 It was a very warm welcome. And yes, I actually feel very much alive. I also feel a bit nervous sometimes, but that goes away after some time as well, so that's really helpful. Yeah, my name is Lena Reinhardt, or Fox with four Fs and UXs in the end on Twitter. I'm a project manager. I'm also specializing in marketing, communications, and community care in tech. I'm currently working on the open source project hoodie, which is an API for making web development easier. And I'm also working, as he mentioned, on Apache CouchDB. I'm also a co-founder and CEO of The Neighborhoodie, a company specializing in software development and consulting. What you can see here behind me is an acre in Southern Germany close to the village where I grew up. And this is where this talk begins. This acre in Southern Germany is cultivated land. This is culture. The English term culture evolved in the mid 15th century and originally meant the tilling of land. It was about who cared for which parts of land and who brought in the harvest. There's also a figurative sense of the term culture, which we mainly use it today. Our modern term culture is the expression of the way we act and refers to a group or community, which shares common experiences that shape the way its members understand the world. I work in the tech world and I contribute to open source. And I love what I do there, because I found amazing people and great projects which have become near and dear to me. Still, I feel that the future of open source is in danger. This is why in this talk, I want us to take a look at the culture in the world of free, liberal and open source software today. And to make it a bit easier for both you and me, I'll just refer to open source from now on. And even if you're not contributing to open source yourself, most of these topics are applicable to general tech culture of our times and to human interactions as well. Technology wise, open source is highly relevant. Mobile, social media, cloud services and big data for very important technologies of our time rely heavily on open source. Open source is also relevant to business. It contributes 450 billion euros per year only to the European economy. So things are going great in open source, right? Let me show you this chart first. The amazing people from the geek feminism wiki are collecting sexist incidents in geek communities, including technology industry, open source, gaming and more. This is the number of sexist incidents per year starting in the year 2000. Today, on October 3, 2014, the number of sexist incidents this year only is 60. Although this graphics does not only display incidents in open source communities, it depicts one thing clearly. There's a lot of things going wrong in our communities and our culture. And here's why. Often, when you talk about open source projects, people quickly refer to the amazing community. So let's take a closer look at it. Diana Harrelson, an anthropologist, did scientific research on the Fedora open source project. 75% of the respondents to her questions agreed that they felt that the Fedora project was a community. Some added answers like the community is the project and without the community, we're lost. But when we're talking about community and appreciating it, there's one essential point we must not miss. As Joseph Reyes, a philosopher, phrased it, if the culture is decaying or if it is discriminated against, the options and opportunities open to its members will shrink. Community is not just about building nice stuff and hanging around with nice people on mailing lists or in conferences like this one. Every troll, every sexist comment, every harassment towards just one single community member will directly harm this person, the entire community, the product that you're building and finally open source in general, its values, its ideas and its existence. One major deterministic of culture is language. What you can see behind me is Mars climate orbiter. It was a robotic space probe launched by NASA in 1998 to study the climate atmosphere and surface changes on Mars. This space probe finally got lost in space. It disintegrated due to atmospheric stresses. What was the reason for this robot's death? It was caused by human communication failure. Four pieces of software producing and expecting data in different units. Two navigators from the teams involved had pointed out those issues, but their concerns had been dismissed. What happened here is briefly described by Conway's law. Introduced in 1968 by Melvin Conway, a computer programmer, it states that organizations which design systems produce designs which are copies of their communication structures. Open source carries Conway's law to extremes. As a scientist phrased it, there's a hybridism of dialogue and code where the dialogue is directly embedded in the code. They called open source a distributed network of people and things that is constructed through the materialization of language. In other words, all communications in an open source project will have direct impact on the product like the software that you're building, all communications, no matter if they're peaceful or violent or if they're none at all. I want to show you a part of the Endangered Languages project. Every dot you can see on this map here stands for one of over 3,000 languages that are currently at risk of becoming extinct. Well, language loss is no new phenomenon and even if a few of those languages disappeared, large parts of the world population could still talk. So why should we care if languages are lost? Let me ask you two things first. First question is, who of you ever contributed to a free liberal or open source project? Oh, that's really plenty of people, cool. Second question, who of you is a programmer or has ever done anything related to programming like editing a bit of HTML or CSS? It's almost the entire audience, amazing. As people who can deal with programming languages, you know that one of their core functionalities is that language shapes reality. One change in an expression of the source code of an application can affect everything or break everything, some of you may have heard of that. Language is an essential part of our culture and shapes the way we express ourselves. And this is why when one language disappears into oblivion, we are all diminished. Thus, silencing people and their voices in open source destables and