Closing a door

This post has been sitting in my drafts folder for a year now. It has never been the right time to post this. I have always been worried about the backlash. I’ve skirted around talking about this issue publicly for some time, but not acknowledging the elephant in the room has eaten away at me a bit. So, here goes.

Here’s the deal: I’m not a Linux kernel developer any more. I quietly transferred the maintainership of the USB 3.0 host controller driver in May 2014. In January 2015, I stepped down from being the Linux kernel coordinator for the FOSS Outreach Program for Women (OPW), and moved up to help coordinate the overall Outreachy program. As of December 6 2014, I gave what I hope is my last presentation on Linux kernel development. I was asked to help coordinate the Linux Plumbers Conference in Seattle in August 2015, and I said no. My Linux Foundation Technical Advisory Board (TAB) term is soon over, and I will not be running for re-election.

Given the choice, I would never send another patch, bug report, or suggestion to a Linux kernel mailing list again. My personal boxes have oopsed with recent kernels, and I ignore it. My current work on userspace graphics enabling may require me to send an occasional quirks kernel patch, but I know I will spend at least a day dreading the potential toxic background radiation of interacting with the kernel community before I send anything.

I am no longer a part of the Linux kernel community.

This came about after a very long period of thought, and a lot of succession planning. I didn’t take the decision to step down lightly. I felt guilty, for a long time, for stepping down. However, I finally realized that I could no longer contribute to a community where I was technically respected, but I could not ask for personal respect. I could not work with people who helpfully encouraged newcomers to send patches, and then argued that maintainers should be allowed to spew whatever vile words they needed to in order to maintain radical emotional honesty. I did not want to work professionally with people who were allowed to get away with subtle sexist or homophobic jokes. I feel powerless in a community that had a “Code of Conflict” without a specific list of behaviors to avoid and a community with no teeth to enforce it.

I have the utmost respect for the technical efforts of the Linux kernel community. They have scaled and grown a project that is focused on maintaining some of the highest coding standards out there. The focus on technical excellence, in combination with overloaded maintainers, and people with different cultural and social norms, means that Linux kernel maintainers are often blunt, rude, or brutal to get their job done. Top Linux kernel developers often yell at each other in order to correct each other’s behavior.

That’s not a communication style that works for me. I need communication that is technically brutal but personally respectful. I need people to correct my behavior when I’m doing something wrong (either technically or socially) without tearing me down as a person. We are human. We make mistakes, and we correct them. We get frustrated with someone, we over-react, and then we apologize and try to work together towards a solution.

I would prefer the communication style within the Linux kernel community to be more respectful. I would prefer that maintainers find healthier ways to communicate when they are frustrated. I would prefer that the Linux kernel have more maintainers so that they wouldn’t have to be terse or blunt.

Sadly, the behavioral changes I would like to see in the Linux kernel community are unlikely to happen any time soon. Many senior Linux kernel developers stand by the right of maintainers to be technically and personally brutal. Even if they are very nice people in person, they do not want to see the Linux kernel communication style change.

What that means is they are privileging the emotional needs of other Linux kernel developers (to release their frustrations on others, to be blunt, rude, or curse to blow off steam) over my own emotional needs (the need to be respected as a person, to not receive verbal or emotional abuse). There’s an awful power dynamic there that favors the established maintainer over basic human decency.

I’m not posting this for kernel developers. I’m not posting this to point fingers at specific people. I’m posting this because I grieve for the community that I no longer want to be a part of. I’m posting this because I feel sad every time someone thanks me for standing up for better community norms, because I have essentially given up trying to change the Linux kernel community. Cultural change is a slow, painful process, and I no longer have the mental energy to be an active part of that cultural change in the kernel.

I have hope that the Linux kernel community will change over time. I have been a part of that change, and the documentation, tutorials, and the programs that I’ve started (like the Outreachy kernel internships) will continue to grow in my absence. Maybe I’ll be back some day, when things are better. I have a decades long career in front of me. I can wait. In the meantime, there’s other, friendlier open source communities for me to play in.

When one door closes, another opens; but we often look so long and so regretfully upon the closed door that we do not see the one which has opened for us.

– Alexander Graham Bell

(FYI, comments will be moderated by someone other than me. As this is my blog, not a government entity, I have the right to replace any comment I feel like with “fart fart fart fart”. Don’t expect any responses from me either here or on social media for a while; I’ll be offline for at least a couple days.)

Edit: I would highly recommend you read my follow-up post, “What makes a good community”

Edit 2: Please stop suggesting BSDs or Canonical/Ubuntu as “better” communities.

235 thoughts on “Closing a door

  1. I’m saddened to see you go.

    I’m saddened by the fact that you, and other people are dragged down.

    I personally (a young white European male) felt very nervous sending in my first patch to the kernel, and after that haven’t had the courage to do more with it. I was treated right and even got compliments, but still.

    I see the stories about people being shouted at, being torn down everywhere.

    So I’m sad to see a voice of reason and professionalism go.

    I wish you well in whatever comes next, and given your history, have little doubt that it’ll be well and truly awesome.

  2. I don’t know if/how much shit you’ll get about this, but just in case I wanted to be a voice of support. If anything, I think you underestimate (or at least understate) the extent to which your preferences in this matter are genuine moral goods and not mere preferences.


  3. Well said. No one deserves to be disrespected this way, especially when volunteering your time and energy.

      1. I am just an ordinary Linux user and know nothing about developing kernels.
        I am nonetheless always sad when I read about discrimination in the linux community, especially discimination against women.
        To clarify that, I would add that I am also very wary of all- encompassing buzz-words like “sexism ” which appears to many to be exclusively anti-male, but that is another issue.
        I am sorry that you have been made to feel so uncomfortable Sarah and unwelcome in the linux community, but , I think like all communities, groups, etc, it is those who are in control of the group/community who are the prime cause of the problem.
        They are the only ones in any position of power to change things and too often, as appears to be the case now, they turn a blind eye to a painful truth, preffering instead to regard personal abuse as ” harmless banter “.
        Incidentally, despite the spelling of my name, I am actually male , at least, physically.
        You have my respect,and my gratitude for the work you have done and my good wishes .
        I wish you every success for your future .

        1. I don’t see this as a male/female thing, but as a matter of mutual respect. There is nothing wrong in criticizing an argument or given approach, although there are always ways of softening the blow. There is never a need, however, to make the criticism personal. That is not honest, but emotional immaturity.

  4. It won’t change because the one person who could change it – Linus – won’t do so.

    Not that it will matter, vendors are giving up on upstream. They commit to company trees, grab changes from each others code dumps or work together on group trees like Linaro.

    The cost of upstream for anything but big servers now outweighs the benefit.

  5. It is disappointing that a community that should be dedicated to good values (open source software, a fantastic product, etc.,) should come to this.

    There are concerns in the R community over a certain member of the Comprehensive R Archive Network who creates a similar atmosphere (without the sexism, but overall equally hostile). I wonder whether this is somehow related to practices that were prevalent and ‘acceptable’ in tech in the 1990’s when these organizations were founded, and whether there is a way forward from here.

    1. Haha. So I’m not alone? I gave up on R right when I wanted to get in. And I’m told I’m too blunt sometimes. Hostility doesn’t have a limit it seems.

      The thing is, it takes social maturity to understand what you’re talking about. I definitely have become much milder as I’ve aged, which means I probably wouldn’t agree with you like 7 years ago.

      Unfortunately respect and peaceful dialog are not valued by many abundant open-source projects and communities and their senior members.

    2. Off the topic. When I submit a package for R, I will spend a lot of time on double checking my codes. I really appreciate R core team can spend time checking every package and be straight about your error, and it teaches me a lot, although it make me fill a little bit scared sometimes. 😉

  6. If you don’t think well in one community feel free to help others open source OSs projects needing help , specially in graphics stack. Then you will be promoting diversity of products too.
    There are projects which have a core team elected democratically by theirs developers.

  7. I’m sorry to hear this. I hope that your departure is both an opportunity for you and serves as a catalyst for change in the development community. Good luck.

  8. Thank you for all your work, and thanks especially for giving a talk at our school last year (Waterloo). Just because you are bothered by it doesn’t mean it’s your responsibility to fix it, and it’s their loss.
    Best of luck on your future endeavours.

  9. sounds reasonable.
    I whish you more fun with mesa – good to see another developer helping out there, we need openGL 4.x asap 🙂

  10. Thank you for the fine work you have done on kernel drivers that I rely on every day.
    As the father of three smart, motivated Daughters, thank you also for being a voice for change.

  11. A very sad day for the Linux community. That said, I completely understand. As a person who comes from a Free Software community who have their own struggles with this dynamic (and increasingly getting better), I can sympathize. We also have our own frustrations with the Linux community.

    The linux community very much tends to cling to the old mainframe / Unix society of the 80’s. There is nothing particularly special about it other than that era (having lived in it) other than the notion of sharing ideas.

    Bet of luck to you.

  12. Thank for you having the courage to write about this issue. We’ve seen glimpses of this kind of bad behaviour on mailing lists and IRC and other forums and it’s great that some open source communities have started with a friendly healthy basis for communication.

    Hopefully the cultural changes a little faster thanks to your post.

  13. Ms. Sharp,

    I believe time and energy spent berating and insulting are completely wasted; I no longer remain where they are the norm. Those are some of the behaviors we at do not tolerate. We’re a small community of friendly people (and an old grouch or two).* We focus on improving and enhancing the system and ourselves. Come visit. And recharge your faith in humanity.


