I intend to play devil's advocate here. My brush may be too broad in some cases because I lack accurate information or it's hard to discern whether my words describe a hypothetical or actual, and my statements may offend someone. That is not my intent, and if I've done so, please correct me.
To me, this conversation originates from too little information. Too much generalization, and very few specifics in the 2017 Road Map. One solution is to fill in those holes as some have began to, but at the same time, some individuals elaborate not at all because they've dismissed some of the concerns. I hope to explain those concerns (as I understand them), why they are concerns, and what details can be provided to assuage these concerns.
Consider that I'm effectively a stranger in the Rust community. I don't know the team well, I've never met them, and we're not close friends. I know you superficially through the questions I ask/read and the answers provided. At the very least, I must say the Rust community is very friendly, inviting, and tolerant of my dumb questions that are probably asked a thousand times already. You all have my heartfelt thanks for that.
Mentally, I'm very slow. At times, dumber than a box of rocks, hence the moniker. Eventually, I'll understand, but it takes excruciating time to get there. As such, I hardly spend time researching how the Rust core team behaves in compromising situations (I'm busy struggling elsewhere), so when I see 81.6% of the community are white males, and that this is a "problem to solve" [1], I ask myself how the core team will solve this, and the truth is I just don't know, I can only speculate. The benefit of the doubt is truly warranted before attacking anyone's character, but it should not hurt to discuss my speculations/concerns, so at the least I know definitively what you won't do, and better understand what you're likely to do.
Here, we have one metric X = 81.6%
that expresses a problem to solve. The only way to solve this over-simplified situation, is to reduce that 81.6%. At what point do you consider this "solved"? When that metric is 70%? 50%? 25%? Is this metric even driving the call for action? Why collect this data and refer to it if this wasn't the case? Will 2017's efforts be considered a failure in 2018 if X
is still 81.6%, or a lesson that despite your best efforts, too few people related to the 18.4% are interested in learning Rust.
How do we reduce this percentage? Assuming we're stuck with the community we have, we're not kicking anyone out. So the only way to adjust that percentage is to alter the growth rate of "new admissions". We can definitely reduce that 81.6% by making it harder for new white, male members to join the community, so that fewer join[4]. Or we might fail by encouraging minorities to join our community, since we can't force them to join us. In this very limited simulation, one of three outcomes are possible.
- If we remain more successful at attracting white community members through 2017, then
X
increases.
- If we manage to attract as many minority members as white through 2017, then
X
remains at 81.6%.
- If we hit a home-run and our minority membership grows faster than the white membership, then
X
decreases.
If we limit ourselves to an arbitrary percentage value, then (1) and (2) are failures, and only (3) can be considered a success.
Personally, I hate the exclusive use of percentages in any call for action/argument because they omit the truth of the situation and erroneously become the focus of any solution/discussion. Let's specify exactly what the goal is here. In (1) and (2) above, we could double the minority membership, which is a success IMHO, while X
still indicates our community is not friendly to minorities. So let's focus on raising the level of minority participation. The call to action is that our minority membership is <1000 and we want to raise that above 1500 by the end of 2018, regardless of what the majority % becomes.
After reading @mbrubeck's post for the fifth time, I can see my hypothetical is impractical. The goal is to welcome everyone.
If "Our goal is to be welcoming to everyone...", I can see how that policy intends to avoid a zero-sum situation. I can also rationally imagine a [hypothetical] misguided effort to reduce that 81.6% that seems zero-sum to me. I don't mean all community building becomes zero-sum if this effort comes to pass, but it does strike me as a zero-sum situation on a smaller scale.
The 2017 Road Map advocates mentoring as a method to improve minority participation. I don't know how Rust does mentoring, so I envision a simple program where one mentor pairs with one candidate. If this mentoring program is key to reducing the gap between minority and majority, then it appears to me that a candidate must be selected based on race/sex to encourage more participation from minorities by favoring them in the selection process (how else does mentoring reduce that 81.6%). I imagine myself in the position of a mentor selecting a candidate. If I have one minority candidate and four majority candidates and a desire to support minority participation, I will select that single minority candidate to further a cause I believe in. To do otherwise would subvert that effort by at least promoting the status-quo, and as a result I've intentionally discriminated against four individuals based on race/sex[2].
I see this as a zero-sum situation, albeit very short-term, because at this moment a finite resource (mentoring) is awarded to some candidates and denied to others. Long-term, those denied may/will/should benefit from the contributions to come from the mentored candidates, and everyone gains from that perspective (sorry @mbrubeck, this didn't fully occur to me until I got this far).
If the Rust team advocates reducing that 81.6% through mentoring, and if mentoring works as I imagine, then I would be alarmed by an organization indirectly encouraging racism/sexism. Those are big if's that have yet to be substantiated, and it's still fiction at this point, and I expect it will remain fiction.
Outreach programs I don't have any argument against. Their purpose as I see them is to reach out to others, whoever those others may be.
With these very simple scenarios, I can understand why @Meai1 is concerned. As the OP's title states, "How do we combat this?" Transparency.
- The team can elaborate more on that 81.6%. What does it mean to the team and why is it actionable?
- What actions are you taking (with details)?
- And what goal are you trying to achieve with that action?
If 81.6% carries any actionable significance, I'd like to see that it's an arbitrary metric to pursue this year and next year we'll be polling other metrics to pursue efforts to reduce disparity of those metrics in our community.
I can see a benefit[3] to mixing up a little what may be the de-factor standard operating procedure of 2016, and push everyone outside their comfort zone.
This is all just food for thought. I'm not demanding answers to these questions, nor am I insinuating nefarious plots. The Rust team is free to do as they please, they've done a terrific job thus far, and will continue to do so I suspect.
The OP and Road Map's "X
% imbalance needs action" struck a chord with me, but after spending six hours writing this mess, I'll return to spectating.
- [1]: My knee-jerk, over-simplified, and yet not entirely unrelated summary of the situation.
- [2]: I don't subscribe to any notion that racism/sexism only applies to minorities. I approach that point from a mathematical perspective.
racism = { A discriminates B | A, B β All Races }
, and similarly for sexism.
- [3]: Parable of the Polygons
- [4]: I realized while proof-reading that if minority growth drops, method (1) doesn't guarantee success.