endangers each and every one of us. We in open source have to stop silencing people when they speak out about threats, mobbing, sexism, or other topics that show our broken culture. Open source is the materialization of language and our community culture influences everything we do. This is why we have to take even more care of the culture in open source communities today because the future of open source will be mainly determined by its culture. Let's see which other aspects we have to take care for in our communities and start this with an excursion to biology. As Charles Elton, an ecologist argued, simple, non-diverse communities are more easily upset than richer ones. That is, they're more vulnerable. In biology, there's a special research topic for this. And part of this field is the so-called stability diversity hypothesis. In short, this theory states that the more diverse the community is, the more stable and productive it is. A great example for a space which is diverse by default are rainforests. Typically, they possess a great deal of species diversity. Around 40 to 75% of all biotic species are native to the rainforests, same as with coral reefs. They occupy less than 0.1% of the world's ocean surface, which is around the size of France, yet they provide a home for 25% of all marine species. Diversity means variety and dissimilarity. It is a state and process of involving people who are different from each other in a group, and it aims to create an inclusive culture. Diversity is in natural settings usually nothing we have to implement or fix. It's the default. In artificial settings like open source communities are, we have to take care of diversity ourselves. And first and overall, because diversity is just the right thing to do. As various studies show, diversity also enables us to solve complex problems better and faster, be more creative and stimulated through persistent exposure to minority perspectives, make better decisions and generate more innovation. This means if a community is not diverse, it's broken. Diversity includes age, ability, ethnicity, socioeconomic class, personal background, gender, and far more. But people do not fit into a single box. Each of us falls into a variety of different dimensions. Every one of us has had different life experiences and perceives the world differently. Thus, we need to take care as well about the intersections of these groups. Let's take a look at some examples. According to the 2013 floss poll, 89% of contributors to open source are men, 11% are women. But this is not just about a binary gender system here. Currently, there's so little space for LGBTIQ people in the open source culture of our times that they're not even numbers. And it's worse for people who are marginalized twice or even more, for example, because they're women of color. We also need diversity in ethnicities. White people are still forming the major mass of contributors to open source. And we urgently need people of all ethnicities in our projects. We also need to aim for diversity in skills and get non-coding people like designers, writers, people with organization skills, and many more on board. We need diversity in ideas and backgrounds. The broader the amount of ideas, backgrounds, and experiences in our communities, the further we can collectively go. And we need to aim for more. A lack of diversity leads to serious issues and destabilizes the complete community. These issues can exemplarily be illustrated with an anecdote like one from Google, which they just published last week. When YouTube launched their video upload app for iOS, five to 10% of all uploaded videos by users were upside down. Were people shooting the videos incorrectly? Well, actually no. The early design was the problem. The YouTube app was designed for right-handed users, but phones are usually rotated 100 degrees when held with left hands. Without realizing it, the YouTube app team had created an app that worked best for their almost exclusively right-handed team. Although this is a very simple example, it can as well be taken as a metaphor. And it depicts working on more diversity in our communities should be a daily task for every one of us. Or to phrase it like Aaron Hartwig recently did on Twitter, when everybody is making technology, the technology they make will be for everybody. Community means the appreciation of diversity and variety. Community culture is our daily answer to the question who, which people, and what, which kind of contributions are being welcomed and valued in our projects. There are strong values and goals behind open-source software development, which are the reason why it exists and why it has kept alive. Some of these values are, for example, providing alternatives to closed-source software, enabling independence and driving innovation. But there's one more aspect, aspect why open-source software is built. So let's think about software development in general. Software is built to perform useful work, to run a computer system, to solve problems, make things better. But make things better for whom? Software is built for people. It's built for users. Any software exists for eventually being used by someone somewhere. Of the complete world population, only 18 million people are software developers. This is 0.26%, and that's the small blue line you can see in that green circle. And this number already includes hobbyists who are just writing code in their spare time. This means that for every software developer out there, blue person in there, there are 399 people who have other professions or just don't code. Now let me just give these 399 people a bit more color to get them closer to reality, and there we go. When we are talking about software, we can't have this conversation without talking about users. Software is built for them. And each of those 399 people over there have their very own individual needs, and even our developer themselves will eventually be a software user. This means that building software is an act of representation. Many of us here, we are very privileged, and our realities are very far away from the wants of millions of people on this planet. The users of our software are a very diverse group, and representation means responsibility. As people who work on software, we have this representation role to act according to. And this is why non-diverse communities working on software