The League of Extraordinary OSes

Every now and then I get handed a paper and asked for feedback. Most of the time, these papers have long boring titles and lots of funny charts with red error bars on them. I've even written a few of them myself, so I guess that's why I get handed more and more. But go ahead - I enjoy another article on clustering (Gordon Bell handed me one a while back) or TOE's (that was one of the SiliconTCP boys). I'll even look at the "let's drop TCP and make it fast" papers, because sometimes the authors are actually pinpointing a real-world problem even if I don't agree with their proposed solution.

But I was recently referred to a paper "Open Innovation: The Paradox of Firm Investment in Open Source Software" by Gallagher and West precisely because it related to the evolution of open source, and since I'm often referred to as a "Pioneer of Open Source" with 386BSD, this is another topic on which I get requests for feedback. So I read the paper discussed, and found their discussion of proprietary work quite good. However, I also found that their isolation of time in studies (1998 onwards) actually missed the primary evolution of every single open source model cited in the paper, which misses the point of the exercise, likely due to ignorance of the topic. So perhaps an examination of Berkeley's influence in this regard would be a valuable addition.

The "free software" approach of Richard Stallman stemmed from the LISP wars of the 1980s and the loss of work (see Lynne Jolitz, Copyright, Copyleft, and Competitive Advantage, among other discussions on open source and Berkeley). His most popular works (EMACs, GCC) were commonly distributed on Berkeley Unix software throughout the 1980's and early 1990's. In fact, much of the Berkeley Unix software was underwritten by DEC, ATT and others as part of the design committee work. (The term "open source" was actually coined much earlier than 1997, by the way, since we were working with Berkeley to do an open source release since 1989).

Much of the BSD releases throughout the 1980's was not encumbered, but began to use the "Berkeley software license" model which allowed it to be distributed and modified with attribution maintained. This was antithetical to the GNU model of securing changes to the source. Since Berkeley and BSD contributors were still bound by an overlaying source license from ATT, Stallman's concerns really didn't impact the BSD distributions, however. By the late 1980's, BSD support was fading as the group did not have good faculty support within the university, and it's primary port was to the CCI Tahoe machine (which was marginalized). It was then that the port to the 386 was proposed (1989 - Jolitz & Jolitz, 386BSD: A Modest Proposal, coupled with Berkeley's removal of "proprietary" code, leading to 386BSD and the X86 open source base derivitives available today.

Much of this work was supported by Jon Erickson, editor-in-chief of Dr. Dobbs Journal, a very popular (to this day) technology journal. It was noted recently by Ken Brown in his papers on open source that Linus Torvalds began his work in 1991, 3 months after reading the step-by-step series on how to port Unix to the 386, and using Minix as a basis, began the Linux project. It is likely he and others would never have begun if Mr. Erickson had not made an editorial call to do a series on this project.

So the "freely redistributable and modifiable" license they cite actually stemmed from Berkeley and the decade of research releases in Unix (kernel, applications, utilities) for universities and their dedication to academic freedom. It should be credited as such.

Much of the open source movement would not be possible had it not been for some like Mr. Erickson in the computer trade press taking a risk and publishing technical works which some said were under trade secret. In other words, the power of the press prevailed over attempts to suppress invention.

And finally, much of the work we use today is actually based on design and development of decades ago, funded by the prominent computer companies of the day at major universities like Berkeley, CMU, and CalTech, among others. That should not be forgotten.

Sadly, the prominent computer companies of today do not follow this model, but instead work in isolated "open source" communities which follow their commercial agendas much more closely than a major university would ever permit. So all this wealth of "invention" that groups from Berkeley to DEC to Dr. Dobbs Journal encouraged is unlikely to be replicated by all the "open source innovation" models linked to companies. Companies prefer to reduce risk by preventing new approaches, not encouraging them.

I hope others write serious works on the evolution of open source models - it's an interesting topic with a long and serious history. Fortunately, most of those people who actually laid the groundwork for it are still around in Silicon Valley so you can actually talk to the "source" of open source - people who actually made the decisions and did the work, avoiding current filtered or biased agendas by third parties - if an accurate picture of the motivations, goals, and history of open source is required to evaluate future trends.