 Before I was living in New York City, basically was super unhappy, a shitty job, I was commuting an hour from New Jersey. I routinely fall asleep on my train on the way home, which was the last train out of the city because I had a shitty job and I routinely worked until eight or nine. It was just not fun and I spent most of my time actually kind of wishing I was back in college. The musical Avenue Q came out that year and there's a piece at the beginning, I wish I could go back to college and I broke down and cried when I went to see that. It was pretty ridiculous. Then I saw a job posting from a couple of people I barely knew looking for web developers in Kansas and before I knew it, I was moving to a state I never thought I would live in which I promise is not as bad as you think it is. The state may be. Lawrence, though, the town that I lived there is beautiful. This is South Park, the town in the middle of downtown in the spring. The red buds are just glorious. We have not one but three locally owned newspapers in the downtown area with cats. Yes, shop cats are the best. An amazing live music scene. This is the Flaming Lips playing in a venue about the size of this room. Really great local food and beer, an awesome college basketball team if you're into that sort of thing. The reason I moved there, an amazingly progressive and really unusual locally owned newspaper, in 2004 we had a team of about five developers working for this paper with about a 19,000 circulation. To put that in perspective, the New York Times digital team at that point was about 10 people. It was really weird and unusual to be doing really cutting edge web development in the middle of Kansas at a newspaper. So a year later we released Django, we took it out of the tools that we had been building at the newspaper and released it open source. We actually, like I went looking for our launch announcement blog and I realized we never actually ended up publishing it because while we were editing it people found the website and started downloading it and using it. So this is our first actual blog post saying, yeah there's a tutorial now, hi, yeah and it was great. We didn't think a lot about the organization of the project when we released it or we thought some about it but mostly we took our cues from the Python community and Adrian and I appointed ourselves benevolent dictators for life, a sort of tongue in cheek, you know it's a weird sort of dictatorship when your community can fork the kingdom and set themselves up as a dictatorship so it's a weird sort of term in open source but it's fairly common, a lot of projects have these sort of dictatorish roles and that seemed right, it seemed to be working, it seemed to be good. So let's cut forward about nine years and Adrian and I resigned as BDFLs and since then my involvement in open source and in Django has basically gone to nothing. My last commit to Django was maybe a year and a half ago and it was fixing a link in the README. I stepped down from my roles on the DSF, the Django Software Foundation, I barely write open source software. At this point my involvement in the open source community is basically coming to events like this and thinking about how I used to be involved in open source. So that's the story, that's the arc of my involvement in open source to date and many people know that story, what I haven't really spoken about or written about until recently is the why, why I went from considering open source to be an incredible important part of my life and my identity and why something that gave me so much has become something that I don't have the energy for anymore. And what it comes down to is there were three things I just couldn't deal with, three problems I couldn't handle and I give up and those problems are burnout, money and toxic people. This is probably a pretty good time to mention that the rest of this talk has some discussions of burnout and depression, harassment and toxic people within our community is nothing specific but I am going to talk about them and how they affected me and others in the community. Also, so look, I'm used to giving like two types of talks, I'm used to giving a talk where I'm trying to teach something and it doesn't have to be exciting or happy or fun but at the end I want you to walk out and know something you didn't know before. Or I'm used to giving sort of the like uplifting, inspiring, yeah let's go do the thing type of talk. So this is weird for me because all I have here are questions, I don't have answers and they're questions that really bum me out and depress me. If you follow me on Twitter you may have noticed in the last few weeks in the every evening I've been tweeting kind of bummed out stuff because what I've been doing is sitting down and working on this talk every evening and until I get too depressed to keep working on it and I set it aside and move on to other stuff. So yeah, I don't know why y'all ask me to speak first, I hope I don't, hey it says feelings right on the marquee, you know what you're in for, alright. So it's fairly obvious that what I'm describing here is burnout and I'm by no means the right person to really talk about burnout in general. I can point to a couple of people who have covered burnout in ways that I really spoke to me. Kathleen has given a talk a few times called Avoiding Burnout and Essentials of Open Source Care. There's a link to her video from PyCon, she's given it a few other times there. And then Julie, who's actually here today, talks about it in her talk, she's given a number of times, it's dangerous to go alone, the sort of second half of that talk as I remember is all about burnout and handling it. So definitely check them out, I did not take their advice and I probably should have, I wish I had had these resources a few years ago. So for me, the thing that helped me identify what I was feeling