While exploring the notion of citizenship, I interviewed a number of people involved in various communities. When discussing about what a ‘good’ contributor is, the conversation often led to the problem of poisonous people in FLOSS communities also given the name of As.h.eS. Donnie Berkholz (Gentoo) has made several presentations about the problem of as.h..les in communities by showing how they severely damage a project and its community (see latest version). These community members are also qualified as ‘poisonous people’ (see Google TechTalk about how a FLOSS project survives poisonous people). I personally prefer the term toxicity to refer to the phenomenon but I am not pretending that it is any better; it is just an effort to use a more encompasing term.

There are two overall perspectives to study the toxicity phenomenon. One is to focus on toxic people themselves and study how they impact a project. A second approach is not to focus on individuals themselves but rather on ‘toxic behaviours’. This perspective is broader as it encompasses the individuals qualified as as.h..les or poisonous but also members that may not overall fall into the category but who one day or another performed an action that has a negative impact on the project. For instance, a person may occasionally send a flaming message which may trigger further flaming from other members or other negative actions in the community. The flaming message has been toxic and has spread but one cannot qualify a member to be a toxic individual for the community for having sent one toxic message when perhaps having a bad day. He may still be a good contributor for the community.

Flaming can be one type of toxic behaviour but there is a need to first define the notion. An intuitive definition of the term could be something like any action performed by a member that has a negative impact on a project or its community and that triggers further similar actions from other members. Once the notion is defined, then there are various aspects that can be investigated such as measuring the negative impact on a project/community, understanding the replication mechanisms of toxic behaviours, or designing a toxicity metrics that would allow to monitor the degree of toxicity within a FLOSS community.

This is an area of research that I think needs to be investigated as it seems to be an important matter for FLOSS communities. Please feel free to comment and  share your insights about the idea of toxicity in FLOSS and its relevance.