Some thoughts on diversity in FOSS:
The lack of diversity numbers (percentages) from Foss organizations in pure technical (read core-developers or women with commit access) roles. For the most part, some Organizations include non-tech roles while releasing diversity data - while this is a good thing (I think its important to respect the technical writer and/or bug squasher as much as the programmer) we all know (and have probably experienced first-hand) that women in non-technical roles are hardly respected.
Secondly, the whole "diversity" brouhaha has given us access to some numbers (which we know are low) but there is hardly any discussion on what changes need to occur on the basis of these low numbers. Hence, I think this data is important for the following reasons:
-
It tells me how much harder and longer it will take for me to get there - Becoming a core-dev with commit access does not happen overnight. The key differentiation lies in the fact that it takes time (a LOT of it) to get familiar with a codebase to be able to contribute meaningfully to it - yes, even bug-squashing is not a trivial task in a mature code base IMHO. Most of the Foss projects have a focus and unless your line of work aligns (or you are hired and paid to work on it) with them, its harder to justify taking time off from paid work to volunteer to write code, unless you are a student (includes researchers and students required to do project work who are ofcourse paid a stipend). How can Foss orgs tap the programming skill-sets of those women who are not CS students/researchers or working in Foss orgs/companies and make it easier for them to contribute and work on a Foss codebase?
-
Mixed Data : Currently, the mixing of non-technical with technical roles seems like fudging with the data to present a false premise of organizations doing enough to increase diversity in tech. Afaik, very few org's have a proper mentoring program that is regular in matching up programmers. For the most part, its a bunch of links on a "how to contribute" page that can be very intimidating for newbies and even regular programmers alike. Before you cite GSoC** - yes, this was one reason why organizations that participate have an awesome list of links and helpful resources. However, GSoC and other SoC programs are a 3-4 month (5 days/week X 8 hours/day) effort and the problem sets on those pages are designed for that purpose only, either advanced topics that cannot be done over a weekend, and/or requiring a ramp-up-period in terms of learning which is not trivial (and intimidating?) for a newbie. I cannot emphasize enough why its so much more easier to work on a technical problem when a core-dev makes an extra effort to reach out and explain the nitty-gritty which may not always be obvious at first glance.
-
Leaky pipeline : Gnome's OPW and other SoC efforts are laudable efforts that have done much to change the diversity equation in FOSS. However, I am unable to find any figures beyond those 3-4 months of efforts where women participate - for example, percentages for the number of women who stay back and continue to contribute after the SoC period. I would hazard a guess that this is somewhat related to the points above but I'm hoping that it isnt the case and that concentrated programs like these increase AND retain the number of women in FOSS.
1. IIRC, some years ago there was a thread (or was it an IRC convo? I dont remember which) on releasing diversity percentages for GSoC participants, which was refused (and I paraphrase) due to reasons of privacy2 and also because this was data that was only available with the participating Orgs who may not be inclined to release it. Fair enough.
**2. I agree with privacy concerns - for example, I'm more likely to participate in a diversity survey if it leads to concrete efforts to increase the number of women than if the data is just being collected as a marketing PR (public relations) effort for the Org/Company.
</brain dump of thoughts on diversity in Foss>