was burnout, was this sort of like burnout model from a couple of psychologists. Wikipedia has a surprisingly good summary of it and I kind of want to talk through like what this felt like for me going through it, like I said it wasn't until fairly late that I really identified and named what I was feeling. But in retrospect, sort of my burnout maps very well to these sort of stages of burnout. So it starts kind of with a compulsion to prove oneself, I mean for me, I felt like we had open source this thing and I wanted it to be really awesome and I needed to make this really great and the more I lost interest, the harder I worked to try to overcome it. Yeah, this isn't fun anymore, but if I just do it harder, I'll make it fun again. Working harder, neglecting needs, I stopped taking time off really or rather I would take time off from work so that I could work on open source, which was also kind of what I was working, it wasn't really, I stopped taking vacations. I remember in particular in Florida with my wife and her family sitting on the beach and reviewing pull requests. At a certain point, I stopped identifying, I don't think I ever identified the problem as my own, I started identifying it as other people's problems, if only these idiots would stop submitting bad code that I would have to review, if only these jerks would stop asking for help, like it wasn't my fault that I couldn't muster the energy to work on this. There was all these stupid people out there and I started making it about other people's problems. At a certain point, you kind of have to, I had to change what my value system was to accommodate my increasing disconnect from the work that I was doing and I somehow made it about like, if I can just do this more and make it better, this will like help me get a higher paying job and then I'll have more money and then I can retire and not work anymore. It's very weird, I can't even really summon why I thought that that was a good plan but it seemed really good at the time. All through it, I didn't really note that there was a slide, I just sort of felt like I was doing the only thing I could under the circumstances. So yeah, you start withdrawing, I stopped participating as much, I would say that I would do something and then just not do it. I would be around in the community but I didn't really acknowledge that I was stepping away. At a certain point, other people start to notice that you're different and that's getting more and more difficult and it was around here no longer seeing myself as valuable, feeling really just kind of like nothing matter anymore and finally entering what really can only be described as depression that I finally sort of noticed where I was. Luckily for me, I didn't reach what these psychologists described as the last stage of burnout which is an actual collapse and requiring medical attention but I think I stopped fairly close to that. So this is the email that I sent to the Django Core team, I haven't shared this before, I feel a little weird talking through it but it was important for me to name what was going on and I really just sort of ripped the band-aid off. I just said, I'm sorry, I know you're relying on me for things, I was even too exhausted to identify the things that I needed to hand off. I just knew that I just couldn't do it anymore and I said look like if there are things that I'm the only one knows how to do, let me know and I'll try to make you know how to do them but I need to step back and luckily I was surrounded by such a great group of people that everyone said please take care of yourself, what can I do to help, how can I, and there was no shaming in it, all the sort of worries that I had about telling my friends and fellow contributors about it was totally unfounded and I don't know why it took me so long. So I clearly got pretty burnt out and I still am, I think I'm still recovering from it around open source to some degree but a year out now I'm able to start thinking about why I got burnt out and what factors led to it and so one of the things that really helped me there was Kathleen's talk and these are a couple of three slides from her talk and she talked about a number of different things that can cause, that can lead to burnout, balance, mismatching expectations, pacing of work and a feeling of a loss of control and for me it was really the first and the last ones I think that were the biggest reasons. Balance was one, I was writing code full time during the day and then going home and writing code full time all night, I was basically only doing the one thing, the only difference between what I did during the day and what I did in the evenings was the license under which that code was written which is pretty irrelevant to the work that you're actually doing and so for me this was really, you know, balance really translates to what I'm going to talk about in the next section is money, right? I wasn't getting paid to do the work that I felt like I had to do and I had to do the work that I was getting paid because like mortgages and bills and car payments and that sort of stuff so I had to do like double work one out of a feeling of obligation and one out of a feeling of like we live in a capitalist society and I need money but they were the same thing and so I couldn't find that balance in my life and then loss of control so I, you know, I was supposedly a dictator, right? So why couldn't I do anything about toxic people in our community? I was supposed to be ruling by an iron fist and so why can't I do anything about harassment and abuse and that feeling of not having control over your own life and seeing no way to sort of change the equation is really one of the sort of key contributing factors to burn out or at least it was for me. What's really hard about this is I see this repeating in Django community again. I've seen the next cycle of