Diversity in Technology and Open Source

http://lucumr.pocoo.org/2017/6/5/diversity-in-technology/

by Spiritus • 9 years ago

41 75 9 years ago

lucumr.pocoo.org

b62• 9 years ago

Nobody opposes diversity of experience, perspective, skill, etc. It really is helpful to have people with different use cases testing and developing the software.

But "diversity" is being used to mean diversity of gender, ethnicity, etc., as if it matters, as if it is a reliable proxy for those other things like experience and perspective and skill, when it's not.

When I was starting to program and talk to people on the net, nobody cared about my age or ethnicity or gender. I was just a name and an area code and some skills. The compiler certainly did not care about my ethnicity or gender. It was perfectly fair to me.

I will not go out of my way to try to rope people into my projects because they have different color skin. If people show up to help, I welcome them. I don't particularly care about their genetic code. I don't have any preconception about what the graphs of the ethnicities and genders of contributors would look like. Who cares? We are not our bodies and I won't help build a future where body attributes are considered an important part of identity.

fapjacks• 9 years ago

This is one of the most well-spoken comments I've ever read on HN.

atomi2• 9 years ago

as if it matters

In terms of programming, where the domain is very specific, I somewhat agree that it doesn't matter.

But in a more general sense, since it's life experiences that culminate into who a person is, their perspective will be based on their identity. And it will vary in unexpected ways.

If you're a large company in a diverse market like the American market or a global market, you don't want to appear tone deaf in terms of your marketing and your products.

So it's important to get differing perspectives before that new launch.

wyager2• 9 years ago

So it's important to get differing perspectives before that new launch.

Sure, which is why your marketing team has to be familiar with regional cultures and customs.

It doesn't mean you need to account for every target cultural group on your IT support staff.

Bartweiss2• 9 years ago

I definitely appreciated the Unicode example in this article. It's a nice, clear case where having more diversity (in this case, someone who regularly uses a non-ASCII language) on a team can increase value rapidly. Having someone around who says "hey, this system won't handle my name right, that's super annoying" can add value to the project for a few billion people.

Obviously team diversity isn't the only way to achieve that gain, but it's a good one that's often lacking.

tptacek6• 9 years ago

This comment doesn't say anything. Except in very rare cases, people aren't premising their arguments against "diversity" on overt animosity for women and people of color. All you're doing is restating the fundamental dispute: "is it acceptable to continue the practices that have led to an industry disproportionately lacking women and people of color?"

The "yes" argument obviously says compilers don't care about gender or skin color. The "no" argument says that arguments in favor of inequitable staffing are a form of the Just World Fallacy.

You need to do better than just restate those arguments to add something new.

b62• 9 years ago

There's no need to add something new. I'm just resisting the rise of these new more insidious forms of racism and sexism.

KirinDave12• 9 years ago

But "diversity" is being used to mean diversity of gender, ethnicity, etc., as if it matters, as if it is a reliable proxy for those other things like experience and perspective and skill, when it's not.

Yes, b6. It is. Your personal experience is of course unique. While your experience may have familiar notes with many people, it's fundamentally different from someone who grew up as a young girl. Your fears and the options presented to you are going to be different.

Identical people come to similar conclusions. Folks who have raised this idea of "diversity of thought" think that it adequately addresses this idea that diversity is a value add for software by saying, "Well it's all about thought." But those initial conditions our minds are so sensitive to are not easily ignored.

You cannot have it both ways in this. Either you simply don't care about diversity (which is a value decision I am not denying your ability to make), or you do and physical indicators that shape how society treats people are a high quality signal of diversity. You cannot rationally suggest both that diversity is important but that people's traits and backgrounds aren't.

People try to turn this into a value proposition here (as it seems like you may be doing) are taking half the situation into account. For a value proposition, you're looking at your project's upside. For example, a product primarily aimed uniquely at women would probably be best served by a qualified woman. But even in this, bias creeps in (the famous study results suggesting men are so acclimated to dominating the conversation that when women talk for half the time, men perceive it as oppressive). In this specific scenario, women have reported being hired and then simply not given the power to do their jobs, or not being present in larger orgs with sufficient numbers to create the impact they were hired to create.

