The Daemon, the GNU and the Penguin - Ch. 15, by Dr. Peter Salus

by Pamela Jones

July 07 2005

Here's the next installment in Peter Salus' ongoing book, The Daemon, the GNU and the Penguin, "Commercial UNIXes and BSDI" -- Chapter 15. It's longer than usual, and I suggest you savor it, because there will be a short break until the next installment, on July 28.

Here are the earlier chapters of Dr. Salus' book:


The Daemon, the GNU and the Penguin

~ by Dr. Peter H. Salus

Chapter 15. Commercial UNIXes to BSDI

In the 15 years following the release of V6 (April 1976), Berkeley was not the only place where versions and clones of UNIX sprouted. While I doubt whether I can even enumerate all of them, the following will give an image of the geography of the field. To me, the most significant UNIX releases were:

I've mentioned Minix and the AT&T, BTL and BSD releases earlier. But several of the others are worth devoting a vignette to them.

Interactive Systems

Interactive was founded by Peter Weiner in 1977. (Weiner had been Brian Kernighan's Ph.D. advisor at Princeton.) In 1978, Heinz Lycklama joined him in Santa Monica. Lycklama had just written LSX, a version of V6 UNIX for the LSI-11 microprocessor. Interactive's product was called IS/1 and ran on most PDP-11s. Interactive's UNIX was an important product for nearly a decade. In 1985, Interactive's IN/ix became the basis for AIX (announced 21 January 1986). Some of the later modifications to AIX were developed by Interactive under contract to IBM.


Cromix was a proprietary UNIX clone of CROMEMCO. The CROMEMCO 100 ran on a Xilog 80 and had 512K of RAM, 50M of hard disk, and an XPU processor, enabling 32-bit processing. Founded in the early 1970s by Roger Melen and Harry Garland, Stanford students who lived in CROthers MEMorial Hall, it was incorporated in 1976. In 1985, it was bought up by Dynatech, and disappeared. But Cromix was the first UNIX clone. The CROMEMCO 100 and 300 ran both Cromix and System V. The 300 ran a 68000 timesliced with a Z80 coprocessor to enable multiuser CP/M WordStar.


Technical Systems Consultants wrote a drive for the then-new 5.25" drives in 1976: DOS MiniFLEX. It was superceded by FLEX for the 6800 a few months later. FLEX was adopted by virtually all of the 68xx SS-50-based computers (even the Tandy Color Computer and the UK Dragon). TSC now turned to producing a UNIX-like multi-user for the 6809: UniFLEX. It was a failure because of the introduction of 16-bit processors and the PC.

Microsoft XENIX

Microsoft licensed 7th Edition from AT&T in 1979. On 25 August 1980 they announced that XENIX would be available for 16-bit processors (Microsoft couldn't license the name, "UNIX"). XENIX wasn't identical to 7th Edition because Microsoft incorporated several features from BSD.

Microsoft didn't sell XENIX: it was licensed to manufacturers who were responsible for the porting. The first ports were to the Zilog Z8001, a 16-bit processor. Altos shipped one in early 1982. Tandy shipped one for 68000 systems in January 1983 and SCO released their port to the 8086 in September 1983. The license had been for V7, XENIX was based on System III.

XENIX 2.0 (1985) was based on System V, and added support for 80286. However, Microsoft apparently lost interest in XENIX after signing an agreement with IBM to develop OS/2. In 1987 Microsoft transferred ownership of XENIX to SCO in exchange for 25% of the company. That same year, SCO ported Xenix to the 386 and Xenix 2.3.1 supported SCSI and TCP/IP.

Xenix became SCO UNIX in 1989.


P.J. [Bill] Plauger received his Ph.D. in Nuclear Physics from Michigan State in 1969. From 1969 to 1975 he was a Member of Technical Staff at Bell Labs. Together with Brian Kernighan, he wrote Elements of Programming Style (1974) and Software Tools (1976). He also writes science fiction, and won the 1975 John W. Campbell Award as the best new SF writer of 1975.

It was while writing Software Tools that Plauger left the Labs. He told me this:

I ended up leaving the Labs. I felt I didn't have a future there and that I'd better move on before they told me to move on. And I was able to get a job at Yourdon . . .

After a few years of traveling all over the world lecturing, I felt that I wanted to get back to programming. Ed [Yourdon] had an opportunity to get a contract to do a commercial C compiler, and I talked him into doing it. I worked around the clock for a week. . . .