    * – Shamelessly stolen and adapted from an old billboard outside Portales, NM.

  14. It must have taken a lot of courage to take this decision and finally write this post. Thanks for all the hard work you did and I hope to see you in another open-source projects 🙂

  15. Come to FreeBSD — we have a code of conduct! The kind of horrible vitriol that is acceptable in Linux is not acceptable in the FreeBSD developer community.

    1. Uh, FreeBSD has a terrible code of conduct and has its own problems. It is not a safe place.

      1. I’m sorry you feel that way. I won’t deny that FreeBSD doesn’t have its own problems, and won’t pretend that it’s perfect.

        1. Possibly the “Do not take it personally” and “Try not to take offense where no offense was intended” language. Sure, that could be better (just strike those sentences).

          I don’t see how that makes it a terrible CoC, though.

      2. I haven’t been in the FreeBSD developer community for over 15 years, so I can’t comment on that part; you’d know better than I.

        However, as a statement of intent, the code of conduct seems pretty good to me on a first reading. Am I missing something obvious?

        1. Look at the comments which happen in the community. Try to send complaints about ones which clearly violate the letter and spirit of the code. When you are told that they have no interest in stopping the abuser you will understand why FreeBSD’s code is little more than wallpaper.

          1. Who were you sending complaints to? When did you send them? I’m surprised that this was your experience, because I had a very different one.

            I sent complaints to core about the person abusing Randi on Facebook (not even in a project context). They took it seriously and the person has departed the project. That process took some time (as there were also efforts at mediation that proved fruitless that I was involved with). They created the CoC based on that incident. They’ve revised the CoC based on feedback that its language didn’t fully reflect the culture that the project wants. They are most likely open to further revisions or rewrites so it reflects the inclusive and safe environment that people in the project want to create.

            I get that Randi has a different perspective on things.

  16. I don’t know if it would be your thing, but we’d welcome you in the WordPress community, and I think you’d find the community you’re looking for. Mostly. 🙂

  17. Thank you for reminding us that the standard we walk past is the standard we accept and to keep pushing for better communities. I hope one day nobody will never need to “fart fart fart” comments

  18. Sorry to hear about your decision to leave the kernel community.

    The older I got, the more frustrated I got about other people’s stupidity and lack of respect for other human beings. I think I know how you feel.

    Lets hope that you find an open source project with a less toxic work environment one day where you enjoy working.

  19. Thank you Sarah for all the work you’ve done. You don’t need to feel guilty. You made the community a better place. Have a good time doing whatever else you’d like to do!

  20. I am sad to read this, but not surprised.

    Sad, both because the Linux kernel is losing a great developer, but also because it mirrors my experience with other FOSS communities and people from FOSS communities working on non-FOSS stuff. It’s been the major reason I have chosen not to join any established communities I was expressly invited to. Like you, I can take technically brutal comments if they are personally respectful.

    I would add that it’s my impression that these communities are not, in fact, meritocracies as they claim to be – though of course merit does factor into rank. Rather they are communities dominated by fairly typical alpha male behaviour, just channeled enough to seemingly be meritocracies. I’ve been in more than one discussion where the technically inferior path was chosen, because it was promoted with the loudest chest thumping by the most vocal team member.

    I’m not saying that last is true of all such communities, nor of the Linux kernel community in particular, but that these self-proclaimed meritocracies at the very least teeter closely to such shapes, because of the communications methods chosen.

    All that said, I’ve also met many, many FOSS developers who were respectful of any attempted contribution, even when they rejected the attempts. Kudos to all of you like that!

    1. Your point about meritocracies is important, and it goes beyond tech.

      Many places that’re dominated by men, and accept aggressive and rude behaviour claim that they’re just a meritocracy, and apparently women aren’t as interested in being involved or as good at it. Like politics – watching Question Time in the Australian Senate is like seeing a primary school playground when the teachers are out, except the budget is billions. Their argument is actually true in that by and large women aren’t as good at tolerating and succeeding in toxic, hostile environments, having not been socialised to it as much during upbringing… but that doesn’t make it a defence of the culture. Rather, an inditement. Their measure of merit isn’t skill at the job, it’s skill at being horrible and tolerating others behaving the same way.

      LKML is a meritocracy, but their measure of merit isn’t purely technical, and the non-technical inputs are pretty nasty.

    2. You’re absolutely right, +unwesen! In most organisations it’s not competence which gives you authority, but guanxi, muscles, and testosterone. But just because the Linux project is dominated by men is no excuse for sexism and mobbing. The CTAN (LaTeX) I have sent a couple of patches and even a complete macro in is an oligarchy dominated by men over 50. They can be quite authoritarian, but they normally won’t start with personal attacks.

      Now, as you left them, they will probably never realise that what they had done to you was wrong. So what: Simply forget about them and find another exciting project. There is so much OSS you can work on. For myself, I’ve just started contributing to FreeVMS.

      Shit happens.

  21. Well, this sucks. Not that you are leaving, although obviously that’s disappointing, but that you are put into a situation where you have to make a horrible decision like this. I fear you’re right on all counts – the situation is not acceptable, leaving is the right thing to do for your own health and well-being, and the community is not likely to change soon. Please do not feel guilty for taking necessary steps to protect yourself – nobody is required to allow themselves to be victimized.I have daughters, and I very much hope that by the time they are old enough to notice, things will have changed for the better… But I am not counting on it. I wish I had something more constructive to say other than thank you for sharing, and on behalf of all non-jerkwad men out there, I apologize for the bullying idiots. We’re trying to figure out how to civilize them without becoming them. (And you didn’t say it explicitly, but it was men who were being jerks, I guarantee it…).

    I wish you happiness, peace and success wherever you choose to go. Be healthy.

    1. I don’t know about LKML specifically, but in my experience it’s rarely just one person. At the very least, other people have to let it happen.

  22. Sarah, thank you for your work. It’s been several years since I had to spend a single thought on USB in Linux. It is the examplar of “just works.” Sorry to see you go (but cheers to you).

  23. As an anonymous bystander, I’d like to say that I appreciate the work you’ve done and the things you’ve started.
    I hope you’ll be happier wherever you go.

    >I have the right to replace any comment I feel like with “fart fart fart fart”.
    Is this a joke I’m not getting? It seems so out of place.

  24. Thank you for your cultural support.

    I am one of those “newcomers” that performs kernel modifications for a private company, and I too have no desire to interact with the Linux culture; I feel it would require more effort to deal with personalities than to deal with technical issues.

    Please continue to fiercely defend your values wherever you are, as many voiceless people on both sides of the sex aisle need your flavor of leadership. A citizen of any community is required to defend the rights of other citizens, not disregard the concerns of their neighbors with crass.

    Fair winds and following seas where ever you go.

    1. Companies are starting to require participation in an OSS community – including the mainline Linux kernel – as part of job descriptions. As someone with decades of experience in operating systems, I can tell you that the 80s and 90s were nowhere near as toxic because most companies had their own version of a closed-source implementation, and companies in the main will not tolerate an overtly abusive environment. Some will, but most won’t.

      This is a major part of why women are not going into and are leaving computing. Companies who profit from using FOSS need to strongly advocate for change, or nothing will happen.

  25. It is rather sad, but am certain you have chosen the right path, in mainly for escaping the dread that changes that you suggest would trigger a knee jerk negative response. The community needs to be able to foster harmony and at least internally diplomatic approach. What I fear is that stress is as infectious as the invective, and one perpetuates the other. One thing I do strongly believe is that women bring something positive and calming to this male dominated environment, and with your departure, things may get worse rather than better, even for the males left.

  26. You are welcome to try FreeBSD. All projects have inmature people to some extent but we try very hard to keep the environment sane. We are more like a democracy; we have no dictators (benevolent or otherwise) and there is a code of conduct.

    Women, and particularly graphics saavy developers are actively welcome.

  27. Hola Sarah, antes que nada te mando un fuerte abrazo y mucho éxito en tus nuevos proyectos, ojalá que la comunidad mejore por el bien de todos.
    saludos y te mando un fuerte abrazo.

    Hi Sara,
    Hope than Kernel developer community changes for good, it’s no nice, see or feel like uncomfortable in any work and neither it’s the best choice for you and for all people, a place insane it’s no good for anybody.


    1. OpenBSD has a bad reputation (against anyone non technical, not specifically sexist), otherwise the BSDs seem pretty welcoming and safe,

  28. You are not at all wrong in saying this. I am 100% sure that the senior kernel developers perceive a need to be abrupt with people. Whether their perception is correct or not, I cannot say.

  29. I just came to watch the comment section for the “fart fart fart fart” substitution. Also, I plan to use this blog post in my tech comm class when we talk about gender and collaboration. Thanks.

  30. Thank you for your contributions to free and high-quality software. You are absolutely right that human dignity should be respected. As you put it, this is not about asking the government to silence anyone, but asking a club of sorts, one that relies on many, many volunteers, to be respectful to the people you work with. You’ve done your part. Enjoy yourself while waiting to see the hopefully inevitable cultural change. Videogames, computer science et al should not become a safe haven for homophobia and sexism–just because many a bullied person find refuge in “nerd” or “geek” culture does not make it okay for them to kick the abuse can down the road. Women and minorities and other sexualities are not the oppressors. Some developers need to check their privilege. Good luck finding positive communities to be a part of.

  31. Sarah,

    No one should be expected to deal with such hostilities, especially when contributing to open source.

    I can only hope that this helps incite change in the system and behaviors of those who have not only pushed away a great kernel maintainer, but many would-be contributors as well.