But the diversity movement is also trying to reduce the downside for individuals who often find themselves on the brunt end of unfair standards, inappropriate behavior and discriminatory teammates. You may talk all day about how "my compiler doesn't care" but this is you looking down at your desk quietly as folks all over the industry go through terrible and unfair treatment. Your compiler doesn't care about that, but you should.

You suggest that changing this status quo would actually be a more subtle form of racism or sexism. This is true, insofar as the top article accidentally falls into a paternalistic role and offers open source housekeeping roles to women as a token offering, thinking this is an incentive. In terms of damaging the fortunes of white men at the expense of offering other people, the truth is that correcting for gender and racial bias in our industry will bring us closer to a meritocracy and offer more potential for talent to emerge than the current state of affairs.

ManFromUranus2• 9 years ago

Identical people come to similar conclusions. Folks who have raised this idea of "diversity of thought" think that it adequately addresses this idea that diversity is a value add for software by saying, "Well it's all about thought." But those initial conditions our minds are so sensitive to are not easily ignored.

I don't know what is being said here. Are you saying that women or people of color would implement software differently from a white male? How would writing an encryption app differ if a women wrote it? You seem to think that there is something being missed by the Poster you are replying to but I can't figure out what it is exactly.

AnthonyMouse3• 9 years ago

You cannot have it both ways in this. Either you simply don't care about diversity (which is a value decision I am not denying your ability to make), or you do and physical indicators that shape how society treats people are a high quality signal of diversity. You cannot rationally suggest both that diversity is important but that people's traits and backgrounds aren't.

But that's the whole issue. It's about backgrounds, not traits. You can't tell someone's background from their skin color, it's just a proxy. And if you optimize for the proxy instead of the thing itself you end up selecting black men from affluent families in the suburbs, which minimizes the difference in perspective they have compared to white men from affluent families in the suburbs.

You would end up with more actual diversity if you had chosen someone from Russia or Japan or Germany (or even a different part of the US) who has an even more different perspective. Even if that person is also a white or Asian male. Instead of selecting for the people who add the least actual diversity to your organization but allow you to check a box.

humanrebar2• 9 years ago

But those initial conditions our minds are so sensitive to are not easily ignored

When we boil people down to their demographic membership, we do exactly that. We decide that a latino son of a doctor from Cleveland has more in common with an Nicaraguan in on an H1B than with a Caucasian son of a doctor from Cleveland.

You cannot rationally suggest both that diversity is important but that people's traits and backgrounds aren't.

First, in specific cases, you can argue that it doesn't matter. A core dump is a core dump.

Second, you can argue that boiling people down to a handful of traits and background is saying the rest of the considerations are mostly unimportant.

What if you have a candidate whose parents were Amish? That's an incredibly unique candidate, so he should be off the diversity charts, right? But the current model we apply to determine diversity lumps him in with all the white boys who go to tech magnet schools in the Bay Area.

belorn5• 9 years ago

Historically, class difference has been one of the major qualifiers for grouping people with different experience. A person growing up poor on a farm and eating roots and tree sprouts as candy substitute (not a joke) is not going to have much in common with a person growing up in the city by rich parents and who could just go across the street and buy something whenever they wanted. Much of last century of political struggle on a global scale has been about class, and much of current political landscape is made from it. Fears and the options presented to you are much more dictated by class than race or gender (which a fair amount of study concludes over a long range of subjects).

The diversity movement do not care about class. A poor white man is hated as much as a rich white man. I can only conclude that "personal experience" and "fears and the options presented" are not factors.

a_imho• 9 years ago

I don't agree with the implied correlation of user friendliness and diversity, let alone the supposed causation. The Gimp example seems a forced attempt to support Armin's agenda.

Disclaimer: I don't know how diverse the Gimp developer base is as opposed to a user friendly competing product.

