I’ve penned – well actually I’ve pecked – an article on open source software. Its not yet been accepted, so I thought I’d see if anyone wanted to read my draft and offer comments before its too late. I should have thought of this before – but there you are – I didn’t.
When I found out about open source software it really excited me and surprised me how few people really knew much – or even anything – about it. It seems to me a genuinely new form of production which as I say in the essay comes “with the additional excitement of bridging ideological divides between left and right”.
I’ve posted the prologue of the essay over the fold. If anyone wants to read on they can email me at nicholas AT gruen dot com dot au . Comments would be appreciated as the essay has not yet been accepted.
________________________________________________________________
No scientific inquirer can keep what he finds to himself or turn it to merely private account without losing his scientific standing. Everything discovered belongs to the community of workers. Every new idea and theory has to be submitted to this community for confirmation and test. There is an expanding community of cooperative effort and of truth. It is true enough that these traits are now limited to small groups having a somewhat technical activity. But the existence of such small groups reveals a possibility of the present. . . . Suppose that what now happens in limited circles were extended and generalized. . . . The general adoption of the scientific attitude in human affairs would mean nothing less than a revolutionary change in morals, religion, politics and industry.
John Dewey
__________________________________________________________________
Prologue
Half way through Peter Weir’s thriller “Witness” a small rural community comes together for a barn raising. As Maurice Jarre’s synthesised monotones transform themselves into a symphony, so before our eyes the lives of those we see are transmogrified in a single day into thing of simple and enduring utility, a monument to the miracle of human co-operation and endeavour.
Those viewers whose hearts have not somehow been turned to stone marvel at what they see. They have one of those moments when they wonder why life can’t be more like this. Perhaps some contemplate how to make it so. Then their reveries fade and they return once more to life in its fallen state.
The scene would be even more stirring if, once built, the barn would last forever, be freely available for anyone, indeed everyone to use and to improve upon with those improvements being likewise available to others. More stirring still if this was just the beginning, so that once the barn was raised, the resources of large profit driven corporations were somehow drawn, quite freely into the same process of maintaining, refining and extending the building for their own and the common good.
If something so extra-ordinary occurred, some suspicious souls might wonder who might try to stop it. They might expect those used to doing things the old way to use all the instruments at their disposal to defend their interests. And what better vehicle to use than the laws of the land both those that existed before the new methods proved their worth and new ones that the powers that be might be persuaded to introduce.
All this both the good and the bad is happening in the world of open source software.
Postscript: I’ve now posted the paper publicly here.
Nicholas
Why not post the rest of the essay using the MT upload facility (see the buttons at left of the editing/posting screen)? Then you could delete the link to the upload page if and when your essay is accepted for publication (if required by the publisher). In the meantime, getting feedback on the entire essay must surely be better than the short teaser you’ve posted so far.
Nicholas,
I began my career as wannabe ideologue through opensource. Richard M Stallman (RMS) and Eric S Raymond (ESR) were my first intellectual heros.
As time has gone on I have moved past the starry eyed position on open source and free software. In particular, I specifically reject Stallman’s claim that “all software must be free”, as an unjustifiable rent by the consumer of software on its producer.
Opensource cannot, however, unite left and right. One only need spend about 5 minutes reading Slashdot to see that left and right opensource types fight like cats and dogs. Most, if not all, active members of the opensource community behave and write as though FLOSS is orthogonal to mainstream politics. It has a self-contained politics with no reference to the outside world. There also seems to be a strong norm to maintain that rule.
So for instance, when free software geeks argue about George Bush, they “step outside” to places like this, Usenet, Kuro5hin.org or Daily Kos. But when Linus Torvalds decides to use closed source software in the most intimate development of the Linux kernel, the doors of Bedlam boom open hollowly and pandemonium is unleashed on the Linux Kernel Mailing List, Slashdot, and a thousand other mailing lists, Usenet groups and websites.
The reason that mainstream politics and political philosophy has not engaged or absorbed FLOSS would probably be:
1. Barriers to technical understanding. Very few people understand what source code, compiling, linking, kernel, mailing list, CVS, etc mean.
2. Denial of conventional models for understanding. FLOSS has thus far managed to shrug off conventional theories and methods of explaining economic, social or personal behaviour. ESR’s first-generation attempts at a “General Theory of Open Source Production” were a sound failure. I attempted (admittedly in highschool) to examine the economic theory[1].
3. Unwillingness to engage. Many geeks are of a technical bent, and roly-poly social theorising annoys them as inelegant and irreducibly complex. They prefer programming, which is in its own way the poetic son of mathematics.
I’d be happy to comment on your article further. I’ve been a “FLOSS watcher” since shortly before ESR published “The Cathedral and the Bazaar”, which essentially got the ball rolling on the navel gazing bandwagon. Just send an email to my address above.