cannot be justified. Creativity is a driver of good design and engineering. It's the ability to see connections and relationships where others have not, and it's the result of making unexpected connections between things we already know. Hence, creativity depends on our life experiences. This is why diversity is essential to good design and engineering at a very fundamental level. Only if we ourselves are a very diverse group of people who thus are capable of imagining the wide array of possible values, we can uncover an equally wide array of possible design approaches and results which then offer built-in solutions for a wide range of people. Only moving beyond self-referential modes will enable us to develop infrastructure, processes, products, and work that resonate with the broader population. By giving people more ways to speak up and more tools to take action, we keep decreasing the gap between what we wish for and what we can do about it. It was recently written by the author Seth Gouda. Fulfilling our representation role properly when we're building software means that we acknowledge the needs of people in diverse communities, that we work on software that resonates with the diverse group of humans on this planet, and that we build tools that empower people and enable them to solve their problems themselves. And yes, part of this is education. This is what enablement, empowerment, and representation really mean. Misrepresentation leads to serious issues. One of them illustrated for one programming languages in this tweet. The JavaScript community loves good enough solutions where good enough means horribly broken, except for my case. We already have representation problems in our developer communities and it's not getting better for people outside of those communities. We have to fulfill this representation roles properly in our communities so we can also finally enable people to really trust the software we're building. I want to give you two examples which depict the trust issue we're exposing our users to far too often. The queer chorus, a choir group in Austin, Texas had a Facebook group for its choirs members. The president of the chorus added two new members of the choir to this Facebook group. What the president didn't know, Facebook automatically told all their Facebook friends that they were now members of the queer chorus Facebook group. Thing is, both these people had decided to not inform their friends and families about them being queer. It was Facebook that did that for them. By inadvertently exposing this information to their friends and families, Facebook made a choice instead of leaving it up to them. A second example. Grindr shows the precise distance between users and therefore makes it possible to locate them. Grindr is a dating app mainly targeting homosexual people. It said that it's being used by 6 million people in 192 countries. This means it's also used in countries where anti-homosexual legislation, in countries with anti-homosexual legislation, where homosexuals are persecuted, imprisoned and killed. When you want to use Grindr on your mobile phone, you share your location with the app and it displays profiles of people and their distances to you. Without telling you in which direction of you they are. Thus it didn't show the other people's precise position, right? This quote you see on this slide is from an email which was sent to Grindr LLC, the company responsible for the Grindr app. It was sent in February this year and it was sent because of problem with Grindr and this distance to you functionality. This problem is called triangulation and it enabled you to see the actual position of a person, even though this was not a direct feature of Grindr itself. It was possible to see where Grindr users were. They were in the US Senate office buildings, the British House of Commons, as well as in Moscow and Kiev, Ukraine. Pointed to the problem with the email you've seen before, Grindr reacted with a blog post, telling users, if you have a problem with the location functionality, you can turn it off any time. It's not a bug, it's a feature, as if this were true. It wasn't as easy as that. Grindr had not enabled their users to make an informed decision. It had kept them unknowing that by enabling that other users could see their distance to them, it would also be possible to find out their precise location. And this misinformation of users even led to serious incidents, leading to headlines like, Egyptian cops using Grindr to hunt gays. Reports say it led to people being arrested. It took Grindr LLC seven months to fix this serious problem. And all of this from a company that then proclaimed in a blog post, there is nothing that matters to us more than the safety and security of our users. Both these stories about Facebook and Grindr are about responsibility. Tech is not neutral. Many more people have been stung by accidentally revealing secrets online that were easier kept in the past. We need to keep examples like these in mind and build software that respects its users. The topic of proper user representation is also linked with one of the core values and goals of open source. Freedom. Freedom is the idea of giving users choices, power, and control over the tools they use. It's often stated that open source enables users to see how the apps they use work, check if they're secure, and change them themselves if they want to. So let me ask you another question. Who of you has at least once read the source code of your mail server that your email provider runs for you? Nobody? Maybe next time. So we have to ask, how many people can actually do this? How many people have not only the interest, but also the resources and knowledge to check the source code of their software? How close to people's actual realities can this idea of freedom practically be? The idea of long-term freedom for users through open source is a great goal which we have to perceive. But we have to take care to not make this a patronizing ideal that forgets about people's actual realities. Instead, we have to build products understanding people's needs and their capabilities in diverse communities so we can make this ideal of freedom a thing which is closer to people's realities than it is now. Another danger which we're currently facing is shown by what happened to the Mayas. Although their culture and civilization were highly developed, they