    I wish you well in your endeavors.

  32. I’m a Debian developer – and mobility impaired. In due time, is there ny chance at all that you might consider something that would help accessibility or the needs of folk who need asistive technology? At least then, you’d be assured that end users would benefit and thank you for it – and the number of core developers is smaller, many of whom are living with one or other condition themsellves.

    Things are getting better in a desperately unequal world but not nearly fast enough. Thank you so much for your contributions up until now.

  33. Thanks for this clear and reasoned statement. I think we are all living through a cultural shift as advanced computer expertise moves into the mainstream and develops more normal expectations of maturity and decency. I appreciate your efforts to move that process along, and I respect your decision to back off from the fray.

  34. Good for you, Sarah.

    People are probably going to argue about whether it’s good or bad for the kernel community to be acerbic, but they’re missing the point: it’s not good *for you*, therefore you left.

    Ignore any haters. 🙂

  35. I don’t know you. I was sent a link by a friend. I don’t know you, but I know your situation. I wish I didn’t. I wish this was an isolated event in a healthy and inclusive industry. It isn’t. I feel ashamed for my fellow men, sad for the lost opportunities and angry for the people who get abused in one way or another.

    I’m sorry. But thanks for trying. And thank you for not trying too hard. Too many have tried too hard. It’s not worth it. You are worth more. Love.

  36. I’m sorry to hear you’ve left after being treated badly. I’m not a kernel developer, but I’ll still try harder not to be personally brutal to people in the communities where I develop, help and use.

  37. Thank you for your courage to post this.
    Thank you for your courage to step down.
    Thank you for your work on the Linux kernel.
    All the best for your future ventures!

  38. Sarah,

    I work in a very male dominated organisation, I rarely see this behaviour but when I do I never accept it. It’s not how I work, although it may well be how others think they need to work in order to complete tasks.

    I feel, could be wrong here, that cultural change can only happen if demonstrated by the organisations leaderership.

    If leadership don’t actually want change then it won’t happen. Some times a change in leadership is actually what is needed.

    Best o’luck in your new adventures!!!

  39. Sarah,

    just a quick comment that I fully stand behind you. You did the right thing. The community loses an important contributer and they deserve it.

    Best to you!

    1. There’s been some discussion on that question at LWN:

      I don’t endorse those recommendations necessarily since I haven’t participated in the communities they list enough to form an opinion, but you may find the discussion helpful.

  40. Thanks for being. Being honest, being direct – being you.
    You’ll be an enrichment for these friendlier communities.

  41. In my honest opinion, all kinds of problems in the ‘community’ are widely overreacted onto by some individuals. Anyways, thanks for work and good luck.

  42. Hi

    That seems sad. I have no axe to grind in this battle, but I do understand what you say about wanting technical directness with personal gentleness.



  43. It takes a lot of strength to stand up and say what you did. I totally respect what you are doing. Take care of you. It’s what’s the most important. You don’t have to subject yourself to nastiness.

    Also, I have no problem if you replace my comments with FART FART FART FART. 😀

  44. That is a sad commentary on a project that is a backbone on the Internet. Ultimately it reduces the number of people willing to help out which compounds part of the problem. We are social beings and this is a group project. People need to learn to communicate in such a way as it facilitates progress not impedes it. Gay jokes/insults should be left behind after high school

    1. What “reduces the number of people willing to help out” is the bad, violent, sexist behaviour of wanna-be-alpha males. Not the speaking out.
      Gay jokes/insults should be left behind not after high school but at the same point where racist and antisemitic jokes/insults are left. Just OUT.
      No more sexism and violence!

  45. Thanks for the wonderful blog on the toxic nature of the linux community. I appreciate your honesty and personal sacrifice. Common decency and basic respect afforded a stranger should be the norm for everyones working environment NOT TO MENTION a volunteer environment. I hope your journey has many joyful moments in store for you! Regards,
    Jim Baker

  46. You should give FreeBSD development a shot. Our community is much more respectful than Linux ever thought about being, with a code of conduct that is actively enforced.

    But I totally understand not wanting to be part of such a toxic culture. I gave up contributing to Linux in the mid 90’s because even then the culture was just too nuts.


    1. Unless a male committer threatens a female committer. We all know how that goes. Don’t suggest unsafe places. FreeBSD core’s handling of that situation proved it is not a community that is welcoming to women.

  47. Your experiences are shared by many. Contributing back a fix or feature to open source projects, particularly the Linux kernel, is often an unpleasant, thankless, and sometimes personally degrading processes. It does not need to be that way. There are other communities in this world who maintain high standards while still maintaining a positive, supportive, and thankful community environment.

    With every interaction, each of us has the opportunity to make this better.

    Thank you for all you’ve contributed to to the USB stack and beyond. These are technically impressive, highly impactful contributions of code and ideas. I hope you take great pride in this body of work – we all have been lucky to benefit from your efforts. We wish you the best!

  48. Well sorry to here you are stepping away from the Cult of the Penguin but I totally understand the reasoning and feel quite sad that it had to happen the way it has. Really I just don’t get why people can’t just get along with each other especially when it comes to sharing ideas like code etc But for what it is worth you will be missed but I am also positive you will find somewhere else to be equally as successful.

    Don’t feel bad it is not your fault but it is the responsibility of everyone who advocates Free software to also have duty of care with peoples freedoms to agree and disagree.

    Be strong , be happy, be successful, but above all, be yourself.

    Donec rursus convenerimus

  49. I’m sorry to see another developer driven away from the kernel. I also thought the Code of Conflict was useless. Thanks for all the work you’ve done.

  50. Didn’t there used to be this thing called “professionalism” that used to mean that you treat your colleagues with respect? I knew that responses to the mailing list were usually a little brusque, but I had no idea.

    Thank you so much for coming forward. You are a true pioneer, and you have shown much courage.

  51. Many thanks Shara for your post that very well confirm my own experience with too many maintainers. I wish you to find a project where you can enjoy to share your very high competencies in a pleasant and respectful way.

  52. I am not surprised. This has happened in the past and this will continue in the future, unfortunately.

    Good luck for your future endeavors.

  53. Dear Ms. Sharp,

    I’m sorry that you decided to leave the “community” of Linux kernel developers, as it’s a loss for the community. Having said that, I think you made the right decision.

    Respect for a person is a human right. Mr. Torvalds is abusing his power by ignoring that.

    I wish you happiness and success in the future

    Yours truly,
    Tom Haberkern

  54. Having a daughter myself, I feel saddened that this is something she has to deal with simply because the majority of people in the existing system aren’t willing to objectively evaluate their behavior and that of their peers and realize the damage they are doing to the community.

    We have made considerable strides already in acknowledging the value that women have to offer in every field. I simply wish we didn’t have so many more strides left to go to get where we need to be.

  55. Thank you for work Sarah, both as a developer and an activist. I wish you the best of luck in whatever you choose to pursue and hope to see you back someday. Take care and don’t give up, entropy is part of our nature 🙂

  56. Hi Sarah

    Very sad that things have gone this way, this is how a community loses good people. You’ve done such a lot. Awesome.
    Thanks &

    Some of the comments have suggested “alpha male” behaviour as a root cause of this, but unless we’re talking about testosterone-induced aggression I really don’t think that holds up.
    I have also seen similar behaviour in local user groups.

    It is just inexcusable, completely rude and people should just get a grip on their behaviour. While it generally doesn’t have malicious intent, it is offensive and seriously off-putting particularly to newcomers but also just in general – when actively engaging with a community, one naturally want that community to have a baseline of friendly & comfortable.

    I think the problem can more accurately be summed up in terms of social skills. Many people haven’t had brilliant examples from home, and then depending on the groups they interact with there’s more experiences and examples that are far from ideal including from some “top people” in the field (and I’d include Linus in there).

    On the plus side, these are very smart people, so once they acknowledge the issue (and see it for what it is, a serious issue that needs addressing in themselves) they should be able to work on resolving it. It may take some personal work.
    It’s not cool. It’s a bug.

    Regards from Oz,

  57. Thank you.

    I’m old and retired and have worked in many environments. What you have written is correct. All the work of the Linux kernel could be done with the utmost civility if people would treat others as they would want to be treated rather than icons on a screen or buttons to be pushed.

    Linus and others are wrong to treat abuse as emphasis. It’s not. It’s bullying, inflammatory hot air, and has no place in the kernel. Treating a stranger like that on the street can get a person charged with assault or held for psychiatric observation. It’s not normal behaviour in modern society. A teacher, priest or lawyer who acted that way regularly would be suspended or fired rather quickly. Computer programmers/developers are supposed to be rational. If they need to vent base emotions they should hug someone or go for a walk not scream into the ether. Linus and others at Linux have the simple option of explaining their displeasure and refusing/ignoring requests. They don’t have to start a war of words. LKML and the Linux code is an enormous and important body of knowledge. The authours and keepers of that knowledge should be pleased with what they have done and strive to do more. Verbal abuse is out of place and contributes nothing to the task.

    1. “if people would treat others as they would want to be treated” – They do. The problem is most normal people don’t handle how they treat each other very well, but kernel developers are not normal people. Comes with the territory. You’re essentially dealing with people who think they’re better than everyone else, and many times they are.

      “Computer programmers/developers are supposed to be rational.” – They are, until they aren’t, and it changes very quickly. People are irrational and they don’t deal with that. The problem is what is irrational to one person may be rational to another. Having a differing opinion on a matter could result in an exaggerated response. You also have the issue of stress. You may be the straw the broke the camel’s back or you may be an easy target to vent.