really active contributors step up in the absence that Adrian and I left, burned themselves out, leave, and have been replaced again with yet another cycle and we are returning through people where we're using our community like an exhaustible resource, burning them out and then going on to the next people who don't know that this is coming for them a couple years in. I don't know what to do about that. I don't know how to prevent this. I don't know how to build a more sustainable, a more long term community that doesn't just exhaust people and then cast them aside. So money. So there is a ton of money around open source. The one that kind of annoys me personally is Facebook buys Instagram for a billion dollars, one billion dollars and Instagram has not donated a cent or a line of code to Django or to Python, which they built those technology that on top of. There is so much money around open source. There isn't a company out there that is not directly more profitable because they're using free software. And yet very, very few people actually are getting paid for their work on open source, a shockingly small number of people. I was actually one of those lucky people. I got paid to work on open source. In fact, I got paid a number of times to work on open source. And it never really worked out. When I was working at the newspaper and we originally released Django, the agreement was kind of part time. Like, yeah, you can work on Django when it's relevant because we're using it at work and don't want it's not. And sensibly, I was being sponsored by my employer still. But what happens with that sort of loosey-goosey arrangement is that over time, work always seems more important. There's never a good reason to give your work away. And especially when you're working in this sort of deadline-driven atmosphere that you are inside of a news organization, it's really hard to say, well, yeah, I know that we've got the big piece on government salaries coming out on Monday, but I really need to fix this bug report. So when I had the opportunity, a company offered to pay me full-time to work on Django until 1.0 was released just to get it out the door. And then half-time thereafter, it seemed like perfect. Like, that's exactly the right solution. Like, they're going to explicitly make working at open source my job. And for the first part of that worked really well, like, was being paid 100%, only working at open source. It was brilliant. It was like the dream that every open source developer has. That half-time thing, I totally blew it. I straight up kept working full time on open source. And I remember really clearly the phone call I got from my boss after about six months of that being like, dude, it's not working. This is not working out. And he was right. He remains a really good friend, despite having fired me. That's a sign of a good manager when they can fire you and you still feel good about it. But I couldn't make it work. Like, if half your job is doing something really exciting and fun and the other half is kind of trying to build a startup so it can get bought. So then I founded a consultancy with the idea that, look, if I work for myself, it will be super easy to set my own priorities. And that's a hilarious idea. Because suddenly when you're the one paying the paychecks, open source never seems like a good idea. And so I finally just kind of threw in the towel. Now I work for Heroku. I'm a manager full time. I don't write code at work. We don't use Django or Python at work. I work in a different field. I work in security, not really in software development. Like, I just kind of gave up. I'm not being paid in any way to work to work on open source. And in some ways it actually feels healthier. So I couldn't make it work. And the fact is, this is not really that unique a story. Like, there are so few people who managed to make working on open source something that actually works out for them. This is David McIver, who works on a tool called Hypothesis, which is a really cool, anyway, it's kind of like a testing tool for Python, sort of doing provably correct algorithms. It's really neat. And being used super broadly. And he basically posted recently about just giving it up. He said, you know, I've done a reality check and I found reality wanting. I've been trying to figure out a way of making Hypothesis development sustainable. And the answer is basically I can't, despite the fact it's clearly going to save people millions of dollars over the course of its lifetime. We've built an industry on free labor. And we've concluded that we'd much rather make people work for free in their spare time than fairly compensate for their labor and get good software. This is the same point that Ash Dryden made maybe a year or so earlier, the blog post, the ethics of unpaid labor in the OSS community. People contribute to open source freely and mostly enjoyably. I enjoyed it for eight or nine years of that 10-year arc. But we've somehow Ash writes that we've been culturally talked into accepting this arrangement and not realizing that businesses are using it to extract value from us. The idea behind open source was to break free of proprietary software, to break free of this pay to play. And instead, she writes, we've ended up in a scenario where we are now paying for the development of software that large companies financially benefit from with little cost to them. We've somehow built this system where, yes, open source one in the sense that every company is an open source company now. Microsoft is open sourcing.net. How much more of an example do you need to show in the open source versus proprietary battle we won. But we won at such a cost. The cost now is that thousands of people are donating money to Microsoft and to Oracle and to Apple and to Google. That's weird. Would you write a check to Google? So why are you working on open source