[1] http://kanga.nu/Members/JCLawrence/docs/oss/extess/
Addendum – having had a look at the full article, I’d recommend at it as a fairly good layman’s guide to the history, meaning and political concerns of the opensource movement.
I agree with Jacques, opensource remains largely a specialist dominated field. I have contributed code and documentation to several oss projects. If you search the Apache website for my name it will pop up with one of my younger polemic-infused rants. The specialist pyramid is alive and well in OSS.
OSS is very capitalistic in that it commoditizes that which should be commoditized. In the case of software it is applications that have very stable and well-known requirements. For instance a POSIX-compatible kernel, and HTTP webserver, a Word Processer. If everyone knows the requirements then collaborative development is easier.
Most friction in a software project is over requirements, and the design necessary to support those requirements. Having a commoditized requirements that everyone knows from their own experiences aids that process.
OSS is a different way of doing things, it pushes value down to the service level. Al-la the Red Hat business model. We did one project that used JDOM. It was a government project and get sent off to a physics lab for review. A question came back from the review, “Who was going to support JDOM?” The answer of course was annyone who wanted to, including the company I was with.
I dont think it bridges anything between left and right. Because it OSS lends itself to being specialist based, things get accepted on technical merit, not political bias. It is just a way for technical specialists to share their work, in the same way scientists share their work in journals or musicians do in a workshop/jam.
I’d suggest looking at blogs instead of, or as well as, OSS. The fundamental issues are similar, there aren’t the same technical complexities, there’s good data from places like Technorati and you know what you’re talking about.
Cam,
A comment I made to Nicholas was that FLOSS effectively drives the price of certains sorts of capital to zero, or nearly zero. It’s a massive accumulation of capital.
Indeed a paper from 2001 estimated the recreation cost for common opensource software would be more than a billion USD and around 8,000 man-years.
http://www.dwheeler.com/sloc/
Cameron’s
“Because it OSS lends itself to being specialist based, things get accepted on technical merit, not political bias.”
Caught my eye. Unlike say being accepted at an Athaeaneum Club in your capital city.
Which is a different, and dare I say, discerning accumulation of (social) capital.
I mention this because there are other conditioning effects outside of the Mr Economy trying to commodify some thing or other, I’ve no idea what they will be, but I hope they preserve those Atheaneums in perspex along the way.
the other thing is reading the comments here is that I am actually beginng to feel old… compared to yesterday…
…anyway..
I suppose the languages we speak are open source, with requirements described by the spec(ifications ) of baby brains wanting to learn to speak, which, of course is a bias we all share.
Another bias we all share (though to a much less degree in psychopaths) is a bias to socially integrate. OSS is part of that.
Mr Economy don’t know about them. Whereof we do not know, thereof we cannot speak.
Mr Economy describes psyhopaths who have to tolerate and deal with each other in a noblesse oblige kinda way at best. Rational perhaps, reasonable? Mmmh?
OSS may not be rational, but its produces reasonable products.
Of course these whereof-we-do-not-knows still have impacts in the world outside of the real world of the market…
…ramble ramble ramble late at night… we despatched five roosters today and made sweetmeatloaf with gizzard. liver & testes…
Does the main article (have I missed the link) use less economic metaphors to look at OSS?
I had a go at this in the Fin a while back
URL didn’t come through
http://www.uq.edu.au/economics/johnquiggin/news/2005-01-27-AFR.htm
Jacques, I think gcc is an extremely important project. It reduced the cost of construction to zero. It wasnt that long ago that Microsoft and Borland charged hefty prices for compilers. It commoditized the compiler to zero cost.
Meika, OSS is important but remains specialist dominated. It is a means to accumulate specialist knowledge, and allows specialists to interact with that knowledge in an applied manner.
But it remains a specialist art form with high barriers to entry. Not everyone can write source code that compiles. It is even harder to write well designed source code; that compiles, passes integration tests, complies with an existing project’s design and convention of doing things.
Note though that you dont have to have a University degree to be able to submit a patch. You just be technically competent enough to create working patches. IIRC there was some twelve year old kid that was a committer to the Apple Darwin project which was based off the BSD project. I think there was an issue as he was under eighteen and Apple got upset about it, bumping him off as a comitter. If he was committing to the BSD project, they would not have cared. He would have been judged by his output, not his age or qualifications.
Compare blogging/internet-publishing to OSS. Both have a zero cost of construction. Both have near zero cost of distribution. Yet the barriers to entry for blogging are lower. Anyone who can publish, can do so. Anyone who wants to comment on a blog can do so. It is a non-specialist, OSS by contrast, is specialist.
