This is definitely a concern. What does it tell us, I wonder?
I'm certainly not saying that there are no individual bigots in the OSS world. With thousands (millions?) of people you'll get all points of view, beneficial and toxic.
A simple 'grep' of GitHub reveals a lot of crap, but are these influential projects? Do these projects have multiple committers who have signed off with pride that "my project contains 'n-----' in the code comments"? Do these projects have lots of users?
It would be interesting to plot the occurrence of racial slurs against the number of participants in a project (controlling for offensive term filters or corpuses of 'real' language). My hunch is that the vast majority of these terms occur in one-time undergraduate projects or forgotten hobbyist repos.
What I mean is, anybody can create a GitHub project for any reason and post anything they want to their repo.
I don't consider a Freshman anonymously posting some obnoxious text on GitHub to be a part of the "Official Open Source Community (tm)" any more than a two year old finger painting is a professional artist.
If such a person turned up at PyCon or RubyConf and starting dropping n-bombs in public conversations, they would get a very fast education in what is acceptable public discourse in an open source community.
theorique is perfectly willing to have a reasonable discussion about it. Chasing after them to take back the 'any' in the first sentence of the first post isn't going to accomplish a lot.
I'm certainly not saying that there are no individual bigots in the OSS world. With thousands (millions?) of people you'll get all points of view, beneficial and toxic.
A simple 'grep' of GitHub reveals a lot of crap, but are these influential projects? Do these projects have multiple committers who have signed off with pride that "my project contains 'n-----' in the code comments"? Do these projects have lots of users?
It would be interesting to plot the occurrence of racial slurs against the number of participants in a project (controlling for offensive term filters or corpuses of 'real' language). My hunch is that the vast majority of these terms occur in one-time undergraduate projects or forgotten hobbyist repos.