Note: this post, and the subsequent replies, were moved from this thread by the moderators.
I would choose “Other” “Not Answered” myself because I don’t see gender to be relevant. Why pick gender in particular over other group categorizations (race, ethnicity, disability status, BMI, skin colour, age, height, baldness, financial status, etc.)?
I support discarding it.
Why not the option “Not answered” ?
EDIT: I mean, it’s fine not to answer, but “other gender” is an entirely different category in my mind, and I respect that the two genders are not a sufficient choice for everyone who wants to answer. (I don’t recall how exactly it looked in the survey anymore, but it’s in the results at least.)
Uh, I’m sorry. I posted I didn’t want to discuss it and here I went. I need to try harder.
I also agree with the conclusion of the marketing team to scrap the “gender” question.
The Gender Identity stats are not looking Great, quite a gap. I’m sure the last thing we’d want is to increase it by not paying attention to it.
That presumes that they don’t, and it would be reasonable that they do judge based on the available data, namely Gender differences and bias in open source: pull request acceptance of women versus men [PeerJ].
This is actually a very interesting open problem for open source in the age of GitHub, profile pictures and real names. There’s an interesting talk about it here https://youtu.be/2CVPRObSQzc.
This paper is risible, and illustrates everything iffy with DE&I industry. They actually found that women’s contributions are accepted higher than men’s, but had to try really hard to find a niche factor where that was not the case (ie., outside contributors reviewed by “gendered” profile). And yet they go onto claim, without evidence, that this niche statistic (that falls outside the norm) is because of “gender bias”.
Idk but posts like this make me feel like DE&I should be taken more seriously not less
Post like what, exactly?
Before anyone wants the community at large―which includes people from varied backgrounds (including the different nationalities)―to adopt what they personally feel to be taken more seriously, they need to rationally demonstrate their case. My issue with gender, in particular, being included in NixOS survey comes down to two points:
DE&I efforts have been repeatedly demonstrated not to work and even counter-productive; and anybody who wishes to mindlessly push this without any considered explanation surely must be unwittingly strutting the virtue signalling territory, which is far removed from actually caring about one’s fellow human beings. One’s efforts will be better spent focusing on encouraging the various people―regardless their identity―to join the community using a shared background of interest and passion in the topic at hand (Nix/NixOS). There are many factors that underlie why a certain group is not represented at the same proportion as the population, and not all of it necessarily comes down to bias & discrimination.
So yes, the “niche” was that if they were identifiable as women, and not established contributors, there was a gender bias? Where else would you expect to find a bias? And yes, they found that a women, if unidentifiable as such had a higher chance of getting merged.
If you find that funny, then there you go, you’re part of what’s creating that divide.
Why pick gender in particular over other group categorizations (race, ethnicity, caste, nationality, disability status, BMI, skin colour, age, height, baldness, financial status, etc.)?
and they actually found that women are more welcome than men (going by PR acceptance).
This is just a very flawed and superficial understanding of the study. It showed that despite getting merged more if unidentifiable as women, if they were identifiable as such, they would get merged less.
DE&I efforts have been repeatedly demonstrated not to work and even counter-productive
That’s a guest opinion, I linked a study. I hope we are all aware there is a difference.
Again dude, I linked a study and a talk, that you didn’t understand, and now you’re linking newspaper articles. Maybe think a bit before calling others mindless. And defending diversity training is not the same goalpost as saying that having this data is crucial.
https://discourse.nixos.org/t/nix-community-survey-2023-results/33124/15
This post (the one I was replying to). But the one I’m replying to now is another great example.
I really struggle to see you engaging with this topic in good faith when this is your profile’s description: Undoing Woke Invasion – Sridhar Ratnakumar
I don’t know about you, but I find these types of conversations pretty toxic. And ironically is worse for our community than any outcome of this conversation. It pushes the people away who just want to hack and be left alone. And without those, the project doesn’t exist.
It’s interesting because we get along pretty well when we talk about technology, but then in topics like those, we find out that we have different opinions, backgrounds and cultures. And instead of rejoicing in the diversity that we have, it becomes a battleground to tell what people should think. I think this is the point where it’s good to stop and take a different approach.
If you want to help with diversity, then help me identify the factors that would prevent women from participating in the community. There might be some sensibilities that are different that I am not aware of and that we could cater to?
I don’t know about you, but I find these types of conversations pretty toxic. And ironically is worse for our community than any outcome of this conversation.
If I don’t speak up when people wanna remove metrics on women and others in NixOS, it might just happen. That’s why.