Plauger went on to form a three-man company, Whitesmiths.
I think we started on August 1st, '78. We were going to sit down and write a C compiler from scratch -- my third C compiler, I guess. I paid a lot of attention to not having any notes from my Lab days or my Yourdon days. . . I wrote like a fiend and by the end of November, we had a compiler.
Whitesmiths' first compiler was for Fisher and Porter in Philadelphia. It was for the PDP-11. "We gave them an 8080 compiler by the middle of '79; a VAX compiler by the end of that year; and we gave them a 68000 compiler in the middle of 1980," he said. "And we were doing Idris at the same time."

Idris was a UNIX-like multi-user multi-tasking operating system, written by Plauger and M. S. Krieger. Originally, Idris ran only on the PDP-11. But it was soon ported to the VAX, the 68000 and the 8086. In 1986, Atari hired Computer Tools International to port Idris to the Atari ST. Whitesmiths was sold to Intermetrics in 1988.

Mark Williams Coherent

The Coherent Operating System from Mark Williams was a UNIX-like OS for PCs. It was introduced in 1983. As I knew that several former University of Waterloo students had worked on it, I asked Tom Duff. Here it is, in his own words:

I was at Mark Williams from roughly August 1 to October 31 of 1980. After leaving the NYIT Graphics Lab, I had 6 months free (later reduced to 3 months) before I was scheduled to start at Lucasfilm. Mark Williams CEO Bob Swartz heard that I was available and asked if I'd like to work in Chicago for a while.

When I arrived, they had a working C compiler, assembler and loader and a version of ed, written by Dave Conroy, hosted on RSX.

Randall Howard was doing most of the kernel work. Johann George, David Levine and Bob Welland were also there, but I'm not sure what they were working on -- Johann was probably doing kernel stuff.

Dave Conroy, Randall, Johann and I were all friends at Waterloo in the early '70s.

This was an amazing crew: Dave Conroy most recently was in charge of engineering the Mac Mini, Randall founded MKS, Johann founded Sourcelight Technologies (Randall and Johann are both semi-retired VCs now), David Levine wrote a legendary early video game called Ballblaze and Bob worked on the design of a bunch of important Amiga hardware.

When I arrived, it was pretty clear that the kernel was pretty much taken care of (though it wouldn't be running well enough for daily use until after I'd left), but nobody was working on user-space stuff. So I opened the 6th edition manual to page one and started implementing commands. In the three months that I was there, I think I did did A through M, (As I remember, I started at make, then jumped back to ar and just plowed through. I remember make, diff and dc being a lot of fun.

And, I did units, because the library research required to dig up the more obscure quantities seemed interesting.

While I was there, Ciaran O'Donnell (another friend from Waterloo) visited for two weeks during which he wrote, in a feat of coding acrobatics such as I have never seen before or since, a complete, functioning YACC clone, working just from Aho and Johnson's 1974 Computing Surveys paper.

Coherent eventually ran on most 286, 386 and 486 boxes. It actually had support for X11.

The Mark Williams Company went bankrupt in 1995.


A/UX was Apple's entry to the world of UNIX in 1988. It was based on SVR2.2 with element of SVR3 and SVR4 as well as 4.2BSD and 4.3BSD. It is POSIX and SVID compliant. From A/UX v2 on, it included TCP/IP. The last version (3.1.1) was released in 1995.


A UNIX-like kernel based on Mach (CMU) with many BSD features and display PostScript with a windowing engine lay at the heart of NeXTSTEP. Previewed several times beginning in 1986, it was released on 18 September 1989. The last release (3.3) came out in early 1995.

There are, of course, many other UNIX-like things one could talk about, but I never found Trusted Xenix nor the RISC version nor Compaq's NonStop-UX very interesting.

Dr. Salus is the author of "A Quarter Century of UNIX" [ ] and several other books [ ], including "HPL: Little Languages and Tools", "Big Book of Ipv6 Addressing Rfcs", "Handbook of Programming Languages (HPL): Imperative Programming Languages", "Casting the Net: From ARPANET to INTERNET and Beyond", and "The Handbook of Programming Languages (HPL): Functional, Concurrent and Logic Programming Languages". There is an interview with him, audio and video,"codebytes: A History of UNIX and UNIX Licences" [ ]  which was done in 2001 at a USENIX conference. Dr. Salus has served as Executive Director of the USENIX Association.

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivs License. To view a copy of this license, visit or send a letter to Creative Commons, 559 Nathan Abbott Way, Stanford, California 94305, USA.

07:15 AM EDT

Copyright 2005 -