Saturday, April 26. 2014
Several recent articles have pointed out the woeful lack of funding for free software development, particularly with regard to the recently discovered "Heartbleed" bug in OpenSSL -- the "secure sockets layer" that handles encrypted transactions over the Internet, used by nearly every website and browser engaged in activities where security is essential, like E-commerce (e.g., see Nicole Perlroth: Heartbleed Highlights a Contradiction in the Web). OpenSSL is one of many pieces of free software that have become essential building blocks of our infrastructure. There are lots of reasons why we prefer free software for these tasks, and indeed we would be much worse off if some company were able to corner the market for these functions and put tollgates all around them. However, there is a downside: with no company making money off a module, there is a risk that it won't be adequately supported. OpenSSL turns out to be a case in point.
I don't know of any serious work in this area, but I suspect that if economists were to take a good look at the software market, they'd find that it is intrinsically dysfunctional. The root of this is that the marginal cost of reproduction, which is what prices in truly competitive markets converge on, is zero. Therefore the only way to make money on software is to create artificial scarcity around proprietary niches. (The entire history of Microsoft, including the antitrust trials that happened and many more that should, offers a textbook on this. Smaller software companies aim for smaller niches where "first movers" tend to go unchallenged so have less need for such obviously criminal behavior.) Free software turns out to be the only viable alternative to monopoly software. Moreover, it can be developed and maintained for a small fragment of the costs that commercial software companies accrue -- even if you monetize the vast amount of voluntary labor that currently goes into it. The obvious conclusion is that the economy as a whole would be much more productive and efficient if we somehow found a collective way to fund free software development.
The obvious source for such funding is government -- you know, the institution that is supposed to be "of, by, and for the people" -- and not just in the US but everywhere. But in the US there is a peculiar ideological blindness, related of course to rampant corruption, to the very idea of government doing something that would benefit people by bypassing the corporate profit-takers. So some time ago I came up with a "plan B" and now that (at least) some people are talking again about the need to better support free software, I thought it would at least be worth a post to dust it off and air it out.
I had something very specific in mind, but you're free to generalize it or just change the specifics to your own situation: that is, after all, intrinsic to the notion of ideas, unless you try to kidnap one and lock it up behind the legal travesty known as a patent.
My proposal was that Wichita State University's computer science department should create a major in free software development as one option for its BS program (or MS, if it has or wants one). As far as I know, such a program would be unique, and therefore it would start to attract students from all over who are especially inspired by free software. (At present, nearly all of WSU's 15,000 students are local commuters, although they do draw some foreign students into engineering programs, and recently have had a pretty good basketball team.) The curriculum would combine basic computer science classes with practical project skills notably including actual participation in free software projects. So this would benefit free software in two ways: it would provide some volunteer labor for projects, and it would train people in both the theory and practice of free software -- skills that they can continue to use wherever they wind up. (Most software engineers do not go to work for proprietary software companies. Many wind up in IS departments of companies that wind up consuming and customizing rather than productizing software. Others may go into education, or on to MBA programs, or elsewhere where their free software expertise may turn out to be invaluable.)
Needless to say, there is a "first mover" advantage for whichever university does this first. Eventually, I expect that working on real free software projects will be incorporated into nearly every computer science curriculum, even without a dedicated major, simply because it is the most cost-effective way for students to learn from actual real world development. (Most students at present spend their time working textbook problems. As someone who used to hire software engineers, I put very little value on education compared to practical experience. In fact, I rarely hired people with CS backgrounds, preferring those who majored in math or science and learned to program on the side.)
I'd also recommend that Wichita State amplify the major program by creating a Center for Free Software Development, funded with business grants (and whatever they can squeeze from the government -- WSU has something like this already for aviation engineering where the local aircraft companies were very effective at lobbying the state). The Center could make it a point to hire celebrity free software figures, who would divide their time between teaching and research. And since we're talking about Wichita, it's worth noting that the one local company that could make this happen almost instantly is Koch -- note, among other things, that WSU's storied basketball program plays in Koch Arena.
It's interesting to speculate whether the nominally libertarian, notoriously political Koch brothers would take an interest in free software. Several things suggest that it might be worthwhile to approach them on the subject. For one thing, they do spend a fair amount of money on broadly philanthropic causes. (Not, I think, nearly as much as they spend on blatantly political ones, and the latter align much more closely with their business interests than with their more philosophical ones.) Second, there is a pretty strong libertarian segment within the free software community, most vocally Eric S. Raymond (author of The Cathedral & the Bazaar and The Art of UNIX Programming), but you can find it elsewhere, especially in Richard Stallman's insistence on "free as in freedom" as opposed to "free beer." The Kochs often make a point of their opposition to rent-seeking in politics. Proprietary software is nothing but rent, protected only by the legal force of "intellectual property" laws. And the model of building multiple, independent free software foundries, backed by private funding, scattered across the world independent of government bureaucratic meddling should have some appeal. Plus, free software is the best deal businesses can get: not only free to copy but with open source code no one can have a monopoly on supporting it and anyone who wants to adapt or modify it can do so. Virtually nothing else on the agenda of Americans for Prosperity actually offers so much prosperity.
On the other hand, the Kochs in practice have a pretty narrow definition of prosperity: very little of it, for instance, trickles down to their employees, nor do they even have to share it with stockholders. They've created a near-perfect symmetry where their relentless campaign for freedom and prosperity seems designed to benefit exactly a nation of only two individuals. So while they should see free software as a net benefit to the two, it might not be benefit enough to get them to help us. Still, it's not too crazy to ask. Just probably shouldn't be me doing it.
Display comments as (Linear | Threaded)
The author does not allow comments to this entry