Wired’s open source myth

Wired has an interesting twist on the tired old myth that “open source is developed by volunteers”, that has been. The new twist is that it reverses the usual myth that it is not possible to make money from open source: instead the wrong people will make the money and it will all fall apart. Unsurprisingly this flies in the face of the facts.

The argument is superficially convincing and has probably got a good few people fooled — I found the Wired article through a post agreeing with it on the usually quite sensible Broadstuff Blog.

I will summarise the argument before exposing the fallacies. The argument is that open source companies rely on unpaid outside developers. When a company make lots of money, which is not shared with the external developers, the latter will be disillusioned, stop work, and development of the company’s products will cease.

The fallacies in this argument are:

  1. Most open source developers are paid, not volunteers, particularly in the case of the sort of software that makes a lot of money,
  2. most of those who are not directly paid, have some other motive for developing the software,
  3. the remaining very few developers who are unpaid idealists, are likely to accept someone else making the money: as the BSD developers were happy for Microsoft to profit from their work.

The first of these is what matters here. What is even better is that it is true of all but one of the companies that Wired offers in its examples: MySQL, Trolltech, Zimbra, Covalent, Spring Source and Mozilla.

The exception is Mozilla, which does have a large number of unpaid contributors, but Mozilla is owned by the non-profit Mozilla Foundation: if huge profits are made, the non-profit Foundation will simply have more money to spend on “promoting openness, innovation and opportunity on the Internet”.

The others develop software themselves and distribute it under an opens source license. None of them seem particularly keen on community contributions. I could not even find a documented way of making contribution on the Trolltech, Covalent or Spring Source websites. Zimbra only accepts contributions from those who are prepared to sign agreements that hand over copyrights, as does MySQL — which will greatly deter volunteers.

I would guess (although no data is available) that many of the external contributors are paid employees of other companies with a strategic interest in improving the software (e.g. Google, which is a major user of MySQL and which has signed an agreement to contribute to it). This could explain why MySQL currently only has 22 tasks (a tiny number for so complex an application) listed for external developers to work on. Incidentally, Mozilla gets substantial contributions from both external paid developers and true volunteers.

No doubt some some people will now claim that this shows that open source does not work. That would be wrong. It simply shows that there are many different models that work for open source. Some involve heavy contributions from a wide range of people. Some involve a main developer (an individual, a profit making company or a non-profit) doing most of the work with only a limited level of enhancements from outside. The latter relies more on different advantages of open source (wider distribution, reassurance for customers) from the former (sharing the work).