declined and suddenly disappeared around 800 to 900 BC. It is said that this was the result of a combination of prolonged droughts and extensive, long-lasting internal conflicts. Some archaeologists argue that the Maya Collapse was merely a collapse of the ruling elites. These theories can't be proven 100%, but the Maya show us that it's neither change nor technology that threatens the integrity of the culture. It is power, the crude face of domination. If the needs of individuals in one culture are continually suppressed, social systems can become unstable. One person in Fedora Project Study, which we've talked about before, said, I used to believe that this project was a community, but it seems more like a grouping of various anarchists and monarchists who think everyone else is like them. We really have to take care to avoid cliques and elites which exclude community members and enforce those unhealthy power structures that destabilize our communities. We have to stop the marginalization of people, the worshipping of heroes, bro culture, rock stars, code unicorns, and people who cannot be criticized anymore because of their status. This leads us to two other core concepts and goals of open source, which are decentralization and democracy. The open source model includes the concept of deferring agendas and approaches in production, in contrast with more centralized models of development, such as typically used in commercial software companies. These decentralization approaches include peer production and public availability of all sources and products and documentation. The development in open source is based on the so-called Bazaar model, in which roles of people involved are clearly defined, and that states that users are treated as co-developers. This was actually the case in the early days of the movement when software developer bases were very close to their users. Feedback from them was not just welcome, it was vital. Feedback from users ensured that the code met actual needs, which powered its development and hence further commits. Thus eventually, users and developers were both contributors to the software. But as the systems matured, users changed from being integral elements of the community to playing an early marginal role. This has led to a widening democracy deficit in open source, with users increasingly isolated from the people writing the code. Today, we're far from a state where many open source projects would even try to engage with normal people. Even data shows that open source is not quite as decentralized and democratic as it proclaims to be. Analysis shows that of 5 billion bytes of open source code, 74% were written by the most active 10% users. Democracy and decentralization in open source will both require more diverse communities to enable more stable democratic processes, getting the products we build closer to their users and building closer relationships with them and achieve real decentralization. Open source projects often also proudly refer to their meritocracy, the belief that those with merit should float to the top, that they should be given more opportunities and higher rewards. Funnily and sadly at the same time, the term meritocracy was originally coined by Michael Young in 1958 as a negative example for a system with highly critical approaches and actually to point out that such a system could never exist. Noah Slater recently described meritocracy as a sort of meta-story which we repeat to each other and which we use to construct other stories which then explain, for example, why people are included or excluded from our projects. Current accounts of participation in projects tend to gloss over the political maneuvering that often takes place around the integration of new members. Instead, they describe this oversimplified and supposedly meritocratic process broadly inspired from the ideal of the scientific world where the best ideas and people get naturally based on their talent and peer review. Meritocracy is often celebrated as objective whilst the homogeneity of open source and various studies show that it exacerbates the lack of diversity and institutionalizes structural inequality. This also includes, thus, that we have to find new models to value contribution. When we are thinking about the future of open source, we have to include those who have less opportunity, less time and less money which would allow them to freely contribute. We need to rethink what is being valued in open source projects especially with the needs of marginalized people in mind. And our main question should be, who do we care for? Contributors who contribute? Or do we care for people? This question is an essential one and a topic which is near and dear to me. People in our communities experience not only good but also bad times, burnouts, mental health issues like depression and much more. These are serious issues which we can't ignore and some of them are even enforced through structural inequality, meritocracy and community related issues. We as members of open source communities have to implement a culture where mental health issues are not stigmatized where we talk openly about them and lead an open discussion about how to avoid people from burning out through open source contributions. A culture in which people are heard and they know that there are people who care for them. We have to make sure to be there for the people in our projects not because we want them to keep on contributing but because we care for them as people. We have already seen the relevance of language. Language shapes reality. In her talk yesterday Kevou talked about asking versus guessing culture and Ariane Asbel mentioned the relevance of language in his talk today. I want to take up the great points we have made and talk about the explicit and implicit here. The explicit and the implicit are both parts of our communication. The explicit is often characterized as the direct way of expressing things in saying or writing. Explicit communication is usually specific information conveyed in written or spoken words. Language is extremely powerful and being explicit about things you feel in or want can be very helpful in human interactions as well as in communities. The implicit, its contrary, is often described as the more indirect way. Implicit communication refers to things we do individually or collectively. It can be body language