Could governments effectively subsidize open-source development?

The open-source dilemma

(Credit: Matt Asay)

At the Utah Open Source Conference yesterday I presented a dilemma. Briefly, the idea is that as open-source buyers grow comfortable with open source they will stop spending money on open source. This leads to tragedy of the commons-type problems and a difficulty in encouraging the creation of more open source.

I therefore asked the question, "Who will pay for open source in the future?" I (and the audience) suggested that the problem may resolve itself over time as enterprises come to recognize that their failure to replenish open-source communities with either cash or code may come to harm the code commons from which they derive increasing amounts of value. I also suggested that Eclipse, Mozilla, and other non-profit foundations provide an answer.

Lastly, I suggested that governments might get involved to shore up funding for open-source software development. As I noted, governments derive massive benefit from open source (and from IT spending, generally). Why not fund more of it?

Europe loves open source. Why not fund it?

(Credit: Matt Asay)

I did not, however, have a clear idea as to the right way for this to be done. France, as noted in InfoWorld recently, suggests a way, as does TechDirt, which suggests that military spending could create the next Silicon Valley (so why not an open-source Silicon Valley, given how much the US military is buying into open source?).

France, the second largest market for open source outside the United States, does a range of things to promote open source, but its focus on open source for the rising generation is perhaps most important:

France's future grip on open source looks particularly strong, as it courts the next generation of open source developers. French authorities, for instance, handed out 175,000 open-source-software-equipped memory sticks to high school students last year. Technical universities have made open source their top priority, and some offer advanced degrees.

"All students in France use open source," says Bertrand Diard, CEO and co-founder of Talend, a French pioneer of open source data integration software. "A lot of universities in the U.S., except probably MIT, use traditional tools like Microsoft, Oracle, and SAP." As a result, open source talent is more prevalent in France, Diard says; development is faster, and software quality is higher because French developers aren't distracted by proprietary and competing technology. "The culture of open source is more advanced here."

This approach presents benefits and problems for the rise of open source. So long as it is fostered by government, it thrives. In my own company's case, France is a hot market for open-source adoption. At the same time, and despite the French use of "libre" (free as in freedom) to describe free and open-source software, in my experience the French often conflate "libre" with "libre to pay nothing."

In other words, France is a perfect example of the open-source dilemma. That dilemma would be resolved in France and other open-source friendly nations if the government stepped in to provide the majority or all funding for open source. But they don't. Nor should they. The socialized approach reduces incentives for entrepreneurs to sell commercial open-source, and that private enterprise approach to open source is critical to both filling in the gaps that governments can't and won't fill.

After a few thousand years, it's clear that free markets are much more efficient in delivering widespread value. We don't need another Soviet experiment in social provisioning of goods, open source or otherwise.

In sum, France's encouragement of open source is welcome, but not sufficient to overcome the looming dilemma in open-source software creation. Someone has to pay for this stuff, and it's not going to be governments. At least, not alone.

[mysqlrocks]

"I therefore asked the question, "Who will pay for open source in the future?" I (and the audience) suggested that the problem may resolve itself over time as enterprises come to recognize that their failure to replenish open-source communities with either cash or code may come to harm the code commons from which they derive increasing amounts of value."

You're assuming that there is a problem (or will be a problem in the future). Can you please provide some facts to support your assertion that there is (or will be) a problem? I don't think you're giving free/open source software enough credit. Yes, companies need to step up and contribute - absolutely. However, the free/open source software movement is *huge*- companies that don't step up will simply be rolled over. It's not survival of free/open source software that is at stake - it's the survival of companies that rely on selling proprietary software. In other words, you're looking at the "problem" all wrong.

[Dox96]

Hi Matt, I find your columns very interesting and informative, but I really can't help but think you've got the wrong idea about open source. You said "someone has to pay for all this stuff", and this isn't the first time you've said it. I thought you were wrong then, and still think you're wrong now. No one pays. Everyone contributes. Even the developers who work on code don't pay, because they're writing code which benefits themselves. The cost of sharing that work they did for their own benefit is negligible, and anyway, sharing also benefits the sharers themselves. Even someone who downloads and uses open source software contributes, by the mere act of using it. (One could even argue the mere act of downloading without even using it contributes - download statistics can generate revenue for open source organisations such as Mozilla, for example.)

Open source has benefited enormously from the resources put into it by for-profit companies, and everyone is very grateful for this. But don't make the mistake of thinking that open source *depends* on funding. Where would we be without MySQL? Well, we'd all be using Postgresql of course. Where would we be without OpenOffice? Why, using AbiWord and Gnumeric of course. Open source sprang out of nowhere, totally spontaneously, with no funding, and in a world where capitalist model of production was the *only* model of production. It is now threatening that model. Given this history, what on earth makes you think that someone *has* to pay?

Of course, you're looking at this through the prism of being an executive of an open source for-profit company, so it's understandable that you're trying to work out ways of making money from this new exciting way of creating software. Certainly, there's no doubt that outside funding helps spread open source more quickly, but it's no requirement and will spread naturally without it. I hope your business model doesn't depend on such funding. If it does, I fear for your company's future.

[russ danner]

Someone has got to pay for this stuff. I think that is *absolutely true.* Software is expensive and to develop. Developers scratching their own itch is most certainly a form of payment. They trade time for value. Time is the most valuable resource there is. You often don't realize it until you are running short on it. In some ways developers who contribute freely are paying the highest price there is. Commercial open source is one of the most important factors in open source because it combats the time dragon. With funding more development can be done and thus more open source can be released, upon which more development can be based. The virtuous cycle increases its range of effectiveness. Yes. Someone has to pay for this stuff -- and generally speaking, investing in open source is the smartest way to invest in software development.

[pete_l]

Short answer: No. It's anti-competitive.
Imagine you were a commercial software developer, with a product that you'd spent a significant amount of money developing. Next thing you see is the government funds some group to produce a similar product USING YOUR TAXES. You would not be happy - especially when it's the taxes you've paid that are now working against you.
Nice idea on the surface, but a moments thought shows it to be commercially naive.

[russ danner]

I'm not entirely sure I would throw in the towel there. I see what you are saying Pete. That would be frustrating at first but perhaps there is more here than meets the eye. From my perspective a situation like this is merely a call to change strategy. Mostly because you have to -- that is the annoying part -- everything was just dandy until --- but so what. Perhaps you could use such a scenario to your advantage by putting these folks to work for you. A few more developers working on the problem never hurts right? And you now have an opportunity to move your value contribution further up. This is good for two reasons, a cheap base stack raises the tide and with it all boats. Your total available market grows. So add value on top of what is being produced, provide what is not provided, stronger packaging, address higher order problems -- and for that charge a premium. Quickly engage and contribute your work making your code base the de facto standard.

I'm not saying this is the answer, and it's certainly not the only answer. But I am a bit more reserved in moving to the thought that government sponsoring (or anyone for that matter) open source development spells trouble. I find it more troubling that the military and government strike deals with engineering firms, funding development and later granting the firm the right to produce commercial derivatives. If government wants to sponsor open source -- god bless 'em. It may be sour news at first but there is lemonade all about.