Cooperation and competition

If the characteristics of open source development models were to be defined by a unique expression, `cooperation and competition' would probably be the one to choose. Indeed, the combination of both mechanisms is visible in almost any open source project, not to mention when we look at the big picture, where every project and company is in some sense competing with others for resources and `market acceptance', while collaborating with the reuse of the same code base. Let us provide some detail on how those mechanisms work, and enforce each other in a very productive mixture.

• Cooperation within open source projects. Developers participating in the same project, usually cooperate at levels higher than those usual in proprietary software projects. The design phase is usually completely open, with all developers collaborating, and during the coding phase is quite common that a developer reads and fixes bugs in the code being developed by another. The flux of information is usually very high, and problems are solved by consensus among at least the core developers. All of this together, in addition to be more effective, usually causes the developers to be more committed to the project, and makes their work much easier, thanks to the help of other developers.

• Competition within open source projects. Within open source projects there is also some degree of competition. Most of them are organized as some kind of meritocracy, where the developer who does more for the project deserves more credit, and is given more weight in decisions. This allows for a certain level of healthy competition between the developers, from which the leaders of the project (when such leaders do exist) usually emerge. This is still a fact not understood by many companies, which need to realize that in order to participate in the decisions about future directions of a project they need to have some respected developers within it.

• Cooperation between open source projects. The very nature of open source licences ensure a high degree of collaboration, even between competing companies. This is for instance the case of the many companies selling distributions of GNU/Linux. All of them share the same code base (the GNU/Linux operating system and applications), and whenever one of them fixes a bug, or improves a package, all its competitors have access to the source code, and thus to the same fix or improvement. Therefore, projects and companies in the open source world are compelled to collaborate by the open source model. And when they try to avoid helping their competitors, they have to move away from the open source model, which usually gives them more problems than benefits.

• Competition between open source projects. But in a world of limited resources, open source projects also compete between themselves. Projects developing software for the same niche compete with each other in a way similar to companies in that niche competing. In the end, they are forced to maintain high levels of quality, or they will lose users and developers, and will finally fade away. In fact, this is one of the mechanisms which ensures a good quality in open source production. When a project starts failing in terms of perceived quality or management, there is a chance that some of the developers will open an independent development branch. This branch soon becomes an independent project, perhaps using a similar code base, which starts to compete with the old one. This competition usually causes both teams to improve their product, and in the long term, one can become mainstream while the other becomes marginal, or they can diverge giving special attention to different issues. This is, for instance, the case with NetBSD, FreeBSD, and OpenBSD, all of them sharing similar code, each with a different focus, and all of them exhibiting a high quality product.

Usually, this mixture of competition and collaboration is not intentional, but the product of the development model, and of the licenses used in open source projects. In short, everybody is forced to compete by exposing the tools (the source code) they are using, and improvements are quick to spread through the competing projects, in a manner quite different to the traditional software industry. Competition and collaboration are probably the ultimate cause of the high efficiency (in terms of quality and quantity of software produced with a given set of resources) which open source projects reach.