Diversity on the governance teams

Don't forget that the field is "programming language development" not "IT in general" and there the average seems to be very low from looking at other languages.

1 Like

It seems to me that you are not convinced that the lack of women in leadership positions is fatefully discouraging to prospective women participants. There is valuable input from two of the few women in Rust in this very thread, and their observations and experiences are supported by academic research (e.g. one, two). It is important that this is clear.

But as an interested male dental technician, would you feel welcome to join the team (or, analogously, contribute to its community and ecosystem)?

4 Likes

Don't forget that the field is "programming language development" not "IT in general" and there the average seems to be very low from looking at other languages.

I understand that you are retracting your previous claim and proposing another. (Since "programming language" is not same as "computer science".)

It seems arbitrary to me, because if I were to argue Rust has less female participation even compared to other programming languages (I am not sure about that, it's just for the sake of argument), you could change the reference category to, say, systems programming language, or even Rust programming language: e.g. "the field is Rust programming language, not programming language in general". Then the argument becomes a tautology.

4 Likes

While I'm not convinced that we're equally diverse as other projects, this isn't really a good argument.

An overarching initiative is beyond the means of the Rust community. Additionally, people will do more for things they care about. People from the community interested in making it more diverse may lose interest when they are asked to fix all of the communities.

What we can attempt is fixing the Rust community itself. Those who want to participate in overarching initiatives; can.

It might be unfair to call out the Rust community in particular for this; but I don't really care about that -- there is a problem, and we should fix it. If I live in a neighborhood filled with dirty houses, that should not change my desire to keep my own house clean.

I partially agree with this. I'm guessing there were few to no options for females in the team; and looking over contributors to code and discussions that seems mostly correct.

However, that we are at that state of affairs is itself a problem.

And people also naturally are turned away from such projects if they don't think they will be treated well there. Prospective newcomers owe us nothing to dig deeper and realize that we are harassment-free et cetera.

3 Likes

I agree with that idea but I think the only way this can improve is to nurture female developers that come into the project. You can't just snip with your fingers and all the sudden there are female developers. The only thing you can do is to encourage female developers to join and to set up a harassment free environment. In that regard Rust so far was good (this thread notwithstanding). Unfortunately Rust targets system level programming and there are naturally less female developers than in some other areas.

Which is why I think instead of continuing this discussion it would be a good idea to go to the (small but existing) percentage of the female contributors and organizations like PyLadies to get suggestions. If interest is there I can probably talk to friends and collegues who do a lot there. That said, I'm pretty sure the suggestion will be to make female targeted events to teach Rust and show why it's interesting.

1 Like

Unfortunately Rust targets system level programming and there are naturally less female developers than in some other areas.

I am not sure what is natural about this?

6 Likes

I think the largest reason why this discussion has been so contentious is this assertion that there is something that the core team did wrong in their selection, which implies that there was something else they could have done that would have produced a better result. And I'm having a hard time seeing what they could have done without spending months on further community building to bring more people into the community.

I've been biting my tongue and trying not to discuss individuals in this entire thread, having discarded several messages because I didn't want to make anyone uncomfortable by singling them out, but I really think that it will help us stop beating around the bush so much. (By the way, if anyone feels that discussing individuals here is inappropriate, let me know and I'll edit it out of my post; since everyone I've mentioned has been prominent and public in their contributions, I expect that they won't mind, but I understand if anyone doesn't want to be called out by name).

By my count, there is currently one woman active in development of Rust, @carols10cents (by taking a look at the contributors list on GitHub, as well as the last few release announcements). She has already chimed in in this thread, to indicate that she is not interested in joining any of the teams:

There are some other prominent women in the community, like Lindsey Kuper, though she seems to have moved on to other projects by now, and Julia Evans, who has written some really good blog posts on learning Rust, writing a kernel in rust, and other topics. They might be good people to reach out to to see if they would be interested in getting back involved or more involved, but I don't think that they are currently active enough to be candidates for the positions.

If I have missed anyone, I apologize, and I'd love to learn about more contributors.

With a small pool of candidates to choose from, it can be hard to find any that are interested in taking on additional responsibilities. It has been mentioned before that for the moderation team, at least, it would be possible to reach out to candidates who are not currently part of the community but have experience with moderation and are interested in joining; that team doesn't really require existing involvement in the project, but just experience with moderation of internet discussion and the desire to be involved. That is one possible way forward.

But beyond that, I'm having a hard time seeing what choices the core team could have made differently. Without women involved in the project who are interested in being on the subteams, how could they have come up with subteams that include women?