      If you look at most of these outbursts, it’s because of repeated mistakes, but not all the time, and I do not assume anything about Sarah’s situation. Senior people get a short fuse when they’re constantly bombarded with people making the same mistakes over and over and over and never learning. Many times the mistakes are a “let me google that for you”, just read the FAQ and stop making these mistakes.

      There really should have been a code of conduct from the very beginning. The issue, as it stands, is the code of conduct can’t do much of anything any more because many of the offenders are corner stones of the entire community. what do you plan to do if they offend someone again? Kick them out? Ohh sorry, but Linux can no longer move forward because 80% of the core developers were kicked out, then they forked the kernel and started their own clique again.

      One of the issue with all of humanity is very talented people tend to already have fringe personalities, but not to say all jerks are talented.

      I’ve heard my share of horror stories of very talented people having to “deal” with other people. Rarely ends well.

      At work, us programmers never want to talk to a customer. Our most senior programmer had to talk to a customer once, they threatened to sue us by the time he was done. I had to talk to a customer once, almost said a few choice words, but I was able to find someone else in time to take over before I hit my limit.

      I love helping people, I love knowing what I do helps people. I enjoy knowing what I do makes a positive difference in the world, but don’t ever let me talk to people, it doesn’t go over very well. I have very strong opinions and if I feel that I can’t get through to you, it will quickly devolve, or I will ignore you. Mind you, I do well with other programmers that are either my peers or juniors. My peers and I rarely disagree and I can be easily convinced with good reasoning. My juniors colleagues are well meaning and listen well, plus I’m paid to work with them.

      Without knowing any background of what exactly went on, I have an immediate feeling of sympathy for both sides. Sarah is a talented person, but seems to do more with outreaches and management. She is a people person. Many of these kernel devs are not people persons.

      I wish her the best and thank her for her work. fart fart fart +- 1 fart

  58. This is highly regrettable and I’m sorry to hear that you’ve left kernel development. I do concur that the tone of discourse on LKML could do with being… sterilized, if not toned down a bit. Maintaining a culture within LKML is one thing, but it also requires maintaining professionalism. Something that I hope people start to address sooner rather than later to detoxify the image that could prevent people from coming on board.

    Best of luck to you on your future endeavours!

  59. You wrote: “I need people to correct my behavior when I’m doing something wrong (either technically or socially) without tearing me down as a person.”

    I don’t agree with your phrasing “correct (your) behavior”. If somebody thinks you’re wrong on a technical matter then the result should be a professional and respectful discussion to understand the problem and how to best it can be address. Frankly, the way you’ve engaged, and tried to get other people to engage, has been to the highest professional standard. *Your* behavior does not need to be corrected; it should be lauded.

  60. There is definitely no need for guilt and I hope you’ve found happyness in your new endevours.

    As someone outside the kernel community, but an avid FLOSS developer and I have often watched with a sort of akward gawk of someone watching reality TV at the brutal emails on LKML. Like recent network TV trends, the worst of it would bubble up and stick at the top of the “Hottest messages” leaving useful and interesting patches, announcements, and critiques hidden away never to be seen.

    When I saw this pop up on twitter, I knew what I was going to find on the other side and I wasn’t surprised. Mostly because the community I’m closely attached to has seen toxic members of the community and weak governance lead to similar departures of talented and valued members. Friends.

    Redirecting or weeding out toxic behavior from otherwise amazing communities without destroying what makes them great is daunting to say the least. I believe we all share the hope that the kernel community can work through and address their own toxic members so more people like you can be the valued members of the community they deserve to be.

  61. Thanks Sarah for everything. I did get a chance to interact with you while applying for OPW and have seen your dedication towards the Linux community.

    This has come to me as a surprise. Personally, I saw you and Greg as the people who were changing the community, as the mentors for numerous kernel newbies like me who wanted to contribute to Linux kernel. You were doing a marvelous job.

    You stood for something and we supported it. This kernel community is not owned by these maintainers or its founder or whosoever. It comprises of all of us who love you for what you do and support you.

    I will like to thank you on behalf of the community. Thanks Sarah for all you did. Best of Luck for all your future endeavors.

  62. Just wanted to add to the chorus of thanks for your work. I’m sorry you had to endure that toxic environment. It’s not one that I could bear myself. 🙁

  63. Sarah, thank you so much for all the though you put into this, it shows.

    Thank you for being an inspiration to me in the greater open source world out there. Running into you at conferences, seeing you speak and interact with other developers and ask for personal respect has given me a great model to work from. You are one of a handful of people who inspired me to go back for the masters CS.

    When I was 17, I quit wanting to be an engineer. There was no emotional maturity in my compatriots, made for long days. Now I’m 36 and look, I’m an engineer. A lot of the culture shift that made me want to come back to engineering is happening because you and others like you stuck it out when yes, it is an emotionally thankless task. Thank you for the care you put into every moment you worked in this community.

    I hope you are enjoying the projects you are currently working on. I know I am. 🙂

  64. Hi Sarah,

    I remember back in 2013 when you made a post about this (which I think was probably something that weighed heavily in your decision).

    Back then, I disagreed with you about the whole thing. I’ve since changed my mind after reading more of what goes on on LKML and having work and life experiences to compare it to.

    I still don’t know what the right balance is between “full on politeness” and the more “liberal” approach, but it’s pretty clear to me now that there is more to situations such as yours than what I realized back then.

    Regardless of how much we agree or disagree on the specifics, I wanted to let you know that I’m sad to see you go.

    I wish you well.

    1. I think that the balance is when ad hominem comments, metaphorical stereotyping, or intentional neglect, from leaders themselves or members who are allowed to continue to behave this way, results in turning off or turning away a significant percentage of community members who exhibit enthusiasm, make excellent contributions, and who support and value others with enthusiasm and excellent contributions.

      I worked with Sarah a small bit through a college project when she was a student. I feel her work is excellent. The end of her continued participation in the Linux kernel is a big loss to that community, in technical excellence as well as leadership. The saga that her earnest efforts regarding collegial collaboration are unwelcome there, is to me a dark smudge on that community. It implies that the Linux kernel community will not be able to bring on board all the best and brightest individuals who have the most to offer. Maybe it’s more accurate for me to say, it won’t be able to bring on and retain all of the sort of contributors *I* would enjoy seeing there.

      I can only imagine that shifting focus away from something which captured her enthusiasm and encouraged her to excel, grow as a professional, and be her very best, is accompanied with tremendous sense of personal loss. The decision to cut one’s losses and move on with life isn’t the end, though. In my humble opinion, Sarah hasn’t, because her integrity shows through, her contributions were made without regrets, and they eloquently speak for themselves.

      Not all communities will welcome the incisive, plain speaking, respectful, socially conscious awareness raising Sarah brings, along with the technical competence and community leadership skills. Some communities just don’t want to hear it. However, the community which is compatible with her focused attention will gain greatly, and it’s my hope Sarah will find much worth in her days.

  65. I’m really sorry to hear this.
    I’ve never interacted with the kernel in any way so I didn’t know it was this bad. I knew there was a lot of frankness involved, but this sucks :/
    Thank you for your efforts and work

    I hope it gets better soon

  66. You’d be welcome in our illumos and smartos communities. Although there is no explicit code of conduct that I know of, we keep it civil almost to a fault because the focus is so intense on the technical changes in the code.

    Even the most heated discussions revolve strictly over code. Most of us there are professional engineers with several decades in the industry. intel corporation also used to have (and perhaps still has?) a team of engineers dedicated to the Solaris kernel, who used to work on OpenSolaris before Oracle closed the code and it was forked into illumos.

    Hope to be able to welcome you to our communities soon.

  67. Hi

    I hoped that you in the group meant that the group had made a turn for the better. That they had realised that they need to shape up the style of communication. I am sad to hear they have not.

    It is a loss for us all as a whole that they fail to adopt new people into their work.

    I absolutely think you made the healthier choice. And I will always look forward to see what you are up to, and I hope to bump into you again sometime soon. (perhaps time to take me up on that invitation to come visit Bergen, Norway?)

  68. Having worked with a kernel developer on a totally different project altogether and having had his behaviour excused with “that’s what he’s used to from his Linux endeavours” I can totally understand your decision. I, too, hope that the culture will change to make open source welcoming for everybody.

  69. Sarah, thank you for sharing this and I also want to voice a support for your decision. I have been part of different open source communities (and I started some) and it’s a good decision to leave when the environment becomes toxic and you feel powerless about it.

  70. First: Bravo.

    Second: I think (IMHO) the mistake you made was going out gracefully, with all your succession plans handed over properly, etc. Because — in doing so — you never force the the community to examine the gaping hole you’re leaving, because you filled it before you left.

    And I understand why professionalism — the respect you promote — compels you to do so. And I applaud you for that.

    But because you have that professionalism, and respect, there is no pain-point for them. There is no immediate “why are we having to scramble to cover all this, what did we do to deserve this?” moment.

  71. I hate to see someone with your skills leave but totally understand, respect and kindness should be the norm. You are welcome to join us over at the Linux Mint Community, but wherever you go, I wish you the very best.

  72. fart fart fart fart

    I’m so sorry about what you’ve experienced. Thank you for all your brilliant contributions, and best of luck in the future! ????

  73. For an interesting parallel see David Boies deposition of Bill Gates.

    Gates tried to run through the deposition with the sort of behavior that was acceptable at Microsoft, and delivered a twisted performance.