HelloNurse3• 9 years ago

Let's focus on GIMP, it's a good example. We know that none, or not enough, of the important GIMP developers who made the important UI decisions had particularly good taste for UI, resulting in mediocre choices. But how can we distinguish lack of diversity due to social and economical factors filtering the pool of potential contributors, perfectly fair self-selection of a project committee of people who simply think alike, and random unintentional lack of skill? Would investigating the issue accomplish anything besides aggravating GIMP contributors? Moreover, useful plans should be about concrete attempts to improve GIMP by contributing patches and convincing the maintainers to make changes; attempting to make the GIMP team more diverse in the hope that it performs better is pure wishful thinking.

alexfi• 9 years ago

I think IT and especially open source is one area, where your gender, background, sexual orientation doesn't matter. The most open source projects I'm using or working on, are made by people I don't know how they look. The only thing that matters is the ability of writing good code.

The author is in that point right, that there are more male programmers than female. But to solve this problem you have to start way earlier, with getting the interest to tech things of a child (regardless of gender) in school or even earlier.

tptacek6• 9 years ago

Major open source projects are disproportionately managed and staffed by people with full-time jobs at major software companies, and the process of obtaining and thriving in one of those jobs is not intrinsically color and gender blind, so this argument isn't persuasive.

microcolonel2• 9 years ago

Major open source projects are disproportionately managed and staffed by people with full-time jobs at major software companies, and the process of obtaining and thriving in one of those jobs is not intrinsically color and gender blind, so this argument isn't persuasive.

How is it surprising that they work in software? It's like finding out that your welding instructor has worked as a welder; surely somebody who contributes to a field in public is more likely to have a job in that field in private, but does that mean that not having a job precludes you from having a skill?

I mean, I've been NEET going on half a year, and I'm contributing to open source projects, do I need a software job to do open source? I was contributing to open source projects before I was ever employed, do I need to have had a job to do open source?

justicezyx• 9 years ago

process of obtaining and thriving in one of those jobs is not intrinsically color and gender blind

So it's ok now to post a comment with a strong statement without any data?

jasode2• 9 years ago

we need more [...] people that are good in de-escalating arguments in bug trackers and mailing lists, people that take care of documentations, people that make software work in new cultural contexts (localization, globalization, internationalization, etc.), people that care about user experience etc

If you're a male, you have to be careful writing out suggestions like that because to females such as my mother with a career in electronics, it is patronizing. Women can also be a "strong engineers" and they shouldn't be relegated to "open source housekeeping duties" as a dangling carrot to attract them.

Yes, Armin Ronacher means no harm but I think men often fall into the trap of thinking they are championing women's causes when they are actually insulting their intelligence.

tptacek6• 9 years ago

I agree. We ran into something similar when we were working on Starfighter, where Erin and I were repeatedly told that CTFs were intrinsically male-oriented because women disfavored zero-sum competition (which was surprising for Erin, a roller derby skater, to hear).

It also doesn't (at least in my opinion) help the cause of balancing the industry to be telling engineers that the only way to do so is to emphasize soft skills and deemphasize competition (and subtextually with it technical excellence).

aeorgnoieang2• 9 years ago

CTFs were intrinsically male-oriented because women disfavored zero-sum competition (which was surprising for Erin, a roller derby skater, to hear)

Surely you and Erin both recognize that it's possible that both (a) women, generally (compared to men), 'disfavor zero-sum competition'; and (b) Erin favors 'zero-zum competition'.

[I'm not claiming that it's true or even that there's significant evidence that women 'disfavor zero-sum competition', absolutely or relative to men.]

belorn5• 9 years ago

I have seen very small variation on how people emphasis things to have extreme effect on the demographic of participants. To take a different topic than IT, a public bath house in the city have a goal that visitors should be 50% women and 50% men. When they first had special a sauna evening called "Relax" with a emphasis on adventure (ie, a few events had names like Everest), they had full booked for men but only about 1/3 ticked sold for the remaining 50% that was dedicated for women. A month later they held a second evening, this time called "Wellness", they also renamed the events. Everest was now called ice sauna, salt sauna was renamed to scrubbing, but other than the name they the exact same events. This time all the tickets for women was sold after just a day but only 2 tickets for men sold in total. A complete reverse. Later events has shown similar pattern where more "adventure"-styled marketing leads to more men and less women, and self-improvement styled marketing leads to more women and less men.