That's why I've been so focused in this thread on what kind of outreach efforts we can get involved in; I think that the project has done a good job of creating a welcoming, inviting atmosphere, has a good code of conduct, etc, but that doesn't seem to be sufficient so active outreach efforts are likely to be the best way to continue to improve the situation.

The situation needs to improve, absolutely. It is apparent that merely having a welcoming community, good code of conduct, and encouraging those contributors who do show up is not enough to overcome the social and systematic biases that tend to keep women and minorities from involvement in open source projects. The software industry in general has a fairly low percentage of women, open source software has an even lower one, and I believe that systems programming also has lower percentage than the rest of the tech industry; Rust falls at the intersection, which I think may be part of why it has such low participation despite being welcoming, having a strong code of conduct, and so on.

I don't know of any way to improve this beyond active outreach. Do you?

4 Likes

Neither can we not do anything and wait for stuff to happen on its own.

I personally do feel that there are more ways to handle this than just nurturing (though that is something that is on the table and I hope will be fleshed out as a concrete proposal soon)

Also on the table, but we don't need to stop there. There are other things that can be done too (scattered through the thread, though I did send the core team a summary of some ideas)

(As I mentioned before approaching others is something I'm not comfortable with doing myself because I don't think I can easily represent the Rust community -- waiting for the core team here)

No, but anyone interested in improving this situation should probably start organising female oriented meetups and things of that sort because that's where you can make change. And to add to this: I already said on Twitter a while ago that I will support any Rust female developer society. Unfortunately I neither have the time, nor connections nor motivation to do this on my own. I live in a place where it's already hard enough to scramble any developer to a meetup. But if something is happening around central Europe, let me know and I will be the first to support it by advertising for it, showing up with a talk etc.

1 Like

Exactly that could have been done: Not putting up that team page right now, instead spending more time building a more diverse team. (This has already been said in this thread.)

Dismissing anyone who disagrees as "people who love controversy" is not helping.

Please, do not use the term "females" when talking about people.

1 Like

Mitsuhiko: this very discussion ā€“ with its questioning, refuting, minimizing, derailing and excuse-making ā€“ is an active part of the problem. I am observing and experiencing that subjective fact myself, and have heard several women observing this discussion comment the same way.

Itā€™s 2015 and weā€™ve been around this block about a thousand times already in the FLOSS world. Itā€™s way past time for it to be excusable to make up reasons for women not participating in a field. Thereā€™s research on why, organized explanations of things that make it worse ā€“ conversations like this one, among them ā€“ and experience reports on how to make it better. Ball is in the leadership teamā€™s court. They really donā€™t need you making further excuses, it makes it worse.

18 Likes

Let me try to summarize the points that I feel I can present fairly. The Rust community is currently very homogenous. As a consequence, neither the core team (which existed before) nor the subteams/peers (which are new) are as diverse as everyone would like. Dissecting how that state came to be doesnā€™t seem to yield anything productive or helpful or pleasant, only get dangerously close to blame assignment and excuses. So letā€™s stop doing that.

This homogeneity has the tangible downside that some people of underrepresented groups have more reservations about joining our community, presumably based on bad experiences (own or reported) with very homogenous communities. If they would join, theyā€™d probably find it very delightful, but they donā€™t because their heuristic filters us out. This filter is quite valid, or even if weā€™d think it wasnā€™t, we should still be concerned that we are missing out on potential community members.

Those who argue against the team announcement state that the above downside has measurably worsened since the announcement, presumably because the lack of diversity is now more visible. Let me repeat, nothing has changed about the demographics of the community at large, nor those of the biggest contributors (many of whom now hold a peer position, hence are more visible). It simply became more visible who the workhorses and great thinkers of the project are.

Since nobody here is a wizard, at least not one skilled in conjuring contributors and experience out of thin air, this will not change easily or quickly. This too bears repeating: All the outreach and community building (that everyone agrees to) can enable us to build a diverse team in the next, letā€™s say, few months. The exception being the moderation team, which does not require the Rust-specific expertise of other subteams.

One proposed alternative was to not announce teams at all, resp. disband them now and go back to the unofficial, undocumented structure that existed before. If I understand correctly, @skade has since mellowed and proposes a compromise that keeps the subteams but refines or shrinks their mission statements, but other participants are apparently still in favor of that? Anyway, the many disadvantages for lots of areas other than outreach have been discussed at length, including by me. It is unlikely to find acceptance but I donā€™t want to predict anything.

Here begins the part where I stop summarizing and start thinking out loud.