    I think Gates looked at his deposition and got the message that no one else was able to send him.

    Gates was smart enough to at least change his public persona (I would have no idea about the real Gates).

  74. Well written. Very constructive and positive, love the way that you handled your departure. What matters less is how you left things, what matters more is how you handled yourself.

    I’m new to this story by way of Slashdot. One thought to all of the critics. Diversity has real meaning and value. Diversity means people have different perspectives and needs. Diversity means that one size does not fit all, and that different opinions should be heard, respected, and considered precisely when they challenge fundamental paradigms. Being open minded means grappling with critiques such as these in a way that seeks to find consensus. So much of the open source community has great values, from the GPL to the Debian Social contract, but a great foundation requires rigorous evolution and development.

  75. Wise words from a person making a difference just by making her voice heard on this issue. Thank you for being brave, which in turn helps us all say the things that need to be said. We must stand up for human respect and communications that reflect that.

  76. as a Linux user/admin I’m always very sad if I hear these things about the kernel development. Thank you for your endurance to make GNU/Linux happen happen.

    And after reading that, I would simply LOVE to sometime read about Sarah Sharp making HURD happen, because Linux Community was to stupid to be nice 😉

  77. This is why I stopped talking to the OpenBSD developer community and even stopped running OpenBSD, it’s simply not worth the hassle of dealing with the OpenBSD developer community. No wonder at the end of their LibreSSL slides they ask for money to sponsor developers, I doubt very many else people want to deal with them.

  78. Thanks for all your work! Free Software certainly owes you much.

    Your reasons for closing this door are more than understandable. You don’t have to put up with that. Nor should anyone. I hope more and more people start saying “enough”, like you did, and that really starts a change in this community.

    In the meantime, I hope you have a lot of fun in whatever you do now 😉


  79. I’m really sorry to read this. I don’t know if you remember me, but we met at a women in tech conference, just after I’d taken a job in a platform team working with the linux kernel, after a long history with other kernels. You were very encouraging, and I certainly needed that encouragement – LKML is very intimidating.

    I ultimately went back to FreeBSD. The community there isn’t perfect – in fact, it appears to have the usual flaws of large established volunteer projects – but I’ll take McKusick et al. over Torvalds any day of the week, in terms of the kind of interpersonal norms they favour.

    Meanwhile, you deserve two medals – one for your technical contributions, and another for your many interpersonal contributions. Good luck with whatever you do next.

  80. This is very unfortunate. It is a shame you are leaving the community, and the shame is on the community. I am one more person who would gladly help with a patch now and again, but I am completely unwilling to put up with the style (not substance) of the criticism that I see flying around various mailing lists. I patch code, and keep it in house.

    It is impossible to measure the scale of contribution that is being lost, but I am certain it outweighs any possible gains imagined by the men-children who are the mainstays of the kernel contribution community. I admire and respect the technical achievements, but I wonder at the support for abusive behaviour.

  81. Add another vote for sad but not surprised.

    The problems of the kernel community (beyond those of tech in general) are well-known, and I don’t see them changing any time soon.

    1. Problems of developers groups in particular, but don’t be that pessimistic, it takes time but it’s slowly moving in the right direction.
      Thank you Sarah, you did more than your share to help that move!
      Shame on the guys who pretend that respect is bullshit and verbal abuse is in their culture. What a culture!

  82. Hello,

    People are often unreasonable, irrational, and self-centered. Forgive them anyway.

    If you are kind, people may accuse you of selfish, ulterior motives. Be kind anyway.

    If you are successful, you will win some unfaithful friends and some genuine enemies. Succeed anyway.

    If you are honest and sincere people may deceive you. Be honest and sincere anyway.

    What you spend years creating, others could destroy overnight. Create anyway.

    If you find serenity and happiness, some may be jealous. Be happy anyway.

    The good you do today will often be forgotten. Do good anyway.

    Give the best you have…and it will never be enough. Give your best anyway.

    In the final analysis, it is between you and God. It was never between you and them anyway.

    — from Mother Teresa’s wall (adapted from Kent Keith)

  83. You might want to take a look at Nonviolent Communication, which was developed by Marshall Rosenberg. It has some very useful ideas about how to deal with disagreements in a productive manner. Satya Nadella uses it, so it has some geek credentials.

  84. Thank you for posting. Everyone that has a community should understand the importance of their community’s culture and what it rewards and penalizes.

    The mode of conflict resolution sounds familiar. It’s not surprising you grew tired of the effort involved in trying to change the culture of the group. The time and effort it takes to changes people’s perceptions and subsequently their actions, is often more than a single person can shoulder. It’s not your fault. You shouldn’t feel guilty. It is admirable that you gave it a sincere effort.

    The community lost a capable member because they were unable/unwilling to change the rules to support people that are different from them. This is not an isolated incident, and isn’t unique to a social group. Often cultures tend to recruit and reward those that assimilate and adopt the doctrines of the group, even if they are unspoken. Once established, it stays stable because people who don’t share all of the doctrines, yet are capable members, have to self-select and exit due to the higher burden of membership.

    My hope is that more communities learn from this and do as much as they can to accommodate a more diverse group of contributors. However, if there truly is an advantage to that diversity, like a larger pool of competent members and an ability to have conversations the other group can’t, it is just a matter of time. Building and contributing to communities that we wish existed is probably a better, more constructive use of your effort anyway.

  85. As so many other people, I’m very sad to hear that the state of development for such prestigious project has devolved to were you personally feel you can’t deal with it. I also understand that there are always two sides to a story. Nonetheless it must have been pretty bad for you to feel that there was no other option, especially as you must have invested a lot of personal time.

    We all appreciate you work for the community and hope you might be able to return when the environment has matured more. Two last word: Thank you 🙂

  86. Congratulations Sarah! It is a tough decision to leave. I think you’ll find that you’ve made the right choice, even if it doesn’t feel like it right away. Great job!

  87. I was sorry to stumble across this announcement. It’s saddens me that the Linux kernel development culture continues to be so brutal, backwards and flat out negative. As an IT professional with 20+ years, I have never had to work in an environment where it was okay to yell or insult people. It’s totally unacceptable in these modern times.

    I commend you for sticking it out for so long and thank you for your contributions over the years and I wish you the best going forward.

    Don’t change!! 🙂

  88. Keep up the good work. I hope your corporate culture is quieter and more respectful than the terminal juveniles in the OSS community with whom you may have to interface from time to time. But I don’t see how that couldn’t be the case, in this anarchic, no-consequences-except-for-the-thin-skinned environment that exists on LKML.

    Code review shouldn’t be an excuse for personal denigration; we all make mistakes.

  89. Dear Sarah,
    You do not know me personally, by the way, thousands of miles separate us, but his story touched me deeply.
    I am a paralegal and law school student in graduate pathways, and a Linux user and enthusiast for Two decades, in a remote city in Brazil, Maceió!
    If today I write you these poor lines in the notebook that usually work and study, I know it’s partly thanks to its contributions to the kernel, stable, fast and flexible, which allows me to make use of the tools necessary for the practice of law, and, so I want to deeply and sincerely thank him from the bottom of my soul!
    I know that his departure is an immeasurable loss for the entire Linux community, but personally I understand his reasons and I identify with what you spent!
    I sincerely hope that time heals all wounds that have been inflicted by those who do not value their work or not his person!
    Good luck in your new endeavor and once again thank you for in a roundabout way you have been so important in my personal and professional life.
    Benedito Carvalho

  90. Thank you for posting this. Based on experience with spiritual groups, I assume that most of the people in your position walk away quietly. Those who are willing to respectfully state what they see happening help everyone who is willing to learn the lessons that need to be learned.

    1. I have done both and just walking away was easier, but left more residue in me and provided less learning.

  91. Respect from Brazil. Sorry for exist idiots into this planet , sometimes I think that the most important attitude is to hug the person and say hello, you are not alone. These people did not understand the value of your feelings. I do and I hope you continue bright wherever you go. Sorry for my english!

  92. its so Bad news that did not expect, even in dreams

    anyway Thanks to the great your contributions to the Linux community
    I hope that you return
    we a waiting for you great achievements

  93. Good for you for standing up and pushing back!

    Nobody needs to deal with this kind of behaviour, not under any circumstance. Common decency and respect for peers is sadly lacking in many communities and it won’t get better if people do not stand up against it. You did the right thing by pointing out the problem and refusing to simply go with the flow.

    Well done on being open and honest and best wishes with your future projects.

  94. Respect, Sarah. You can be proud of your contributions to the community, and you can be proud of your decision to shut the door.

    I’m not a Linux contributor but it did not took me too much time before I realized the amount of social misbehaviors that spread the community.

    It’s easy to use words to kill someone, and this seems to be a tradition inside this community. And it’s even easier to do that when you sit down your ass in front of a screen and a keyboard, rather that physically in front of the person you’re talking to.

    Of course, I bet that less than 5% of toxic people are haunting the community. But they are responsible for 95% of insatisfaction (but thanks to the remaining 95% of people who remain honest).

    While this project (Linux) could be a really attracting, amazing and awesome project, it turns out to be an internal war field, with “dominant males” fighting for supremacy. It seems that key people have reproduced here all the management mediocrity that can be found in some private companies.

    They should be ashamed ; just because their behavior makes talented people like you leave the community.

    One thing they should be aware of has been written by a french humanist of the XVIth century, Michel Montaigne : “On the highest throne of the world, you are only sat down on your ass”.

