Slashdot Mirror
Economist's Take On Open Source Development
An anonymous reader writes "Economist Dean Baker outlines alternative funding mechanisms for software development in a new report called "Opening Doors and Smashing Windows" [PDF Warning], available at the Center for Economic and Policy Research. One proposal is to create a US government-funded Software Development Corps of public software corporations, which compete and produce only free and open source software. Baker estimates that through the resulting lower prices in software and computers, the government would recoup its annual $2 billion appropriation to the program and US consumers would save $80-120 billion each year -- all while 20,000 software developers are supported to work specifically on open source projects."
...sounds more interesting to me. He proposes an "Artistic Freedom Voucher", whereby people would be provided with a voucher for, say, $100, which they could direct to a person engaged in creative work (like writing open source software). This sounds rather nifty, since it would allow folks to "pay" for the projects they find most useful personally.
:-)
Of course, another way for open source programmers to make money is to publish a book. Programming in Java? Give it a look! Think of it as sponsoring an artist
The Army reading list
Despite the subject dear to most of us, we shouldnt ignore the fact that he is essentially claiming that developed software is free. He is totaly ignoring the costs incured in developping the software, and only accounting for the costs incured in copying it.
In his t-shirt example he is claiming the price of $20, which without doubt is probably 99% manufacuting expenses and remaining 1% design expenses when spread over the first 100.000 copies. However, for software the ratio is the opposite, with 1% material costs (packaging, manual, cd, etc) and 99% design expenses, again spread over the first 100.000 copies or what not.
Because, money is evil. Vouchers require administrative personel. Staffing. Oversight committees. Planning boards. Watch dogs. All of which comes out of a little administrative fee added on to vouchers. Let's see cash do that.
Plus, I still don't want my government endorsing open-source companies. I don't want them endorsing anything. Not companies. Not churches. Nothing. I want them to do the three or four necessary things they're obligated to do and stop trying to push utopia through government process. If you and I want successful open source, we can do it without a fucking government mandate.
"War"? Oh, you probably mean "proxy war", where the US government sponsored assassins, fascist states and terrorists. Though you might possibly be referring to a real war where you had the shame of killing over a million humans in the name of "freedom".
Government paying for free software? send in the troops...
www.hup.harvard.edu/catalog/WEBSUC.html
The url above is for "The Success of Open Source" by Weber. Another take on open source is by Clayton Christensen in his books on innovation. I highly recommend both.
The thing about open source is that it puts the lie to the notion that people only do things for monetary gain. It is a poisonous notion when it is used as the basis for economic policy. In that light, the notion of massive government subsidies for open source efforts, is ham handed. IMHO, economists and policy makers should make the effort to understand how open source actually works before they propose to spend billions of taxpayers' dollars. I suggest they start with The Bazaar and the Cathedral. It's available for free download.
There is a place for publicly funded research. There is a place for publicly funded open source work. The model for both is probably similar. The idea that private enterprise should fund all research and software development produces bad results. For instance, having drug companies do all medical research means that only profitable drugs are produced. A free cure for cancer won't happen in such a regime. Similarly, pouring money into private corporations to fund research is usually a massive waste of money.
I'm not against public funding; I just don't think that this proposal is sufficiently enlightened to work.
Here's a link to Google's copy of this report in HTML.
The amount saved per person has nothing to do with how much they are paid. Just because there's two numbers and one's bigger doesn't mean the right thing to do is divide the bigger one by the little one.
The theory is that $2 billion pays for 20,000 programmers. Calculating this out will show you an estimated cost of $100k/year/programmer, which is a reasonable figure for salaries plus overhead. The savings are not that those 20,000 programmers don't have to get paid elsewhere, but that their code will be more widely used than it would be if they were writing proprietary code, and as a result, the economic value to our society, in the form of lower software costs, would be something like $80 billion.
Which is frankly not a particularly unrealistic notion.
My blog: http://www.seebs.net/log/ --- My iPhone/iPad app: http://www.seebs.net/seebsfrac/
While his argument about copyrights was genius, I didn't really like the way the conclusion seemed to force a choice between closed software and socialist government. IMHO, we are better off with neither.
You don't have to agree to a license to use a GPL-licensed program. Those who possibly had to agree to it are those who gave it to you in the first place.
You only need to agree to it if you want to *redistribute* it.
I'm not so sure it wouldn't work. It has been done already with Blender 3D, see http://lists.suse.com/archive/suse-linux-e/2002-Au g/2538.html
My supporting evidence can be found in any basic textbook in public economics. Obviously Microsoft is not a pure monopoly, since we have Linux, OS X, Unix etc, but it's pretty close. The main reason why Microsoft is not, and can not be, a pure monopoly is because all OS'es are not exactly the same, and thus people will not choose their OS on price alone. If that was the case, Microsoft could set their price lower than any competitor (because of increasing returns to scale in software production), and thus push all competitors out of the market. They would then be able to overprice their product. And sure, Microsoft cannot bar competitors from entering the market, but the market itself can, since the OS market has pretty a high entrance cost (writing an OS is not cheap). As i said, the argument is not perfectly applicable to the software market, but i think the Microsoft near-monopoly shows that it is not entirely flawed either. As for "socially optimal amount", that is the amount that maximizes the utility of society, in other words, the amount where the marginal benefit to society equals the marginal cost to society. As for the more general discussion on whether government monopolies are ever optimal, how do you suggest we should pay for city streets and squares? Privately owned? Then who pays? Should you have to pay a toll to go to the shopping street in your city? How would that effect the business of the shops? What about the police, or the legal system? Could these be better operated by private firms?