As I understand it, the problem with the team announcement is one of communication. The announcement and web presence of the team unintentionally sends an unfortunate message, or can be read as sending such a message. (As above, it doesnā€™t matter how valid these concerns are, they cost us community members.) That message is that Rust is, to quote a few ways to express it: ā€œAll white dudesā€, ā€œnot diverseā€, ā€œa boyā€™s clubā€. Unfortunately, that is the truth. But itā€™s not the whole truth, thus misleading, and we can do better. If you want to, you can think of this as public relations rather than social justice.

So how can we do better? Obviously, the main thing is to fix the problem, that is, reach out and grow our community not only with the ā€œdefault audienceā€ but also, specifically, with groups that are currently underrepresented. Everyone can help with that, even though the leaders should (and I do imply responsibility here, not just my own wish) take part in this. But since this is established to not work overnight, what can be done while itā€™s in process?

Since this is a communication problem, I say: Communicate better. Once the inevitable ā€œdiversity initiativeā€ is taking off, document it on the web site and announce it to the public. Do this well! Take the hints from the kind community members who have given doā€™s and donā€™t of outreach above, and not only apply them to the actions taken but also make clear in the press releases that itā€™s being done like that.

Even before that, emphasize whatā€™s already great about the governance model, including on the team page. That Rust is developed by an open and inclusive community, that the team merely sets directions and breaks ties in votes, that team members are community members first and foremost, that the teams are not set in stone, that new teams can and will be formed, that membership on the team means more administrative work than any real power. Maybe put the moderation team first, or right below the core team. I feel like that is ā€œcheatingā€, but whatever, itā€™s marketing, right? :smiley:

As for the actual outreach and community building, I donā€™t feel comfortable asserting anything about it except that Iā€™m in favor and if someone near me (Rhein-Main-Gebiet, Germany) wants to do such work Iā€™ll see if I can help. I am not part of any Rust user group or any similar group from which a Rust user group could branch off. I also lack the time and the experience to creating one ex nihilo. I am stuck with armchairing about how to behave on the Internet.

7 Likes

Core team member here.

First, I wanted to thank everyone for the open and honest discussion on this thread already. I was away with family during the weekend, and havenā€™t been able to respond until now.

Second, itā€™s clear that the recent team announcement has caused pain to people organizing outreach efforts, possibly damaged those efforts, and is sending a poor signal to those in underrepresented groups. As a member of Rustā€™s leadership, I am deeply saddened by those effects, and Iā€™m sorry.

A bit of context. The newly-introduced subteams are part of a larger scaling up of Rustā€™s governance, which has several aims, including moving Rust to a greater degree of community ownership, and stepping up and formalizing our commitment to the community code of conduct. I believe both of these aspects are necessary, but certainly not sufficient steps in ensuring the health and diversity of Rust and its community in the long term. (Note also, with respect to explicit goals for the subteams: thatā€™s laid out in a fair amount of detail in the RFC itself, and additional details about subteam operation were laid out in the subteam announcement.)

That said, as the original post points out: we are sorely lacking in diversity right now. This is true of the Rust community in general, but it is even more troubling at the leadership level. It would be easy for a person in an underrepresented group to get the impression: this project is not for you.

Looking backward, I see at least two mistakes:

  1. By far most importantly, the core team (of which I am a member) has failed to make diversity a visible, front-and-center priority. While we all want more diversity, and we have done work to foster a reputation as a friendly, safe, and approachable community, we havenā€™t been sufficiently outspoken about addressing the lack of diversity, nor have we invested heavily enough in doing and supporting outreach to underrepresented groups, or to removing obstacles in their way.

  2. Specifically, in making the subteam announcement we failed to acknowledge the lack of diversity and highlight concrete plans to improve it.

As to the makeup of the teams: I absolutely think our leadership should be more diverse than the community, for many reasons, and along many axes. But, like others in this thread and elsewhere, I have thought long and hard about whether there are members of underrepresented groups in the community who would fit these roles but were overlooked, and I am struggling to find them. In other words, I think that the makeup of the subteams is a symptom of a deeper problem: our failure to attract attract a more diverse community in the first place, and in particular to foster leadership among underrepresented groups.

I do not think the answer is to roll back the subteams. We need to scale up Rust and its community involvement in this way (as the RFC argues in detail), and have already gone through a community process to establish this new structure. We need to move forward, in some way that begins to address the more fundamental problems.

