After the Software Wars (Conclusion)
Creativity and Compensation
Curtis has come to a few conclusions about creative compensation that don’t flow for me. He states that, while software, being science, should be free,
Fortunately, I have concluded that there is an important distinction between static and dynamic intellectual property. A song, once created, is not typically edited by a user. Software is often edited by a user because it often won’t really work until it is fixed to support an individual scenario… Books, music and movies are often made by one person, or small teams of people, and are meant to be used as-is without the need for modification… Writing has been a way to make a living for many thousands of years, and I am not ready to propose that we tweak society at such a fundamental level and remove longstanding incentives to create.
I don’t think it flows, first off, that those who choose to write, using the “science” of language and symbols which so many others before us have improved, are exempt from the rules Curtis would apply to programmers and geneticists just because “writing has been a way to make a living for many thousands of years.” (Actually, that’s not true – the ancient historians weren’t compensated, the ancient bards were compensated for their performances not their content, which was “open source” in its own way, the Iliad and company being modified by each person singing the song. Only since the printing press have writers had the opportunity to profit, and really only since the 17th/18th century and the birth of newspapers and pamphleteering have they been directly paid for their output.) The question has to be, in my mind, whether the unique product of an individual’s genius is her property or not; the content channel cannot affect the philosophical basis of the decision.
And as far as books and movies being “meant to be used as-is without the need for modification,” I could think of lots of novels that needed plenty of modification, and I’m just the man to fix them! How about that ending to No Country For Old Men? Yeah everyone dies but the bad guy, I could fix that for ‘em. The idea of content being used “as-is” is in fact less and less true as it all becomes digital – music is mashed up and remixed; Gnarls Barkley’s Black Album becomes a dynamic reworking of “static” content. Fanfic takes J. K. Rowling’s static conclusion to the Harry Potter series and replaces it with Harry becoming a professor at Hogwarts (which seemed to me the only logical conclusion, and the one I would have written) – or gives Hermione a bizarre sex life. Video artists take Back to the Future and Brokeback Mountain and create Brokeback to the Future. Do the Beatles and Rowling and Intermegaglobal Consolidated Entertainment Keiretsu reserve the right to maintain the static state of their creative properties?
Curtis suggests that “Musicians could start selling the data files they use to make their songs. This would allow people to enjoy the music of their favorite musicians, and re-mix it in new ways.” But the problem with that is that, it being a data file, you make one sale, and then it’s on Bittorrent. You can make one sale of a digital product and then income on all future copies may be and probably will be zero. Honestly, I don’t know how we can prevent all content from becoming “free,” but I don’t think we can say software has to be free but duplicatable art is exempt.
The Open Source Movement
The best part of this book to me is the section where Curtis addresses the OS movement with best practices for moving forward and gaining adoption by users like me:
I believe the free software community ought to make their top focus a fully functioning Out of the Box (OOB) experience. People need to be able to take a Linux CD, stick it into a new, or old, computer, have it recognize your hardware, install everything, and migrate your data…It is important that free software programs, and their codebases, be approachable. Wikipedia has a motto of “Don’t bite the newcomers” and it is a great motto for every free software organization… For example, the most popular high-end graphics editor in Linux, poorly named the “GIMP”, is as rich as Photoshop and suitable for professionals, and completely free! However, like Photoshop, it isn’t particularly approachable for a new person to jump in and start using. The first time you use GIMP, it might take 20 minutes to figure out how to crop an image. It takes time to learn any powerful tool, but a lesson in engineering is: “You only pay for what you use.” Simple things should be simple to do, and today this is not always the case.
Politics are part of all group efforts, and Curtis tells the story of two versions of Linux – Debian and Ubuntu, highly similar products which suffer because the highly similar code base and bugs are being worked on by two sets of developers. Curtis urges Debian and Ubuntu to put their differences aside and merge into a stronger, larger and therefore more effective Linux distribution. Microsoft, he notes, maintains market share through practically giving software away to maintain its installed base in the face of Linux (and, some Slashdotters maintain, by allowing a certain level of Windows piracy to thrive). Curtis cheers the nature of the all-volunteer armies of OS:
There are benefits to a flat command structure: people are forced to convince others to do something based on the merits of their ideas. Because decisions are made on their merit, issues generally resolve themselves, with the person doing the work making the ultimate decision — free software is sometimes called a doocracy. Everyone wants to do their best, and if they don’t do a good job, bugs will come in and people will complain, so the situation will usually reach an acceptable solution.”
But again, this presumes mathematical interactions between individuals, wherein the more elegant solution on one side of the equal sign must be accepted by those who’ve written the solution on the other side. The “free market of ideas” doesn’t take into account the human factors such as ego, ideology, cult of personality, factionalism, etc. that come into play in any passionate community organization and skew the results off the strictly logical. Only a community of open source Vulcans will ever make decisions based purely on logic.
These schisms and sectarianisms are to be expected – when everyone is a volunteer, everyone has more say in the project than they would as employees, but this leads as often to conflict in groups as to resolution. Schism is a byproduct of any religious or political movement, as each set of true believers fragment into smaller groups with the “righter” ideas and increased hostility to the unbelievers (i.e. Debian v. Ubuntu). Progress is arrested as everyone competes to have their idea adjudged the winner. I read a book last year called Dreaming in Code about an open source effort – despite the author’s game-faced attempt to tie it all up as being “about the journey,” the fact was the story didn’t succeed as he’d hoped because the software never got done and therefore the book had no conclusion. Everyone sat around and discussed how it should be done or redone or re-libraried or redocumented, went and got a sex change in the middle, came and went as they pleased, and since nobody’s livelihood depended on making a deadline, deadlines didn’t get made. Top-down management with its firm deadlines and decisions makes the trains run on time, and profit incentives make those deadlines of value to those doing the work, but inarguably the side effects of poorly run autocracy (Dilbertarian management styles, Six Sigma Scientology, “brand identity” hooha) can derail a project just as easily .