    Well, my vision is a little bit dark, but hopefully there are some counter examples, for example the Adobe Brackets community, where each newcomer is treated with respect, patience and pedagogy. It’s a pleasure to work in such environments.

    As a conclusion, I would like to say : don’t be sad with your decision : this was the only realistic one. Developing software should remain fun for us, not a burden.


  95. FART fart FaRt fArT

    that is art

    On a more serious note: You did what you had to, even if it doesn’t change a thing about the whole issue you did what gave you (hopefully) comfort and closure.

    Well like I said, even if you didn’t change a thing it’s just good that you didn’t decide to either play their game or turn an eye on it. In the end feeling good in life is what matters.

    Despite comedic relief I mean the best for you. I go back farting in my closet now. Someone grab a HEPA filter or something.

  96. Sarah, we’ve only met once, I think, during and after your ‘GardenGeek’ presentation, but I’ve been following the public side of your efforts to shift the LKML culture toward greater respect and professionalism with great interest. You should at least have the satisfaction of knowing that those efforts have *not* gone unnoticed by other technical communities, nor the lack of meangful response to your very reasonable requests.

    Simply as an extreme example of what *not* to do and how *not* to react to constructive criticism, the LKML has prompted many other existing communities to pull their act together, and the very visible long-term effects of not dealing with these issues at the outset has prompted many new communities to adopt codes of conduct and explicit standards for behavior at the point of their formation.

    So merely by being a reasoned and determined voice on the deficiencies you’ve noted in the Kernel community you have had a positive impact on the F/LOSS community as a whole, and that is no small accomplishment.

    There are many projects I can see that would greatly benefit from your skilled contributions (both technical and non-), and I for one with all be watching with great interest where you choose to direct your efforts and attention in the future (as one possible suggestion, the GPU-accelerated nature of most current AI and ML libraries could use much more hardware vendor diversity).

    Best of luck, and thank so much you for all you have done to make so many things ‘just work’, despite the ongoing personal cost, and know that your best efforts to change the culture you were embedded in for the better were far more than anyone could have asked for, and succeeded in unlikely ways far from the community you were actually trying to change.


    Michael Bernstein

  97. Kudos.
    Some of your complaints could have been directed at me in my early career.
    I have grown as a person and have hopefully matured past most of these behaviors. I would expect(request) anyone that I’m in contact with in any way, to point out such behavior to me, as I agree they are unacceptable.

  98. God damn it. Another fine woman leaves the field because men just can’t seem to behave themselves. I have seen the kind of behavior that you are talking about, Sarah. This is no way to run a world, people. I expect better- for all of us, and for my daughter, for all girls growing up now and in the future, and for all of humanity for crying out loud.

  99. That stinks to hear.

    Crap like this usually starts at the top. I’m sure it doesn’t help that Mr Torvalds himself is a brat prone to temper tantrums whenever anything doesn’t go exactly his way.

    You can have all the technical skills in the world, but you are useless if they are not accompanied by good interpersonal skills, and an ability to work well and politely with other people.

    It has long been my belief that this aspect of the linux development community has held it back, and it would be better off without these people, including Mr Torvalds himself.


  100. well, I guess the bucket goes to the well until the bottom drops out. :-7
    Hopefully you left before taking lasting harm.

    If you ever feel like doing some kernel work again: FreeBSD and Illumos have been mentioned; another option would be “my” NetBSD, which is good for the occasional squee. 🙂

    The mailing lists are usually public and have archives (for the NetBSD kernel list: Sampling these for a good fit in style should reduce both scarring, and frustration about the length of the decision making process.

  101. Let this be a huge wake up call for all people working in tech. Silence (or shouting in some cases) should not be the victor here. We just lost an important voice in the community. I beg you all to not be a bystander or observer, or participant for that matter. If you see behavior that is inappropriate, speak up. Help create a better tech world.

  102. Sarah
    I have been myself in a similar position, when quitting a ‘good’ job, in the agribusiness area. It is not the same as you described, but very similar. It has been 8 years now since I quit. I have found other great opportunities on the way, and the more I try, the more I see I made the right choice. I have become an independent worker, and recently wrote a whole master thesis analysing the company I left 7 years ago. It was a great feeling to be able to share it, and to put those bad memories into that work. I am sure the technology organizations still have much to improve, and by quitting and saying why, you are doing your part. Good luck in the future! I am sure it will be great 🙂

  103. Good on ya. I don’t know you, but I can tell you that those who tear others down to build themselves up, well, they build on shifting sands. I hope that removing this burden from yourself gives you the freedom to pursue something more fulfilling.

  104. Pingback: Network Online
  105. As a maintainer of an open source project I fully agree. A good communication style paired with code code is a long-term motivation for people being interested in the project. Not just among developers but also between developers and users. As a developer I’m just glad seeing the software working and well-documented and discussing with the community people about it. Affability is what counts in public discussions.

  106. Good luck in your future endeavors and I hope you find the environment you seek with work that fulfills your needs.

  107. I haven’t seen your technical work, being someone who is not a developer, but you have my personal respect.

    I hope you don’t feel guilty about your decision. I hope your feel supported. I hope you’re viewed as a role model and an example for others.

    The easy route was either to take their behavior and not speak up about it or to have the bad experience and quietly “walk” away. Instead it sounds like you respectfully brought the issue to their attention, brought it to the attention to a wider community, and when the problem wasn’t corrected or even acknowledged as being a problem, you respectfully and responsibly exited. At each step, you opened yourself to criticism and attack and didn’t back down. From my outsider perspective, it seemed like you were attacked for your position. To me, the behavior seems like bullying and you stood up to bullies for yourself and for others. Some of these bullies are the same people who may have themselves been bullied and should know better. In spite of all this, it seems like you still respect the work and still support the work and the technical abilities of the group.

    You deserve personal respect because you stood up for your beliefs, you stood up against bullies, are a minority in your field, and advocate for more inclusion in your field.

    I hope you’re finding an outpouring of support and wish you the best in your future work.

  108. I’ve no idea that this kind of things happened in the community. I’ve seen some senior maintainers cursing others, but I didn’t know that this goes so deep that they do so with their own community just to increase the productivity. For me, Linux has always been the idea of fun. Fun, where there are no boundaries or deadlines to worry about. You do what you do because you love it, and not because someone expects you to do so. I wish you the best in life, we’ll surely miss you.

  109. Unfortunately, I’ve seen this repeated throughout the industry, and the level of abuse seems to correlate with how technical the work is in nature. I think there are still some remnants of the time when it was unfashionable to be a geek, so the old “me against the world” mentality is still evident in some quarters.

    There are open source projects that actively encourage participation. No, really, there are! However, these are exceptions to the rule, and I turned my back on the industry, for a career in one which is founded on tolerance and acceptance. I hope that things change, and I hope you feel able to remain involved in the open source community.

  110. As a lot of people here I’m sad and most of it I’m shocked about this way of speaking. I love Linux, I work (love to help) freely for and when someone is going to the dark side of the force, everyone helps him to leave or to recenter. but bad attitude is a hard thing to manage. Sometimes leaving is not being scared, is a proof of wisdom.
    USB 3 won’t be so free anymore. have a good life wherever you contribute.

  111. Cheers from Finland. And Today is the day I’m also being sorry of being a Finnish Citizen, because of Mr. Torvalds. May the future be bright for you, and all the best!

  112. Sarah,

    I find it really sad. Not the fact that you’re leaving, because everyone has the right to chose his future. The fact that you *felt* you had to leave. To me it means that you’ve been seriously hit by this community and that’s sad.

    You may remember I’ve been one of those defending the Linus way of managing his project a few years ago. And I still do, to some extents. I think his method is suited to long-established people who understand all the emotional spectrum between “thank you” and “how did you manage to live”. I’d say we are lucky to have someone as transparent about his emotions as he is, it’s just that there are many cultures around applying different emotional mappings between what they feel and what they express, resulting in the reverse mapping to be done on what they read/hear/receive.

    The problem arises when these different cultures work together. It’s well known that in the US for example, people show over-enthousiasm. When my US boss tells me “great, amazing”, my translation is “OK why are you wasting my time telling me such unimportant things”. If I tell him “fine”, he thinks I’m in a bad mood while I’m happy. It’s well known that western europeans tend to be under-enthousiast and like colorful language to express what they think and have difficulties to express. If I’m saying to a coworker that he did a piece of crap, he will understand that he could have done better and will not feel bad for this. And in some eastern european countries, they tend to be more transparent and say good or bad as they feel it, which often appears very rude to western europeans themselves at first, though they can quickly adapt once they know it.

    Interestingly, Linus seems to have a mix of the last two, but is more on the last one. So he expresses his feelings and adds some colorful images to them to be sure everyone gets the picture. For people who already know his spectrum, that’s not a problem and it’s even a very accurate measure of where his satisfaction cursor is. For newcomers this can come as a shock. Especially when made in public because the person could feel humiliated by imagining that other people will laugh at her. In fact only newcomers or journalists will laugh at what he says, people who’ve been working with him for a long time will just read “hmm this guy made Linus really angry”.

    It’s really a matter of culture and mapping of emotions. I’m lucky enough to have my own emotional spectrum approximately aligned with his so that helps me understand his feelings when he rants about someone or something. He already told me things like “willy shut up you’re totally wrong and stupid” and demonstrated why I was wrong, and he was right. I didn’t take offense for this at all because I could have said the same to someone else who had drove me nuts. What people do not realize is that he expects others to be like this with him as well. I’m pretty sure of this, I’m complaining all the time that people around me don’t tell me when they believe I’m wrong and that’s a big frustration in a workplace.