Here are a few ideas for immediate steps we might take (in addition to others that have already been discussed):

  • Form a community subteam, as many have suggested. Such a team has been in mind since we first discussed subteams, but we had intended to wait until the initial teams were fully running before bringing more online. We should fast-track the discussion of this team, and consider an explicit element of its charter to be leading the effort, and making policy decisions, related to diversity within the Rust community. That is, while diversity is a problem all of us should be aware of and working on, this team would be tasked in part with leading our efforts, including raising awareness within our community.

    Ideally, this team would be led by a member of the groups we wish to better serve ā€“ which would mean inclusion on the core team as well ā€“ but as @mbrubeck and others have said, we need to do this in a way that doesnā€™t ask people to ā€œdo disproportionate or unpaid work to fix problems they didnā€™t createā€. In any case, the first step is to get the ball rolling on the creation of the team itself.

  • Seek to expand the moderation team with representation from additional groups that it is trying to protect; doing so will help the team to better achieve its goals, and will send a more clear signal to those groups that we care about the safety of our spaces for them. Again, though, we need to find a way to do this that does not ask for disproportionate or unpaid work.

  • Publicly address our diversity issues. This should start with a blog post on http://blog.rust-lang.org/, a very visible platform, that draws explicit attention to the lack of diversity on the teams and within the community, and outlines the steps we are taking to address it (i.e., the action items that come out of this and other discussions). In addition, we should look for more permanent places to put this information. For example, a similar acknowledgement might go on the ā€œteamsā€ page, to make clear our dissatisfaction with the status quo, our strong desire to find leaders from underrepresented groups, and links to our ongoing efforts to do so.

We also need to identify additional avenues for long-term investment in diversity, and I hope this thread can start moving in that direction once we have some agreement about steps to take for the immediate situation.

18 Likes

I can't believe my eyes.

I hope this isn't meant seriously or that there's some kind of misunderstanding, because the way this paragraph is worded is terrible. It sounds like you basically assumed a very negative opinion of the Rust community (a "red flag" as you've put it) without any actual negative experience with the people. Only based off the fact that they have a certain skin color / sex / etc., you assumed they'd be discriminative against others. That is the very definition of prejudice and that's the very thing Rust people are trying to avoid here.

Gender and/or color of skin of authors of a language should no way be an acceptable reason for someone to stop learning that language.

8 Likes

Thank you, @skade, and I apologize for this. I was hesitant to use stronger language in my initial post here, not because I don't think it's warranted, but because I hoped to start the conversation in a way that didn't immediately put people on the defensive. However, you are right. This is not okay, and it was not unavoidable. I am personally ashamed at my own part in failing to prevent this mistake.

4 Likes

Usually such assumptions are based on experience with groups in similar fields with similar make up. These are warranted assumptions and based on these assumptions, prior experiences and opportunity costs, people in under-represented groups make the decision not to get involved.

4 Likes

These assumptions are in no way warranted. I can't believe you're seriously suggesting an assumption that some white people are going to be bad just because some other unrelated white people are behaving bad is warranted. Again, that is the very definition of racism - you assume something negative about someone just based off their skin color. This can't happen. This is a violation of Presumption of innocence.

I can't believe this is happening. Please someone wake me up.

2 Likes

No, just no. Stop it right there, please.

The whole "reverse racism" argument is absolute nonsense. Please read something on this topic, before arguing about this further. I recommend something like this: 7 reasons why reverse racism doesn't exist

Also, no one said this was specifically about race. It's just that people tend to be extra cautious with groups that strongly resemble other groups that they had negative experiences with.

After a certain amount of negative experiences, people do this just out of self protection, not discrimination.

7 Likes

One fairly straightforward way to avoid asking for unpaid work would, of course, be to offer it as paid work.

3 Likes

I also hesitated to talk about this for similar reasons, but I think it's important to address it. For the record, my ancestry is half Asian and half European. @Manishearth has mentioned that he is non-white. While I don't know for sure how any other members of the team identify, I think it is fair to say that the racial diversity of the teams is just as much an issue as the gender diversity. If the teams, including the moderation team, are all or nearly all White and Asian, then we lack representation from the groups that face the most racial discrimination in tech. (I don't want to minimize racism experienced by any Asian people in our community, but just to acknowledge that we are not one of the highly underrepresented groups in this field, at least where I live.)

I hadn't yet mentioned representation of gender and sexual minorities, again for similar reasons. But I know it's an equally important issue for both the moderation team and the community as a whole.

One team will never include every intersection of interests, but we need to do better than today. And where we do lack a point of view within the team, we need to at least be aware of that and take active steps to counter it.

2 Likes