Executive Director at The Outercurve Foundation. Member of the Board of Directors at Girl Develop It.
Erynn Petersen
Authored Comments
"Increase the number of women on open source projects makes it more enjoyable for those of us who are already contributing" refers to the women who are already contributing, of which, of course, you're likely not one. I wouldn't presume that it would change how everyone feels abut their work, or their environment, and certainly don't refer to you or the impact that it would have on your work or environment specifically. It's also possible to have a women- even two!- on a team, and not have them there colluding on a feminist agenda. Perhaps as a professional developer you haven't had this experience of being on a team with several women; as a dev, it would be more unusual for you to have been on such a team, and so you may not yet have experienced that women can be on a team with a bunch of guys and have no agenda other than hitting a project deadline in concert with their peers another as a full and complete dev team. As developers moving toward a release date with an increasing pileup of bugs they're more likely to be colluding with the rest of the team about last minute requirements thrown their way by an over caffeinated product manager.
There are a lot of developers who write code for private projects funded by companies, and it's difficult to have them submit samples of this code for review, especially in a hiring process. It's just easier for developers to share code samples from personal or open source projects. Let's assume that 100% of software developers are people. (Safe bet.) Let's go with the NYTimes data that 18% of these developers are women. If, when I request resumes from candidates, about 1/5 of my resumes come from women and about 4/5 come from guys, I have an indication that I have drawn my candidates from a representative pool of the available candidates. If, though, I additionally require open source code samples, and suddenly 1/5 of my candidate pool evaporates before I even interview people, then I have a good indicator that I am not fully assessing the available talent, and need to reconsider how to elicit code samples from candidates. One can level the playing field by having everyone come in and whiteboard out an exercise but it sure is easier to just have all the candidates produce examples of contributions to open source projects and review them all together as a batch.
People get hired onto teams because they have the technical capabilities to beat out other people looking for roles on those teams. One way to develop those technical capabilties is school; one way to develop them is professional exposure; and another very good way is the practical hands on experience gained by contributing to open source projects. Regardless of gender or the way that one develops one's coding skills, producing code to show one's abilities is generally only way to get hired into software development roles, on my teams or on anyone else's.
Good to see we agree on the salient point that getting more contributions to OS projects is a net good. More people knowing how to think logically, express themselves clearly, and write code is just generally a good thing, and open source projects are an excellent way to faciliate meeting that goal.
Sure. If we includes me, and I think we can safely agree that it does, than finding ways to increase the number of women on open source projects makes it more enjoyable for those of us who are already contributing. The "we" looking to bring more people into OS projects doesn't need to include everyone in the OS community.
Solving the problem also means that when I am looking for people to hire, and go find their public code samples, I can find code samples drawn from a distribution of contributors that is representative of the overall percentage of people in software development, and not just representative of the oddly skewed percentage of men and women in open source projects today. When I am looking to hire people, I want to be able to draw the best people from the largest pool I can, and having a distribution that reflects a missing chunk of the overall available developer population leads me to believe that I might be overlooking great candidates.
The larger problem to address is called out well in your comment: most *people*, not just most women, choose not to be software developers. Increasing the net number of people going into software development benefits the industry as a whole. Increasing the number of devs overall, including women, available to contribute to open source projects means OS projects have more resources to distribute across them. Very few open source projects are looking for fewer contributors, and even a tinier percentage (likely near 0%) of open source projects would like to see fewer people rather than more available to make contributions to OS projects.