for them? Very strange. So maybe this is getting better. I think there might be some signs that we're starting to get a little better about money in open source. Certainly in the Django community we've seen several successful Kickstarter campaigns that have paid people to work on large, long-standing problems. Django for the first 10 years didn't have any sort of built in migrations framework which was probably the biggest usability problem that we had. And Andrew raised about twice what he asked for to build a migrations framework for Django and it's great and it solved a big problem and he got paid and everyone's happy. So that's good. I worry that these are proprietary platforms and we're sort of hitching our hopes and dreams to someone else's profit sharing or profit center again. We have a fully funded Django fellowship now. We pay one person full time to work on Django. This is a little bit weird. It was easy to fund the first year. I wonder what's going to happen when we go back to that well the next year. And it's also strange because of the laws around non-profits in the U.S. what we're allowed to employ someone to do is complicated and we have to be somewhat careful about how we structure that work. Maybe the answer is better corporate citizenship. There are other funding models. I have some links here you can read about them. So maybe we're starting to figure it out. But on the other hand venture capital is also investing in open source. And because clearly the way to make open source less exploitative is to get venture capital involved. I don't know maybe we need something more radical. Like maybe this is not a technology problem. Maybe we need basic income. Maybe that's how we solve this problem. Maybe we need maybe we need a strike. Can you imagine if everyone who worked on open source said fuck you pay me. So I couldn't make it I couldn't make it work financially. But for me the thing that really sapped the remaining amount of my will was was worth the toxic people in our community. Sorry I'm a little bit nervous talking about this. This is a tough story to tell. So Jeremy Duncan was a friend of mine. I considered him a friend for almost 10 years. And when I found out that he had been systematically creeping on and harassing women in the community for many years it was a it was a gut punch. He was banned from from double union which is a feminist hacker space in San Francisco. The only person to have been banned. I found out about it slightly before this this blog post was published when a friend reached out to tell me what was going on. And this is difficult to talk about because I really can only talk about what's what's been published publicly because to do so otherwise would betray confidences. And I'm not going to do that. And this is really common when you're dealing with toxic people. They take advantage of a power imbalance to ensure that the people that they are to ensure that the people that they are harassing are going to be less likely to be able to speak out about it. My my willingness to participate in a space with someone like this is basically zero but I also have been unable to entirely kick him out. And this also is common because we rely on this idea of having enough evidence that someone is actually a harasser. So these are some of the reactions that people had to that blog post. I'm not citing a source here. I'm usually try to be fairly careful about citing sources because I really don't want to be sort of shaming the people who said this that several of these quotes are from people who I who I respect highly. And I think that they're coming from a good place. My point here is to show how how our sort of patriarchal expectations reinforce a place where we allow toxic people to stay. These people are coming from a good place but they're reacting to an accusation of harassment in a somewhat unfortunately predictable way. You know I didn't find these emails creepy but he's a fantastic ally. Why are we publicly shaming someone. I don't know what happened. It didn't seem creepy. Jeremy is a good person. I haven't seen enough evidence. And these are these are really common responses. To understand this I turn to a collection of essays that I found really fantastic. Revolution starts at home confronting intimate violence within activist communities. Now it's specifically about partner violence in in activist communities. But it provides there there's a lot of these collection of essays and they are all somewhat tough and also really amazing to read. But it provides more it speaks to more than just sort of partner violence and activist communities. It really talks quite a bit about how and why abusers are allowed to stay within our communities even when they are explicitly set up to combat those types of of abuse. So there's a section at the end called community accountability within people of color progressive movements which is just an amazing resource to understand that what goes on when we're refusing to see abuse in our communities. The author's right patriarchy upholds and supports gender oppression. And they identify four ways in which this works. Denial, minimizing, victim-blaming and counter-organizing. And so I want to go through those and then we'll look back at those quotes and you can kind of see how this functions. So denial looks like characterizing issues as personal or private. Writing off sexual harassments as misunderstandings or flirting or oh he just likes you. Brief aside the authors of revolution starts at home always use the pronoun he to identify abusers because abusers are predominantly male. But they acknowledge and I want to also that that's not always true that's simply the predominant gender in abusive relationships. Minimizing looks like again calling something a misunderstanding saying that it's taking away from the real or the important work but this person is