Transcription Chris Wright Interview
Interviewer Jenna Gretsch
April 2010
 PDF Version

The views and opinions expressed in this interview are solely those of the interviewee and interviewer and do not necessarily reflect the views or opinions of any employers.


Chris Wright has been doing Linux kernel development for the last 10 years. His primary focus has been security and virtualization. He's been involved in the development and maintenance of the LSM framework, is a Linux kernel security officer, co-maintainer of the -stable Linux kernel tree, and helped develop and maintain the paravirt-ops framework in Linux. Now he is focused on KVM and hardware assisted virtualization. He is employed by Red Hat where he is happily hacking KVM and the Linux kernel in sunny Portland, OR.


JG:  Hi Chris Wright, welcome. Let's start with what you currently do for work.

CW:  I am employed at Red Hat. I work on virtulization and I focus on sort of comprehensive systems view of the products we produce around virtualization. So in the past I've been more focused on specific aspects of the technology and now I focus on making sure all of the different disparate pieces fit together and work as a system and consequently as a product. It's pretty different from the bits and bytes of the low level technology.

JG:  Is that what you used to work on, the low level technology?

CW:  Yes I was doing a lot of the sort of infrastructure work that make something like virtualization function. Now I do alot of work of trying to understand where the technology is going and how we can sort of take it to the next step, how we can move the industry forward. It's interesting, it's a whole different set of challenges.

JG:  How did you get started working in technology?

CW:  Like many people that I work with, there's an aspect, a root of interest in technology that goes all the way back to childhood. I had a computer as a ten year old. It came with a manual and you could do a little bit of programming. My parents enrolled me in a couple of courses at a local college that were geared towards kids learning how to do computer programming. But it really wasn't until I was in college studying physics that I got exposed to Unix systems and I liked what I was exposed to. Thought somehow it was more intuitive and made sense to me compared to the Windows or DOS systems that I had used just as a person writing a document for a paper or a class or something.

JG:  Do you remember what type of computer you were on as a kid?

CW:  My first experience was with a VIC-20 which is around the same time, but not as powerful as the Commodore 64. Shortly there after the classes were focused on TRS-80 and then a little bit later there were some PC specific classes but still it was mostly BASIC programming. I lost interest as a teenager when I was more interested in using a computer to play a game and I was more likely to be outside or playing sports. Then in school when I started to have to deal with large amounts of data for the labs we would do in physics we would have to use these clunky, very command-line oriented text based interfaces to process the data of whatever statistics package or various tools that we had. Some were commercial tools and some were just local homegrown at the university, and I kind of liked that --- I got the feel for it, I kind of liked it. And then as an undergraduate graduating with a degree in Physics there's not alot of ways to apply that without getting an advanced degree and I wasn't that interested in doing an advanced degree in Physics but I had started to develop an interest in computers, or redevelop an interest in computers throughout the last couple years of my degree. It seemed like the most logical way to leave school and find a job if I had to do something vaguely physics related, since I had started to specialize in fiber-optics, I thought telecommunications had maybe had some computing aspects and sort of by blind luck my first job was in a telecommunications company. It was a very sofware-centric Unix programming, producing products based on Unix platforms. I learned a huge amount of Unix and programming skills on the job and that's really how I got started in computers.

JG:  When you applied for that job was your approach totally arbitrary or were you looking specifically for a job in computers?

CW:  There was an arbitrary element but I had decided that I thought I was interested in telecommunications, Unix and that I was interested in traveling. I applied for it and the luck part was that the hiring manager was a physics major in college and with some of the skills I had acquired from being at the university and using computers and having a background in physics, he felt confident that I could learn and he was willing to roll the dice with me.

JG:  Would you say he was your mentor once you got hired?

CW:  Yes, definitely.

JG:  What is the most demanding aspect of your job?

CW:  The most demanding aspect now is really the time associated with work. It's intellectually demanding just because things are always changing there's a lot of new technology there's always new things to learn or things that you know that you don't know how to apply -- you know there's always something to learn. But the part that is most demanding, in the sense that I feel the most drained, is just from the amount of time it takes to be focused and successful, it really takes a lot of commitment, time commitment. It was easier when I didn't have a family. With a family I find it harder to balance my time.

JG:  So you must not have always thought that you would end up in technology?

CW:  I did not. I was pretty convinced that I would be a french speaking lawyer when I was in the eighth grade.

JG:  Do you employ any sort of strategy for balancing your work and family life? Does Red Hat offer flex-time/flex-place?