That Slashdot posting I mentioned earlier had a comment that I think Curtis should incorporate in his book, because it has the ring of truth for me (emphasis mine):
In general, every widespread product drives gradually towards commoditization, where profit margins are driven to the lowest levels that capital sources will allow (any lower and capital flees to better ROI). But the marginal cost of software is zero; no capital is necessary to produce additional copies. For custom and niche software, there are too few people interested in the software to drive commoditization, so the software can be profit-generating. For software that is consumable content and differentiable, like games, there is opportunity for profit. But for stuff that everyone needs, and which provides few opportunities for differentiation, commoditization is the natural course.
Without the presence of open source in the market, the commodity price level would remain above zero, even once the software was “perfected” (meaning real differentiation is no longer possible), but if open source enters the game, as long as it is sufficiently functional that it is cheaper for some individuals or organizations to fix the ways it fails to meet their needs rather than buying commercial software that does, then the open source code will continue to improve, which increases the segment of the market that finds it acceptable, or nearly so.
Without the presence of open source in the market, the commodity price level would remain above zero, even once the software was “perfected” (meaning real differentiation is no longer possible), but if open source enters the game, as long as it is sufficiently functional that it is cheaper for some individuals or organizations to fix the ways it fails to meet their needs rather than buying commercial software that does, then the open source code will continue to improve, which increases the segment of the market that finds it acceptable, or nearly so. And remember that cost isn’t just dollars. Ill effects of monoculture, perceptions of vendor lock-in and even just plain dislike — even if irrational — are all “costs” that some people apply to commercial software.
So the proverbial MRI software remains profitable to the few who code for it, but the “franks and beans” of operating systems and word processors become like blue jeans – all pretty much the same with only some layer of marketing “added value” bullshit about how only Calvin Klein can make your ass look good to drive customers from one to the other.
I think open source is a greater good; I think it is essential to progress in areas like AI. I agree 100% with Curtis that “The reason we haven’t solved AI is that there are no free software codebases that have gained critical mass. Far more than enough people are out there, but they are usually working in teams of less than five, in old codebases.” This, after all, is why I’m making Christopher’s as-yet-unnamed brother an OS prophet in the novel, and OS will have a huge role to play in part two of the book, when Alex becomes a for-profit AI based on the work of so many OS volunteers (sometimes I forget all this reading and blogging really does have a literary purpose, in establishing character and motivation for the novel that I’m writing!).
But I also know that to succeed, OS has to have a clear and coherent underlying philosophy that takes into account economic realities and human nature; that it needs to be less loosey-goosey in its application of its principles (i.e. software v. art); that it needs to shed metaphor as argument (trees and loaves of bread, etc., and Eben Moglen’s argument that creativity is some kind of inevitable natural phenomenon like magnetism, that will happen whether or not people are compensated or rewarded – see under Ages, Dark for what happens to creativity and innovation when society crumbles economically).
And I think we need to work on establishing a successful economic model for OS developers, trainers, support staff etc. Messianic talk about joining the Movementarians and giving all your property away to sing Hare Krishna and save the world will inevitably appeal to a small core of those who (like the Islamic engineers noted yesterday) need a world-bending mission to motivate them. But the other 98% of the world needs to know that they’re not committing to a life of sackcloth and ashes. And I don’t think just throwing “services” out there in a vague form will do the job – there need to be compelling, Malcolm Gladwellesque stories about gee whiz look what all these people are doing in the new paradigm! That brand of evangelism is not the kind of thing engineers are comfortable doing; they think that if they lay out the facts, that should be sufficient, but it’s not enough. Joe Software Consumer reading the New Yorker on the plane needs Malcolm Gladwell to tell him how awesome Linux is and how many cool people are doing super stuff that’s neat. So we need stories, because stories motivate more people than reasons.
[Talk about timing: Just about half an hour ago, Slashdot opened a thread asking “Without Jobs, Will Open Source Suffer?” It asks its readers, “Have you thought about scaling back open source work as the economy continues to contract?” I’ll repost some of the money quotes tomorrow after this thread has run its course.]
I’ve been highly critical of what I’ve read so far because someone has to play Devil’s Advocate; and I haven’t seen anybody take the role. Nobody in OS is challenging the thinking of anyone else, and the only opposition to OS has been from closed source, and the only reasons given have sounded like the seagulls in Nemo – “Mine! Mine! Mine!”
Thanks to Keith Curtis for his feedback on my postings, which you can see in the comments section of each part of this review. I never intended to start this project by researching OS so thoroughly; I imagined I’d go through all the AI stuff first and then circle around to OS when I got to part two of the novel, when Christopher’s brother comes into the story. But that’s the creative process for you – it has a way of setting its own timeline. I’ll write more about OS in the future as it comes up, including at some point Against Intellectual Monopoly, but my next goal has to be another chapter of the novel, and digging deeper into the hard stuff like Computational Linguistics and hard AI again.