such a good ally but this person is such a good activist their work is important or deliberately addressing issues in a very ineffectual and non-toothy way. Think of the conference that I think was reader Khan that reacted to an abuser in their in their community by banning him for two years. That's addressing the problem but what makes you believe that after two years he's not going to repeat the same behavior. Victim blaming among the other ways it works is by characterizing the abusers as victims they're nice their heroes they're important to our work why are we publicly shaming. And counter-organizing looks like deliberate reaction to two accusations threatening to fire people as happened in this case discrediting the person who raised the concern rather than dealing with the substance of it. Questioning the legitimacy of the concern. Questioning the need for accountability in the first place. So I did not find these emails offensive is denial and minimizing. I have always found them to be an ally this reeks of public shaming victim blaming counter-organizing denial minimizing victim blaming victim blaming denial counter-organizing and again these are people I respect and think highly of and think genuinely want to deal with these problems these are not you know these are not anonymous trolls these are not you know anti-feminists these are people who are trying very hard to grapple with these issues and are still falling into this trap that our culture lays for us. And the problem is most abusive men don't seem abusive except to their victims. And that observation comes from Lundy Bancroft why does he do that. The best thing that I've read into that really helps understand the psychology of men who harass and abuse and why and why they do that. I knew Jeremy for 10 years and I didn't have a clue. The problem is that we approach abuse and kind of this legalistic frame you know we need we need a conviction beyond a shadow of a doubt or you know preponderance of evidence. And this is a mistake we're not a court we're not a justice system we don't need to have proof that someone is an abuser I mean let's let's take a thought experiment right let's say let's say there's someone in your community like this and you have a suspicion that they're behaving in some in some crappy way. So you've got two options you just get you can kick them out without much evidence and so what's the ramifications if you do that well they can no longer be part of your community. Maybe it makes it a little harder for them to get a job probably not especially if they're a white guy because we benefit from assumed competence but okay maybe it's a little bit harder because they can't claim to be part of your community anymore. They don't get to come to community events. Is that really a big problem? I don't think it's huge. Let's flip it around let's say you let them stay and the accusations are true. Well how many people are they gonna are they going to exclude from your community because they feel harassed and abused. One, two, ten, fifty? The harm of letting someone stay is so much greater than the harm of kicking them out even if you're wrong and yet we still approach these things from this idea of needing evidence and even as BDFL I did not have evidence I still don't have evidence to really be claiming if I if I need to prove in a court of law before I can get this before I can expel this person. The problem is that being fair and balanced benefits the abusers. Bencroft writes that it's not possible to be truly balanced in one's view of an abuser and an abused woman. Neutrality actually serves the interests of the perpetrator much more than those of the victim and is not neutral. In reality to remain neutral is to collude with the abusive man whether or not that is your goal. We live in an unbalanced society so neutrality is unbalanced. So the worst part about this the thing that's really difficult is this was not the first time. This was not the first time that someone I knew and considered a friend I later found out to be a toxic person. It wasn't even the second time this has become common this has become a thing that happens. Somebody tweeted I went looking for it I couldn't find it so if you know who to who to site here please tell me that if you're a man and a woman hasn't told you about an experience of harassment it's because nobody trusts you enough. It's not because you don't know women who have been harassed and this wasn't the last time. There's a person right now in the central to the python community who I even once considered a mentor who is acting in some pretty shitty ways. I have an opportunity right now to name a name and I don't think I can I'm not going to because the harm that it would do to the people who can't come forward is greater than the benefit that I would gain by calling this person out. We've built systems that don't force accountability on people who behave in this way and I don't know how to change that equation. We have codes of conduct in their start they at least tell us where the boundaries are but they're not enough if we can't actually do something about people who cross those boundaries if we can't change the legalistic frame that we that we view these these problems from we have to figure out a way to remove toxic people from our communities because if we don't non-toxic people will leave. That's what I did I left and I've stopped trying to get new people in because I don't know how I can in good conscience say you should come be a part of our community when I know that we don't have the tools to deal with the creepers. So I had three problems that I couldn't solve and I still can't solve I still don't know how to fix these problems. The reason that I wanted to give this talk in particular to this group is I think if anybody has an idea how to solve them maybe it's maybe it's some of you. Thank you.