A small leap I will make is that people of both genders look towards gendered emphasis to validate their choice. If its for their gender then it is a safe choice, and if its not then doubt arise. The effect being a small bias that depending on what one is doing can have a rather extreme results.

dionidium• 9 years ago

"(which was surprising for Erin, a roller derby skater, to hear)."

When someone says "men prefer X" or "women prefer Y" they are quite obviously speaking statistically.

the_mitsuhiko2• 9 years ago

As the author I want to clarify something: the word "gender" appears once in the article as an example of diversity. The enumeration you gave is more an example of lack of diversity in roles and behavior than it is gender.

jasode2• 9 years ago

the word "gender" appears once in the article

But people reading it will see the following words that are all in the same semantic space as "gender":

"men", "woman", "women", "male", "someone" (implied female)

As used in sentences:

When your team is 4 men, the first woman which joins will make a significant impact.
When your team is already 20 men you need to get a lot more women on board to have the same impact.
Tech for recent historical reasons is very male heavy but society is not.
We don't need more [...] more strong alpha males but people that are good in de-escalating arguments
So when someone cancels a conference because the speaker lineup after a blind selection was 100% male it just shows how bad the imbalance

Therefore, it is not unreasonable to get the impression that the article is heavy on gender imbalance and not so much on skin color. (No explicit mention of "white", "black", "African American", etc.)

Bartweiss2• 9 years ago

Simply counting that up seems a bit disingenuous... The word "gender" appears once, yes. But literally all the specific examples given are gender - adding a woman to a team of 4 versus 20, the gender imbalance of tech, not needing more "strong alpha males", and having 100% male conference lineups.

The line under discussion directly contrasts "males" with role diversity. It doesn't explicitly label the other roles as feminine, but it was certainly the impression I got when reading "not more strong alpha males but people that are good in de-escalating arguments".

None of this contradicts your understanding of what you meant with your writing, obviously. And the Unicode point is a good example of how a shortage of how other developer diversity (e.g. in languages spoken) can motivate more useful products.

But for whatever it's worth, I had to read this comment to find out that you weren't writing the entire piece with gender in mind, and writing that diversity of roles line specifically to state what women would bring to the table.

microcolonel2• 9 years ago

You know, I never stopped to figure out the ethnicities and genders of co-contributors on a project, it truly is an unnatural thought. I'm concerned that the people who are most concerned about diversity in "open source", do so because they consider themselves a threat to it.

Maybe white people like diving computers or lake sonar, maybe women care more about topography or teledildonics; but if nobody cares to even know the accidental characteristics of the people who send the patches, then is it an issue when the numbers bear that out?

mafribe3• 9 years ago

    everyone agrees that such overt 
    exclusion, if it exists, is racist 
    or sexist.

I certainly do not agree with this. Your opponents simply deny that the compiler has the ability to exclude or be sexist or racist. The argument that the compiler or test suite discriminates on anything but merit is for you to make. Good luck.

   you can't shut down the argument 
   by claiming that it's somehow racist/
   sexist against white men.

On the contrary, this is an extremely good counterargument, and the reason you try to re-frame it away is precisely because you know it is a good counterargument. Dare we speak of "ptacekian fragility"? There is another re-frame that you maliciously use: you talk about white men, when you should really be talking about white and asian men. The spectacular success of asian men (and -- to a lesser degree -- asian women) in SV (and any profession that needs high degrees of education) absolutely, brutally burns to the crips any claim that there is racism/sexism at play. And you know it -- hence the ongoing reframes.

tptacek6• 9 years ago

No, it absolutely does not. The definition of "whiteness" has changed dramatically over the 20th century to accommodate and include the successes of formerly disfavored minorities while maintaining the lower status of latinos and, most especially, African Americans. Italians, Greeks, and the Irish also used not to be "white".