CW:  Well there's a variety of things. Certainly there's the base where we have vacation time and personal time that you can take annually spread out throughout the year. On a day-to-day basis our company is head quartered on the east coast and I live in Portland, Oregon so I work completely remotely, I work out of my home. My hours technically are flexible, practically I have to do enough real-time personal interaction with people on conference calls who are all further east than me so I'm tied to a schedule that is sort of not my natural schedule. So it's not quite as flexible. When I was completely focused on programming things were really flexible; it didn't really matter whether you worked all night or early in the morning or whenever. When I was working in that aspect, collaborating with a community technical project with people all around the world, timezones become quite irrelevant and you have a sense of what you're trying to accomplish and as long as you're accomplishing what you're setting out to do, you're being successful. Now since my work is more focused on the product aspects of the technology and some interactions with how we work with our various corporate partners, things like that, there's a little more time-structure and less flexibility. But again we're all in different time zones -- I can slip out in the middle of the day for some exercise and lunch and feel reinvigorated because I had some time to myself and by the end of the work day I can reengage with my family and that brings the balance back.

JG:  So it's not frowned upon to leave mid day for some time for yourself?

CW:  Maybe it's my own personal work ethic because I'm pretty driven to stay focused on work so normally I wouldn't take a lot of time mid day without calling it a half day off but to take an hour to go for a bike ride and some extra time for lunch is completely reasonable. The benefit of working in an environment where a lot of our communication is done through email I don't need to be instantly in touch with somebody.

JG:  Is this flexibility understood as a policy throughout the company? It's not something that is just conferred upon programmers or senior developers?

CW:  You know I think that that is a great distinction because it may be different for programmers because there are aspects of jobs at our company that are not really technology focused, not really programming focused and there may be a much more traditional understanding of when you're supposed to be in the office -- I really have no idea because I have pretty infrequent interaction with people in those groups. From the tech side I think most people have a sense of flexibility and most of the people I work with are incredibly driven so oddly enough that flexibility means that they're just taking time out from an exceptionally long work day.

JG:  What projects have you worked on at Red Hat or elsewhere that you are most proud of?

CW:  That's tough. I've done a wide variety of things, I think what I'm most proud of is maybe two different types of things. I'm proud of some of the technical work I've done but it's a little hard to put a specific finger on it just because a lot of our work is requirements of existing -- you know you work on a large community project it's hard to say you added a red widget when there was already a red widget and now you've changed it to a widget that goes from red to pink or something. I mostly focus on the Linux kernel, in the Linux kernel development community was creating a whole framework for plugging in security modules so that you could redefine how you do access control in a Linux kernel. More recently having been focused pretty exclusively on virtualization some of the things I've been proud of aren't purely in the technology space but also in the work that we've done as a group within our company to bring together new technology, mainly an open source virtualization solution from sort of nothing to a really sophisticated and competitive product. Taking something from essentially a research project to something that's an enterprise ready, production quality piece of software is a huge amount of work. So a lot of the work I was doing was not just the technical work but also mentoring other people and helping bring people to the team. We started with a pretty small group of people and we ended up with a few dozen folks making a huge amount of progress. From an open source community project point of view it was a different experience for me since I had been mostly focused on the technology or the research aspects of primarily the Linux kernel. It's harder to see what it takes to turn something from the initial concepts, the prototypes the research into something that is really high functional and works for a large variety of uses and doesn't just crash every time the user does something the developer wasn't expecting. The amount of work involved in that I had experienced in earlier proprietary development efforts but I hadn't really experienced anything like this in the open source world and it was a good experience for me.

JG:  What draws you to open source?

CW:  There are two or three main things. One is as a geek, and engineer, a technology person, the ability to lift the hood and tinker. I have that insatiable desire to find out how something works and when you have all of the open source code available to you you have all of the tools you need to figure out how something works and then the next step is to modify something and see how it works after you change it and that can be really satisfying in and of itself. So that was definitely my first primary interest, just the fact that I could learn how something worked. The next thing I discovered by being more involved was how amazingly talented all of these different people are working on a project. You sort of have a sense when you're not involved that it's a motley crew of hobbiests working from their basements and there absolutely is an aspect of truth in that but it doesn't really capture the fact that you're dealing with really some of the most talented developers on the planet who are all exchanging ideas -- you're exposed to high quality work and a collaborative environment that's just engaging, it's just kind of contagious once you get involved it's really hard to imagine doing things any other way. So the three main points are an interest in tinkering, the collaborative environment, and the quality of people involved. People are really dedicated and coming at it from so many different angles that you're always surprised at somebody's uses or ideas, things that you wouldn't have necessarily thought of. When you're collaborating with a large group of people it's a very creative environment.

JG:  What about the methodology from the business model side, obviously tinkering with code and having it not be a black box is desirable but are there other reasons that you think that produces a better product in the end?