And @TeofilC does make a great point, at least to me, that if @srid has a whole section on “wokeism”, then perhaps there isn’t as much good faith as I presumed, and it makes it easier for me to realize that engaging with him might be a waste of time.
I’d take a step back and say “what are factors that prevent people in general”. Here, the lack of a project wide code of conduct is an obvious step. Hopefully those RFCs pan out.
Then, instead of speaking of women, we might just generalize to non-men (since women and other are both low). I’d say that one thing we could do is not become blind to their participation, as @srid suggests. Keep the gender identity statistics.
If we “first do no harm”, then perhaps more opportunities will present themselves. I could spitball on those, but it’s usually good to keep somewhat focused on a few issues at a time.
This is about diversity training, not gender bias in open source. I’ve never said we should have diversity training.
By the way, did you know that most published research findings are false?
Yes, I’m very aware of the replication crisis. Now, if you were suggesting that some of the studies that were used as references in the study I linked, I’d say yes, those are affected by it. I however don’t think this falls as cleanly under that scrutiny.
Moreover, it is glossing over the most important fact that women’s contributions overall in general are accepted more than men’s anyway.
They don’t gloss over that. It’s a central point. It’s surprising that women get merged more if it’s not obvious they are women. It’s literally the finding that enables the other conclusion, that if gender is visible, their “apparent” better merge rate becomes much worse. If they got merged less whether or not it was visible, it wouldn’t be as easy to conclude there was a gender bias.
he study you linked to on the other hand is just male bovine faecal matter1.
I don’t wanna attack your character, but I am skeptical of how you could be a positive influence on a discussion about increasing inclusive of women, specially given your views Wokeism – Sridhar Ratnakumar.
I also wanna note that a particularly view you have makes me no longer interested in engaging with you. I’d call other readers attention to Section 2.4 of this briefing produced for the European Parlaments INGE Committee, and I hope others find your subscription to such ideas as concerning as I do.
as, unlike you, I care about actual results not virtue signalling
For eg., I do encourage the people in my network – men or women – into getting excited about tech and have been doing so from many years ago.
Your encouragement is just words. I produce results, you don’t. I mentor women, I make welcoming spaces for women in free software, and I take the burden of dealing with people like you so others don’t have to. To you this is just about people seeing your blog and your twitter. For me, this is about making NixOS an actual alternative to Guix for Women, including myself.
I’ve learned that it is not such much an imagined discrimination as victim mentality that prevents someone from learning and engaging in anything.
Then you must have quite the victim mentality given your inability to understand scientific studies, and an equally paired inability engage with things that you don’t understand.
Ultimately, your criticisms are projections.
I moved the whole conversation out as it was derailing the discussion. Remember that other community members are reading these conversations.
see Nix Community Survey 2023 - discussion about gender stats - #11 by zimbatm
You can’t tell from my username or profile picture, but I’m female. I’ve recently retired after 38 years in in software development and AI/ALife research and I do a bit of mentoring as well. I enthusiastically support diversity of all kinds, not just because I think it’s the right thing to do, but also because it makes life more interesting. I wanted to speak up and share this information about myself in the hope that it helps others feel a bit more welcome.
I think a significant number of us software folks are introverts who have always felt a bit different in some way. I myself was socially awkward and geeky/brainy. Lately I realised that by making everyone feel welcome, I myself feel more accepted in the community.
IMO, ideally decisions like these should be made depending on goals.
For example, let’s say there weren’t enough contributors. A small feedback button is added somewhere “How did you feel your contribution or attempt at contributing went?” → multiple choice and a free text field. If out of that comes that people attempting to contribute feel unwanted because of a lack of accessibility (or something else), then we should look at how that can be improved and which questions to ask and metrics to track to follow progress e.g “are you disabled?” → “do you consider X accessible?” would be one way to do it for this example.
Tracking gender, ethnicity, nationality, political alignment, religious leaning, etc. (these are just examples) should be included to achieve an agreed upon goal. The entry/answering thereof should also be made optional.
Since we don’t live in an ideal world and everybody has limited time, the options I see are
All in all, if they’re optional, I don’t see the harm in including them 
I genuinely fail to see any reason to remove them if they’re optional and give us data about the community.
When the argument for removal is “but what about baldness”, then that can be used on any of these metrics. This becomes a very convenient slippery slope, and I think we should outright reject it. The whole point of a community survey is to survey the community. Removing data just means removing insight.
A bit of a meta point on the discussion: the only people who need to be convinced are the survey team members. They have the autonomy to decide what questions they put in the survey.
Personally, I believe that one question like this doesn’t hurt. But it has to stay the same every year as the primary utility is to be able to observe trends. Finding what to do to fix the imbalances requires other, more useful questions.