Linux Magazine France, November 2006 (article online, December 2007):

On n’utilise pas des mots comme “turnover” (taux d’attrition) pour les volontaires, mais il est intéressant de regarder pourquoi ce phénomène existe. Quels sont les caractéristiques qui font que certains projets ont plus de facilité à garder leurs développeurs que d’autres ?

Like professional projects, some community projects are better than others at getting new contributors, and keeping existing contributors engaged over time. This article (in French) explores some of the characteristics of a good community.

(read more for a translation).

Recently, Matthew Garret’s resignation from Debian, and my own from the GIMP have caused some discussion about what motivates free software developers and our community.

This is not a new phenomenon  –Jamie Zawinski (as a Netscape employee) turned his back on Mozilla in 1999, and people less well known, but just as important to the life of their community, stop working on free software all the time.

We don’t use words like “turnover” for volunteer projects, but it’s interesting to look at why this phenomenon exists.What are the characteristics which make some projects better at holding on to their community members than others?

The first criteria which a developer/user interested in a project has is the welcome he receives. How likeable are the people he first meets in the project? If his first contact goes well, then he will be more inclined to return – a great step forward. If, on the other hand, his first contact is more difficult, then it doesn’t matter how much he is predisposed to helping the project. He will not make that leap and invest in the project.

After the welcome, there must be a sense of justice. Does my opinion count as much as the opinion of others? Is what I propose criticised on its merits, or does the opinion of one of the “clan” mean more, without my understanding why? Is there a well-defined path I can follow to go from new contributor to full-fledged member of the project (how to get an SVN access, an email address for the project domain, and so on)? Are my patches considered quickly, or do they wait for weeks without comment?

Finally, there is what I call the adaptability. Freedom gives us the ability to do whatever we want. In a quest for justice, there is a tempting trap to avoid: rigidity. If the rules of your community become more important than the goals of the project, then your project will suffer. If a project member does not feel that he can change things via his participation… he will look elsewhere.

Jason Clinton recently wrote about communities he has participated in, and his conclusions are interesting: it doesn’t matter what your association or community does, “in all instances of those things which are demotivating, it is a single individual or small group of individuals which are ultimately responsible for the anti-social behavior” – in other words, we destroy ourselves.

There are some good examples of communities who have succeeded in these three points; they have been welcoming, and have had sufficient structure to ensure justice within the community, without having so much structure as to stifle creativity. Inkscape and Subversion are two projects which come to mind, and their success is a testimony to the way that they have systematically reinforced good community structure.

In a world where we are free to choose which projects to support, it doesn’t matter how much we have already invested, a volunteer will always want to take part in a project where he feels appreciated.