a simple, coherent system which pushes a few good ideas to the limit".
In June 1979 came Version 7 Unix, perhaps the definitive version. Look at the commands we owe to V7: awk, find, m4, make, tail, sed, to name but a few. Just think for a moment of what it would be like to use, or try to use, a system — any system — which didn't have those available. Perl is surely inconceivable without its predecessors: awk and sed, as well as sh. Version 7 also introduced a new, restrictive, Unix license. One of the consequences of this was that a certain professor of computer science was moved to write his own replacement for the source he could no longer use as he wished. To quote Salus: "Tannenbaum's solution was to write a new operating system from scratch that would be compatible with Unix....Tannenbaum called it MINIX". The rest is history.
Unix was for much of its life free software. You got in touch with Dennis, he sent the tapes, you hacked the system to make it run on your hardware, your changes maybe got fed back into the source tree. This attitude continued into the early days of BSD. However, with the licensing changes, came legal difficulties. The account of BSD's legal problems is good background reading for those who have difficulties understanding Stallman today. To make it plain: this is what the GPL was designed to prevent.
One small criticism occurs to me. This history of Unix is also Linux prehistory, so to speak: but there is nary a mention of Linux. Perhaps that is not so surprising; the book, while published in 1994, will have been written in 1993. In those days, Linux was really just a enthusiast's system. Still, no mention at all? A footnote at least would seem indicated. No doubt there will be a whole Linux *section* in the notional "50 Years of Unix" in 2019?
In conclusion, this book recounts not only twenty-five years of Unix history, but twenty-five years of free software history, our history. Indeed, it is the definitive history of Unix to date.
Paul Dunne 2001
Copyright © 1995-2007
Paul Dunne,
Sponsored links (requires javascript):