It is, if you stop to think about it, a little silly to argue that the success of Asians means there's no meaningful disadvantages to being African American.

mafribe3• 9 years ago

   "whiteness" has changed

Maybe we should change the meaning of tptacek to Osama bin Laden -- and argue that you are dead?

Yes, the meaning of "whiteness" has evolved -- for sinister political reasons: equating "whiteness" with "successful" serves no useful purpose, other than covering up inconvenient facts like immigrants/Asians doing extremely well, which doesn't fit the left's convenient narratives. Twisting the meaning of "whiteness" really is an incendiary political mobilisation strategy, and you know it. It is a well-worn strategy of political mobilisation going back -- like almost everything on the contemporary political left -- to Lenin. I recommend you reflect on this moral failure of yours: Why not use something like "successful" instead of "white" from now on, otherwise you will continue to be part of the problem. In order to solve a problem, one needs a clear understanding of what causes the problem in the first place.

   meaningful disadvantages to 
   being African American.

Yes, so? The success of Asian in SV (many of them immigrants from poor countries) shows clearly that the convenient narrative "blame whitey" is not helpful. If you want to help, why not start with a correct analysis of the problem's root-cause?

humanrebar2• 9 years ago

It's also silly to argue that since white and Asian men are overrepresented in tech, that somehow means that impoverished white boys in Appalachia, Khazakstan, or Lithuania get labelled "privileged" and "not diverse".

Honestly, the failure to address that concern is one of the big reasons why nationalist and populist arguments are so persuasive.

err4nt• 9 years ago

Italians, Greeks, and the Irish also used not to be "white".

Are you thinking of the term "WASP" in place of "white"? Italians and Greeks are 'W' but not 'ASP', and Irish are 'WAS' but not 'P' — but I think all three would be considered 'white' by people all around the world and I don't think that has changed.

Perhaps where you live where's a different capital 'W' version of 'White' (contrasted to 'white'), but that's not how most of the world looks at these groups. Do you think the people living Italy, Greece, and Ireland believe they are excluded from being 'white'?

metaphorm2• 9 years ago

you're perilously close to the precipice of post-modern identity politics.

Tomte• 9 years ago

That's a purely American definition, right? To me it sounds incredible that anyone would call Italians or Greeks non-white.

frons2• 9 years ago

Your opponents simply deny that the compiler has the ability to exclude or be sexist or racist. The argument that the compiler or test suite discriminates on anything but merit is for you to make. Good luck.

That's not really how enlightened discussions work. If you have the opinion that the "compiler doesn't care" and that it's somehow related diversity you have to make an argument to the fact. You can't just say that "it doesn't care" and have it be your opponents job to make an argument that disproves your opinion. At least not if your expect to base your opinions on merit.

The notion that compiler reflect the background and culture of those creating and using it isn't exactly far fetched. Of course something like a hobbyist background (historically only available to certain groups of people) is far more important in certain types of language and that the demographics of people differ between languages.

But that not even the point. What you value in the code you're writing is as much, if not more of a factor, in projects. So it doesn't necessarily matter if the compiler is unbiased since what's good code is going to be judged by people.

mafribe3• 9 years ago

   isn't exactly far fetched.

Why don't you discuss this with Jensen Huang, Chi-Chih Yao or Satya Nadella?

dang2• 9 years ago

This comment crosses into personal attack. We ban accounts that do that. Please post civilly and substantively, or not at all.

We detached this subthread from https://news.ycombinator.com/item?id=14488081 and marked it off-topic.

sparkling• 9 years ago

I wont accept your pull request because you are [black/female/gay/asian/latino/transgender/whatever minority of the week]

Said no one ever

KirinDave12• 9 years ago

I will accept your pull request, but be aware on my twitter I was just talking about how much I despise you and every liar like you. Thank you for the free work, you disgusting child-corrupting liar.