    And I can understand why he says that by being more gentle and polite it will degrade the quality of the kernel. It’s because you’re asking him not to reveal his opinion. He needs to give his opinion on contributions. If some project leader asked you to just use “maybe” and “maybe not” as a response to any question, you would probably not feel at ease either.

    It’s important for the project’s progress that everyone respects each other’s work and time. And it’s true that having people do mistakes is never fun but is normal and expected because we’re all humans. Mistakes can often be corrected by the first one who notices it. Faults are different. They happen when people are lazy and try to save their time, resulting in wasting other people’s time. This situation is not acceptable and being very clear about this is important. I tend to consider the first fault a mistake (ie: “it’s not the right process here, you’ve been warned”) and next ones to be faults. Linus yells at faults, not mistakes. And sometimes, as he’s human, he will make the mistake of taking someone else’s mistake for a fault. That’s where the other ones must tell him “Linus shut up you’re a moron” so that he realizes that he hurts someone due to his own mistake.

    I’m having issues on other projects where instead of doing useful work I’m spending my time either chasing bugs that were introduced in code done in a hurry (easily noticeable in copy-pasted comments that become irrelevant), or repeatedly fixing basic coding style despite having spent a lot of time documenting how and why people should respect a style. Two thirds of my time spent on this project consist in fixing others repeated mistakes who ignore my explanations, appreciations and criticisms just because I’m still making the effort of being kind. In the end sometimes I feel like you, I’m wondering if my efforts are still worth on this project because others don’t respect my time. So you see it happens both ways. And in order to save some of my time, I’ll have to be a bit meaner, otherwise the project dies by lack of time for the maintainers in place.

    There’s a difficult balance to find here. Either demotivate established contributors or demotivate newcomers. In any case, we must keep in mind that humans are lazy, especially those working in IT, and that most contributions to a project will only contain the minimally acceptable amount of time leading to the work being accepted, with the remaining time being put on the other ones’ shoulders. There are exceptions to this but they are rare (and we’re lucky to have a number of them in the kernel).

    I hope you made this analysis before taking your decision. If not, I would kindly suggest that you revisit some of the conflicts that have affected you with the explanation above in mind to see if this was not just a problem of mapping as I explained. I’m not saying the problem should not be fixed, but that you could possibly be one key to help fix it and that this help would be welcome both for the project’s future and for the community’s health.

    In any case, thanks for all what you’ve done.

    1. Willy, I think you’ve hit upon the exact spot where I and most of the senior Linux kernel developers disagree. I believe you can be technically brutal without being personally brutal and still get your message through. In fact, most times, your explanation of the issues will be clearer, because you’ll focus on expressing what they did wrong, rather than your own emotions.

      As for your comments about the emotional mapping of Europeans to what they say, we will have to respectfully disagree. If you saying “I wish someone would kill you” is equivalent to feeling disappointment over someone’s skills as a maintainer, that mapping is just broken.

      What do you say when you’re past disappointment into anger at a larger broken system? Well, in Linus’ case, it seems that he slips into homophobic slurs. That means he thinks that being gay is worse than being dead. What kind of message does that send LGBTQ developers who want to get involved with your kernel community? (I almost said “our community” there but it’s no longer my community.)

      The most frustrating thing for me is that as a woman, I don’t get to participate in the same skewed emotional spectrum without harming myself professionally. I have had other kernel developers imply that I’m being “too emotional” and that I should “calm down” when I raise my voice even in the slightest. Women are socially trained to care about the community around them and other people’s feelings, and they get called nasty sexist slurs when they don’t have empathy.

      From reading articles and talking to other minorities, they also feel the awful double standard here. Black men and women get labeled as violent or deviant when they speak in anger. Or get shot by police if they attempt to assert their rights. If they express anger at a system that oppresses them, they get told to pay attention to white men’s feelings. They can’t win.

      When you say Europeans have a habit of exaggerating their emotions, to the point of tearing down other people, what minorities hear is “I have the privilege to not be able to care about other people’s emotions.”

      I would highly recommend checking out Scalzi’s post on privilege, “Straight White Male: The Lowest Difficulty Setting There Is”. It explains privilege with as gaming metaphor that I think most people can connect to.

      1. Hi Sarah,

        I think we more or less agree on most points internally but disagree on the form because of our different perceptions. In France we have some humorists who would probably be arrested in certain countries for the horrors they say. Most of the times their horrors don’t hurt people because these people know this is humor and take it with a second degree. It’s a matter of shaking each one’s culture to make people think before reacting. That’s an interesting exercise and people tend to become much more tolerant after such an experience. However people who don’t know them can sometimes feel hurt. When I read some of Linus’ posts you’ve pointed above, I find two completely opposite readings, one from a very angry man who suggests anyone to kill his ennemies, but that would be foolish from a smart man to ask for this in public, so I turn to the second interpretation which is the angry guy taking on himself and using colorful language to exagerate his feelings and make the situation a bit funnier to spread the heat. Linus almost never uses smileys, his emoticons are his words instead. He’s a sarcastic poet in some sort, and I very easily understand that he can hurt all people who don’t take it this way. His mistake probably is to mix this with his very authoritarian attitude. If a humorist would use his words he would have a lot of success. If a manager applied his authority, he would have a lot of success as well, but the two don’t mix well because they can too easily affect a lot of people who don’t know what parts he really thinks and what parts are just images.

        You know, I often find it fun to be a bit provocative to see how people react, because most reactions are only cultural or educational. But once in a while I see that people have doubts about a possible double meaning about what I said, or feel hurt and then I know I need to explain myself or to apologize if I unintentionally touched a sensitive cord. I have made some jokes about very close relatives to the point that other ones were ashamed of having laughed about that. The important rule is that the “victim” wasn’t there to hear or preferably was dead to be sure she would never ever risk to feel any pain about this. Yes that may sound awful to some people, which is why such words are restricted to the few people who are able to withstand them. And contrary to what you could think, I *do* care a lot about other people’s emotions, in part because I can’t stand seeing people in pain and that immediately affects me. And quite frankly, a joke which makes 100 persons laugh and one cry is a very bad one, just like the one which makes some people keep silent because they don’t want to cry in public. So I don’t feel like I have any privilege on anyone else, I just expect people to react about how *they* feel and not how they were *instructed* to pretend they feel. I’m often thinking “what are the worst ever things I’m possibly willing to hear about me and am I willing to hear these people say it ?” If the response is “no”, then some reserve is required. I try to avoid public shames as well for the same reason. And sometimes I fail, as anybody in any situation.

        However it’s important to continue to teach people how to distinguish humor, illustrations and attacks in order not to condemn free speech, just like they also need to be able to manipulate these with the least possible ambiguity, otherwise we’ll end up with humor totally forbidden because the basis of humor is always to make fun of someone in a given situation.

        So it’s not a matter of privilege, white vs black, men vs women, etc. While it’s true that some people may be more commonly aggressed and then obviously have a lower attack detection threshold, that doesn’t imply other ones must be in a privileged situation. Otherwise you’re creating a classification and that breaks any society, because just like you can always insert a real number between two others, you can always find differences between two groups of people whatever the group size.

        Maybe in the end you’ll find that I’m overly tolerant regarding speech in general and Linus’ in particular, but I don’t think I am given that I expect that if he acts like me, he’s willing to take the return with the same strength (you’ll see I’m not using the word “violence” here). Maybe he doesn’t realize how far some people take this emotionally because he has never felt the pain of seeing someone cry at his statements. I don’t know. All I can say is that *I* am not shocked by his words at others just like at myself, and am pretty sure I’m not the only one.

        Thanks for the links BTW, I’ll finish to read the last one later.

  113. Cheers from Brazil. That was a hard decision to take, and it is a shame that those people are disrespectful. I hope this gets better soon. Keep up the good work! Good luck and success!

  114. Gracias Sarah por tus contribuciones.
    Para mi Linux empezo siendo una afición, sobre todo formando parte de mi distro favorita Debian. Y todo ello gracias en gran medida a personas como tu que marcan la diferencia. No es solo trabajo y software, es una filosofía y forma de hacer las cosas basadas en una ética y moral que comparte una comunidad. A veces el camino que siguen los acontecimientos no es lo que esperamos, a menudo con desilusiones.
    En mi caso como simple usuario de Linux, sin un nivel como el tuyo, recuerdo con ilusión cuando buscaba si Linux podía soportar usb 3.0 incluso sin saber quien lo desarrollaba, porque tanto esfuerzo a veces pasa desapercibido.
    No eres la única que se siente mal. Después de años de esperar una mayor convergencia de Linux y sus distros comtemplo tristemente que hay todo lo contrario divergencias. Pero si algo he aprendido es que realmente no es como cerrar puertas ya que es como un rio que se abre a nuevos caminos, en tu caso sin duda encontraras uno bueno por el que seguir (solo hay que ver el gráfico cronológico de las distros para ver como hacen camino, unas se estancan, otras bifurcan y otras surgen desde casi cero).
    En mi caso dada la decepcion que me he llevado con Debian busco alternativas con pequeñas distros como puppy, quirky y espero si todo va bien y la comunidad devuan prospera y se mantiene fiel a unos principios dignos que sea una buena alternativa. Pero todas siguen necesitando de un sólido desarrollo del kernel y los que todavía confiamos en su funcionamiento y para los que queremos que siga siendo democrático la marcha de personas como tu nos da que pensar al respecto.
    Muchas gracias por ser asi, no cambies.
    Saludos / Best Regards

  115. You got my support.

    Regardless of our beliefs and as much as I believe in liberty, this kind of conduct should not occur. It really is all about integrity and nothing in this world can replace it.