and even the decision to not communicate at all is implicit communication. Implicit communication can be just as powerful and effective as explicit communication. And sometimes our implicit communication can even erase everything we communicated explicitly. Culture, which you've talked about before is part of our implicit communication. Now we as people usually send and receive both messages the explicit ones as well as the implicit ones. And ideally the explicit and implicit communication we receive and send a line and lead to consistent, coherent and harmonious overall picture. On both layers the explicit and the implicit one. Thus, every one of us has to take care of what we actually communicate. We have to watch out carefully what we communicate explicitly. Here's a few examples for what to watch out for. If you don't have one yet implement a code of conduct for your projects and events that sets explicit limitations to behavior. Still, any code of conduct is completely useless if there are no people who really take care of it. It is not worth any millisecond of loading time on your website if there are no people who actually will enforce it. And it's worth rubbish if the people in your community or at your event can't trust that it will definitely be enforced. Neither will a code of conduct fix any diversity problems. It can support work on diversity topics but it can't be the only thing that is done in terms of diversity. A code of conduct also will not directly make your community a safe space. We have to think specifically what we can do to provide safe spaces for people in our communities and how we can specifically support marginalized people. We also have to take care of what we actually communicate in our language and how we do it. What language is being used in our communities? Is it comprehensive and accessible, also for people who don't know our insider jokes yet? How do we handle swearing, which some people are very opposed to? And how do we react when people say they're feeling uncomfortable with a specific use of terminology or language? Do we talk about diversity in our communities? Do we talk about in which terms we mean diversity, what our motivations behind work on diversity are, which approaches we could choose and how we'll work on this? Have we ensured that not only marginalized people work on diversity themselves? Have we asked for feedback outside of our communities and also from people who don't have the same background as we do? Is improving our community culture a topic which you regularly talk about and which everybody is working on? We have to take care as well about our implicit communication. Is our project hosted on GitHub or a similar platform? Then it may be open for developers, but it excludes non-coders like designers or writers by default. Is the major mass of people in our projects white men? Then it can be extremely hard for women or people of color to join us. Is our so-called diversity work limited to posting one blog post per year, a statement on Twitter, or sending one or two emails? Or are we really putting effort into it? How is diversity handled at our events? Is it just about making sure that at least one or two women are speaking, or do we really care about this topic, read about it, learn about it, and see what matters if we want to create inclusive events? What about our software documentation? Is it referring to users as he and using mail terminology and non-gendered language? Then why should women feel addressed? Do we discuss everything extensively on mailing lists and even make decisions there? Staying up to date with this discussion is a lot of time and time that especially marginalized people usually don't have. For example, because they have less spare time or have to do care work. It excludes them by default. Or do most or even all of our community members contribute voluntarily. This is great for us. But this again excludes by default people who have less spare time or earn less money which would allow them to freely contribute. There are many, many more aspects in addition to these examples. We have to raise our awareness towards what our words and our actions tell people, what messages we send and which messages they may receive. We have to find ways to become better humans, better communities, and better places for people and work on improving and finally getting away of our toxic culture. You know that in our bodies we have arteries and veins, both part of our circulatory system. Arteries are the blood vessels that carry blood away from the heart, mostly oxygenated blood. There's a special coronary artery which is located directly at our heart. Its job is to supply blood to the heart muscle and thus keep it working. Still, it sometimes happens that there's plaque building up along the inner walls of the arteries of the heart. This plaque narrows the arteries and reduces blood flow to the heart. It can progress without anyone noticing evidence for it for years. And as it progresses, it leads to lack of oxygen in the body's cells, can cause chest pain, and finally the two heart attack and death. In open source our communities are our hearts. They nourish the entire body, our hearts, without our communities we are lost. Narrowing those communities by limiting and restricting their spaces leads to serious problems for individuals, communities, the products we are building, and finally open source in general. There are already many initiatives and individuals in open source that are working hard on all of this. These people and their allies have a significant amount of time on improving the culture we currently have. Some of them are, for example, Black Girls Code, Ashtradon, LGB Tech, Model View Culture, TransHEC, Rails Girls of course, and many more. The least we all should do is listen to them, support their work, share what they're saying. We must not wait for others, especially marginalized people to educate us. Each of us has to educate us ourselves. Each of us has to take care every day to make open source a better place for everyone. And each of us has to transfer these things we are learning to the communities we are all in. We have to widen our communities, welcome and appreciate everyone, to ensure that our heart keeps beating, and to ensure that open source is better. Because open source can only have a future if it does everything to be inclusive. Thank you.