Said lots of people just last week.

jp_sc• 9 years ago

Really? Show me five Github projects where the main developer has said something like that at some point this year.

jankotek• 9 years ago

There is a old saying: "On the Internet, nobody knows you're a dog."

It can not get more diverse and inclusive than that.

brighteyes• 9 years ago

To really understand the issue, you need to look at all the facts. And yes, while women are underrepresented in open source, LGBT people are overrepresented - almost double their percentage in the general population in fact (7% vs 4%).

So any theory of "open source drives away women" has to explain why it has the opposite effect on LGBT people. There isn't going to be a simple answer there and maybe we aren't even asking the right questions yet.

HelloNurse3• 9 years ago

There's no conflict, it depends on whether these LGBT people are women. I suspect that few lesbian, bisexual, transgender women follow the typical social patterns that keep women away from open source software: few uneducated housewives, few early marriages and unplanned pregnancies interfering with education and work, little inclination in general for taking care of household chores sacrificing themselves so that their S.O. has some free time for open source work, few and/or late children, etc.

metaphorm2• 9 years ago

There's a bit of an elephant-in-the-room here which I wish Armin had discussed more.

I agree with Armin on what he did mention, which is that there is great need in OSS projects for people with greater focus on skills besides just hard boiled coding.

I wish he had discussed the issue that seems to be creeping up in many places now (recent example: cancellation of Electron Conference) which is that there are many activist groups who are seeking to implement diversity quotas. That is a really scary idea and a step in the wrong direction.

zmoreira• 9 years ago

If “Diversity in Open Source Is Even Worse Than in Tech Overall” and there are less barriers to participation in Open Source it could be that this "lack of diversity" accurately reflects the interests and capabilities of the different groups?

It could be that the different kinds of humans are not identical and given an even playing field different outcomes will result.

It could also be that the lack of diversity in tech is the mirror image of the lack of diversity in other fields such as healthcare and education.

If that is so it could be that the "lack of diversity" is the result of diversity in human inclinations and it would only be stamped out through increasingly authoritarian measures.

HelloNurse3• 9 years ago

One has to be even more privileged to spend time on open source work than to hold a well-paid IT job; there are additional barriers.

For instance, open source contributors need free time, which is usually a consequence of relatively short working hours (i.e. being highly competent and educated) and short commutes (i.e. an expensive house in a convenient location), and not having to spend a lot of time, when at home, on housekeeping, children, old and ill family members, hostile environment, distractions, etc.

Klockan• 9 years ago

Isn't it strange that diversity gets worse when you add anonymity?

aeorgnoieang2• 9 years ago

All of the members of the [Pocoo Team][1] seem to be white men.

[1] http://www.pocoo.org/team/#team

Is that a reflection of your hypocrisy? Or the difficulty of recruiting people that are not white men? I'm guessing it's mostly the latter.

You wrote:

When you start an Open Source project today, in particular one which is further disconnected from frontend technologies there is a very high chance the organic community development will be everything but diverse.

and given the following, from [this comment][2] in this thread, which seems probably true:

Major open source projects are disproportionately managed and staffed by people with full-time jobs at major software companies, and the process of obtaining and thriving in one of those jobs is not intrinsically color and gender blind, so this argument isn't persuasive.

[2] https://news.ycombinator.com/item?id=14488000

we should be tempering our judgements of open source projects, e.g. that they're not welcoming to people that are not white men.

I'm confused as to what principle or principles you think should actually be adopted. Should all open source projects reflect the 'diversity' of the entire world?

Consider the following, from [this comment][3] also from this thread:

LGBT people are overrepresented - almost double their percentage in the general population in fact (7% vs 4%).

[3] https://news.ycombinator.com/item?id=14488689

Assuming that the above is true, is this a cause for concern? Would you similarly be concerned if it were true that, say, Asian people, or even just Asian men, were over-represented in technology or open source software projects? Is that not a cause for concern for you too?

It sure seems like the only cause for concern is that there are too many white men. Would anyone ever criticize an open source project for not having 'enough white people' or 'enough men', let alone 'enough white men'?

Diversity in Technology and Open Source

Related Stories