    I hope that your workplace is as technically challenging as respectful and creative.


  116. I know you have done an amazing job implementing USB 3.0 support into the kernel and that awesome. I just want to say that as a person who does IT and is learning about Linux I want to know if the community and kernel development will keep going i’m sure it will. Also what are some other friendlier open source communities that use Linux as a fork for their projects? You don’t have to reply Sarah but anyone else feel free to do so thanks so much and keep on going don’t give up.

  117. I am sad that you *had to* leave, but given the circumstances I am glad that you leave. Nobody deserves to cope with such behavior, and quite frankly it is their loss.

    FWIW I think it is a very brave thing to do. It is such a wonderful thing all the effort you’ve put and the work that you’ve done, I’d just like to say “thanks!”. Trying to change things is great, but to keep suffering the behavior of emotionally-challenged people is out of the question. I feel it is the right decision and I can only respect you more for that.

  118. Sarah, I hate to see anyone get so frustrated with the culture of a project that they feel that it is time to leave.
    Do you think that the atmosphere that you describe is in no small part the nature of a project that started on Usenet?

    Your description of the kernel development reminded me of Usenet, on a good day. On a bad day, the place was thermonuclear.

  119. I don’t think we’ve ever met, but your work has inspired me. I’m sad to see you leave, but more than that, I want to express my gratitude. I wish I could say or do something that could change the environment. Respect is essential, and I am glad that you are taking care of yourself.

    Thank you for what you have contributed. You have shared gifts with the open source community beyond your technical expertise.You are a light!

  120. I really feel sorry for your decision, but you are a great professional and seram open new doors on your way.

  121. Really sad to hear that; although considering the tone in what I’ve read of the mailing list, I personally wouldn’t have bothered sticking around half as long as you did (even if I did have half your skills). Good luck with your new adventures!

  122. I’m sad that the Linux Community doesn’t work well.

    Thank you for your efforts and good look with the new doors.

  123. Sarah, I feel and have felt your pain. I spent 7 years active in Local Politics in the UK, and there is an enforced behaviour between politicians, with multiple faces. Like you said, nice people in ‘real life’ can be unpleasant and aggressive in a public arena. Differently to my experience within the software engineering profession, because each engineer is clearly skilled to some degree, there is a level of acceptance, which is only controlled by company culture. The Linux Community has evolved like the wild west, without that culture. It’s wrong, and belies the fact that humanity requires a degree of public behaviour, otherwise known as manners. Sorry to see you go, thanks for all your input. I hope in time, you’ll be able to come back, and see it’s grown up some.

  124. WTF does “people with different cultural and social norms” mean? This is a very honest, non-offensive question. I’m sorry, but respect is universal, I really hope this is not an euphemism for “non-American people”.

    1. You don’t have to be European to be an asshole. However, that being said, many kernel developers have attempted to excuse their behavior with “but I’m not an English speaker and people misunderstand my terseness for rudeness” and “I’m European, we’re all blunt, and you sensitive Americans and Asians should get over it” and “I’m British and you just don’t understand the purity of the English language and are deliberately misinterpreting my words.”

      Intent is magic. *shrug*

      1. I get it. I have to disagree, though.
        People DO have different cultural backgrounds and speech is an important part of that.
        I can definitely tell you that I’m an educated person and people have described me as a “polite girl”. That being said, in my country (Mexico) we really don’t care about political correctness, we don’t appreciate tone policing and we do swear a lot (in Spanish). I’ve been working with US teams for a couple of years and I can tell you, I’ve had to “censor” myself many times. As a polyglot person, I’m still my sweary Mexican self LOL, and I’m not mean. I understand that this can be tricky and people can abuse this resource, but most of the time it happens to be genuine. So I think we should ALL do our best to adjust to global cultures when working in distributed teams, not just make foreigners adjust to US standards 🙂

        1. There’s a difference between swearing and tearing the person down emotionally. A ‘shit’ or ‘fuck’ here or there really doesn’t bother me. What bothers me is the verbal abuse that is designed to silence criticism and cower people into agreement. Sometimes you just need to blow off steam about a bad situation, but do that privately, over beers with friends or with a counselor, not on a public mailing list. Then come back, explain the situation, and work things through calmly.

  125. People that are good in science are control freaks. This is how science works.
    Your interaction with the Linux community is an interaction with said scientists

    1. I’m confused by your statements. Are you implying that because someone is a control freak they lack empathy? I see scientists as curious folks who are willing to test hypotheses, and are able to change their mind in the face of evidence against their biases. I’m curious where our internal models of scientists differ.

  126. The person who wrote ‘given enough eyeballs, all bugs are shallow’ needs to consider what happens when people who report bugs or try to fix them get belittled for their efforts.

    I have been doing security for as long as Linus has been doing the kernel and I have been a senior member of the only project that can claim greater impact and scope – the World Wide Web. And the only reason I am mentioning that is that reading the IPv6 coding rant circulating right now there is absolutely no question that Linus is wrong.

    IPv6 was designed about 20 years ago and follows an approach that was popular at the time but we now know to be problematic. Specifically, if you use a length field to specify the end point of a list of variable length items that are also length delimited, this creates a situation and attacker can exploit.

    Assume that the list 11 { 3 { a} 4 { bc } } is well formed, there are ten objects in the list, including the braces. Now lets suppose an attacker sends 11 { 100000000000 {a} 4 {b c}}. In the right circumstances, the attacker can use that to create an attack in which the processor fails to notice that the inner element is too large to fit in the outer.

    In normal circumstances, this can be avoided with a simple check. But here is the clever bit. We create a list 11 { 100000000000 {a} 10000000000 {b c}} such that when the two values are added together, the result overflows an int16 and gives us the expected value (7).

    That is what allowed the heartbleed bug and it isn’t a very shallow bug at all, it is a very subtle vulnerability.

    Instead of taking the time to find out the reason for using the macro, Linus bullied the submitter into capitulation. And that is pretty easy when you are recognized as a ‘King of Geeks’. But winning the argument isn’t the same as being right.

  127. Sarah,
    I am not sure how many of those members were rude, control freak or disrespectful to you, but in the end you left and they stood there. I think after a point we make choices on how much somebody can hurt us. I

    I have faced scenarios like these and initially left or fought back and then gave up but then i realized that there is no such place where you may not find them. You or me always* don’t have a choice to select people we are going to work/debate/discuss or be a part of in a group, specially in today’s world where collaboration is the key. In the end the Linux community lost a great contributor and you.being a part of the great community.

    Sachin Gopal

    Sachin Gopal

  128. It takes amazing strength to be both this honest and vulnerable. I have a very deep respect for you… I just find it sad that your next step to driving change has had to be leaving.

  129. The “fart fart fart” thing is very apropos.

    You expect people to bend to your liking.

    That is narcissism.

    I read the entire LLKM thread that you started and you completely missed Linus’ points. I don’t know if it was because you are incapable of understanding opposing viewpoints or you simply lack reading comprehension. Even Linus’ “victim” in the breaking user-space flare up disagreed with you.

    You want fake behavior there are plenty of software shops that are totally PC but are a vicious den of passive-aggressiveness and everyone talking behind other backs. I have worked at places like that and was miserable although I suspect you would enjoy it.

    One less SJW in kernel development is a very good thing. Although, at least you contributed something in the field unlike most of your kind(ie the dongle-gate lady, et al), not that maintaining USB3 is demanding. It is up there with maintaining serial port kernel code.

    Anyway, good luck with your future endeavors, I hope you find the oppressive shop you are strongly desire.

    1. Now you have me curious about the number of patches I merged versus the number of patches I authored. Here’s some stats:

      sarah@xanatos:~/git/kernels/xhci$ echo “Year | # Patches authored | # Patches merged as maintainer”; for i in `seq 2009 2014`; do SIGNOFF=`git log –abbrev-commit –pretty=oneline –grep “Signed-off-by: Sarah Sharp” –since=”Jan 1 $i” –until=”Dec 31 $i” | wc -l`; AUTHORED=`git log –abbrev-commit –pretty=oneline –author=”Sarah Sharp” –since=”Jan 1 $i” –until=”Dec 31 $i” | wc -l`; echo $i ” | ” $AUTHORED ” | ” $(( $SIGNOFF-$AUTHORED )); done
      Year | # Patches authored | # Patches merged as maintainer
      2009 | 98 | 0
      2010 | 51 | 28
      2011 | 98 | 91
      2012 | 69 | 68
      2013 | 41 | 106
      2014 | 15 | 73

      Regardless, my opinions as a human being have worth, no matter how much I contribute. Expressing desire for basic human respect is not narcissism.

      I’m sorry you didn’t like your passive-aggressive co-workers. There needs to be a balance between tearing people apart and bottling up emotions to avoid conflict. Imagine being in an environment that you aren’t allowed to express feelings other than anger or frustration in random outbursts, where conflicts go unresolved for years because no one wants to talk to each other offline and work things out. That is the Linux kernel community. Avoiding conflict to keep things “PC” breeds resentment and backstabbing politics, but actively encouraging unnecessary conflict causes people to hide their emotions and issues until they have no choice but to blow off pressure. There are lots of healthy ways to constructively discuss issues and blow off steam that don’t involve tearing someone apart on public mailing lists. It’s unfortunate that people think there is no middle ground.

Comments are closed.