Neither are a guarantee of quality. While articles on OSS tend to focus on OSS’s poster-boy, Linux; a quick search on freshmeat.net will show just how many OSS projects there are out there.
John;
Though I’m sure you’ve received extensive email about this since the article was published, but Unix is not a programming language. It was and is an operating system.
As a term it’s generally used as a name for a family of related operating systems, including direct descendents and systems which take design and inspiration, but not code, from the original line.
“Unix” is also a trademark, so strictly speaking there is only one operating system which is actually Unix.
open source software has a number of out and out winners on the board. apache and php to name two. these open source software products have destroyed the marketplace. there is no longer a market place in these. Nich’s thesis that oss collapses the right-left debate has merit. It renders nugatory the whole idea that X requires a market-based solution. X becomes as public domain and as “democratically” modifiable as any other public idea.
Software and open-source is a natural fit. Microsoft is the exception that proves the rule.
It’s only when spruikers for software houses and people who have no idea what software is become involved that the argument becomes at all muudy. There is no debate here. History is moving faster than the speed of any polemicist’s pen.
The horse-power of the mega companies is naught when compared to the synergistic power of the networked open source dev community. And let nobody mistakenly assume that these people are ideologically driven. They do this because the project drives them. It is their community. They need the OSS project for their own spiritual nutrition. They are being paid. That’s the kicker!
Thanks all for your comments.
To clarify a couple of points. I wasn’t planning that OSS would bring about peace and harmony between left and right. (Perhaps Jesus will be able to do that on his second pass by the planet). I just suggested that the unique way in which OSS simultaneously addresses both community and private interests is exciting. I got excited by it anyway.
John, as I argue in the essay itself, I don’t agree that blogs or Wikipedia and similar ventures are examples of the same kind of thing as open source. They are of course in some ways of talking, but for my purposes
wbb,
Open source is not democratic: not everywhere. There are clearly defined boundaries of property and control, which are self-enforced.
Some projects are anarchic. Some are dictatorships. Some are radically democratic. Some are collegial. Some are like clubs, requiring sponsorship.
All of them have source code management and secured servers holding their main source trees; all of them therefore commence from a sort of property.
What’s different about open source is that software is non-rivalrous. The controller-owners of it can grant infinite easements to download sourcecode. What very few of them grant is unlimited access to write to the main trunks of code.
I think that the other thing that’s oversold is opensource as a process. Opening the code does not solve hard problems of design, code construction, testing and the allocations of work. This incidentally why opensource projects thrive on technically interesting problems – advanced memory managers – while “boring” nuts and bolts stuff tends only to be picked up by the paid programmers.
JWZ said something about this, if I could only remember where.
Nicholas, I think you over-state IBM there. It is only recently that they decided to do a marketing campaign to push opensource and even then only “Linux”. I think Log4J is the only IBM opensource library I have used. Sun has been a big backer too, openoffice is from their own staroffice. I did the SSR book in Openoffice.
IBM is seeking to commoditize their complement. They sell hardware and services. Having low cost, high quality software means they dont have to make any odd licensing deals with software makers. Especially ones that have been adjudged an illegal monopoly in the US. Also means they dont have to sink development money into AIX for running on dinky i386 platforms.
Also if you have a look at all the opensource projects, Linux is just one of many, despite being the sexiest in the media. I am tapping this out on an iBook. OSX uses Darwin(BSD) for its kernel. I am also using emacs as the editor. SSR/Newcopia run on Linux though.
Apple has also embraced opensource, but only where it commoditizes their complement. They too are a hardware supplier. They differntiate themselves in the market by their hipness and usability. OSX is awesome for that, and unsurprisingly is proprietary. Yet the rendering engine for Safari is KDE’s khtml. Browsers are commodities. Musical experiences are not, and iTunes remains proprietary as well.
But those remain big projects, with big sexy names. freshmeat.net has umpteen and oodles of opensource projects. All filling some niche.
If science did go opensource it wouldnt create some kernel, it wouldnt start from the middle out. It would just be a bunch of people, unusually interested in something and then sharing what they know without care or recourse. They might just require that if you use their publications, that you publish any derivative of that work.
IMO it is important not to see opensource as just Linux.
As to the blogs issue, they have effectively replaced the op-ed sections. I dont read those in the newspapers anymore. I come to places like here for that kind of commentary. I expect blogs to innovate and segment more, replacing the mass media entirely one day. Opensource may do that one day too, it depends on how capital intensive software development remains/becomes.
Fair point Cameron,
But ‘boots and all’ was intended to refer to the fact that IBM have as I understand it about 600 OSS coders. But I’m not suggesting they are wide eyed supporters of OSS as such to the exclusion of any money making opportunities they see around them by closing their source code.