CW:  Well having sort of been lured in by my own interests and that contagious aspect of getting involved I started to see the definitive benefits of the collaborative environment. When you're working together with such a large group of people all focused on the same project coming at it from different angles that's harnessing a huge amount of creative energy and talent that is really diificult to do in a proprietary development model so from a business point of view it's hard to compete with an open source development model. You've got a large number of people and not all of whom are employed by a single employer, so from one company's point of view they're leveraging the work of a lot of people they're not paying for directly. From a sort of innovation view it really fosters moving the technology forward rapidly. For a proprietary development envoronment you have fewer resources, you have very focused deliveries throughout product dates and you're really scrambling to get the next marketable feature into your product rather than just focusing on advancing the technology. When you're focused on advancing the technology you may do things that sort of aren't conventionally done -- conventional wisdom or intuition might say this is an expensive set of changes and it's not worth doing from a business point of view but from the point of view of evolving the technology making changes regularly to your code base may actually benefit and create something that is moving forward rapidly and it's always adapting to the changing environment. So from a business model point of view I really feel that open source is not "the" way to approach the problem, certainly there are lots of ways to approach the problem, but I think it's one of the most efficient and effective ways for certain aspects of technology. One of the benefits you get from something like the Linux kernel is a huge number of different interests from pure hobbyists to a lot of different competing corporate interests and as you get further and further into a niche environment you'll find fewer and fewer interested other people and the communities will dwindle and I'm sure there's a tipping point where if your community is too small you'll just have a stagnate project that's just not doing the innovation and creativity that you'll see in a larger community project and at that point perhaps a model that's more structured around just features and delivery needs would work as well.

JG:  How has money affected your career choices?

CW:  I don't know that money has played a huge factor other than I work in a field where I'm reasonably well compensated so I haven't ever looked to leave the field but I have not made career choices within this field based on money. I've always been interested in the technology side so really it's just about being able to do the technical work that you're interested in doing and finding someone who is willing to pay you to do that.

JG:  Was that always easy for you to find?

CW:  I didn't always do open source programming. In the first few jobs of my career it was completely proprietary development. Again reasons for changing companies were not really about compensation. In one case the company went out of business in another case the project we were working on was clearly not going anywhere so the reasons for leaving the job didn't have much to do with financial compensation as much as how secure it was in the long run or if it was just getting so mired in corporate politics that it was time to move on and do something else. But from the time that I started doing open source development as a career, finding opportunities was really not that difficult. Especially in Portland at the time, the late 90's we sort of had the dot com bubble growing and a lot of online companies and Linux had gone from unknown to reasonably well known as a hacker platform to something that was a significant portion of the internet.

JG:  Thanks to Red Hat.

CW:  Yes and certainly many others. So in that time there was a lot of growth, there were really more positions available than there was talent available. So I felt pretty lucky that way. It's not like I had a pick of absolutely anything I wanted but it really never felt like a tight scramble doing open source development. At this point I've been in the open source development world long enough that I've established some credibility for myself. I don't know if it's the same now or not, somebody coming in now may have a different experience than I did ten years ago but it's been a pretty rewarding experience that way, feeling like you're in the drivers' seat. People doing Linux kernel development, most of the folks I know about have had similar experiences where they're essentially sought after.

JG:  What is the gender makeup of your current technical team?

CW:  The gender makeup of my current technical team -- if it's not one-hundred percent male, it's in the very high nineties. It's just not a well gendered balanced team.

JG:  Can you specify the team you're referring to?

CW:  What I am thinking of directly is the internal team I work with at Red Hat. Depending on how you define our team I'm thinking mostly of engineers working for a specific portion of the product. Essentially they're all men. If I expand it out just a little there are essentially a few women, and then if you expand it beyond purely the developer aspect of the work, that brings in a few more women.

JG:  When you say outside of the developer aspect, do you mean outside of programming?

CW:  Right exactly. There are a couple of women that I work with that do documentation. One woman who I don't work with personally but she's helped out our project does -- it's often called user experience and she provides some of the artistic content.

JG:  Is this for the virtualization project, is that what you're referring to?

CW:  Yeah and again confined to the actual corporate umbrella of Red Hat and the folks that I work with directly at my company. Once you get out of the technical side of the project and include project management and folks who manage the corporate machinations of the company, the balance shifts quite a bit. Which I would expect is not unusual from my experience with other companies.

JG:  Let's refine our focus and talk about just kernel development for a second.

CW:  Just kernel development that would be taking an open source project that has some people who are employed at Red Hat and many who are not, that community is not completely devoid of women but it's still a very male dominated community. In that community there may be maybe a dozen of women doing development and that certainly pales in comparison to the hundreds of men I've worked with on the same project. It's a flagship project for open source projects, it's well known and I don't know how well kernel development represents the gender balance in computer technology fields or developers in general or even specifically in open source. If you include a whole variety of other projects I don't know if it's completely typical or if it's tipping the balance in being really unusually male dominated.

JG:  It is actually unusual, by ten-fold compared to proprietary cultures which are greatly disproportionate too. It's 1.5% women for open source.

CW:  Yeah that kind of resonates with my personal experience of having worked for both proprietary and non-proprietary, open source worlds. Most of the jobs I've had where we did proprietary software development have had better balanced teams.

JG:  Have you ever had a woman technical manager?

CW:  I have. I've had just one woman technical manager and that was at a proprietary software company and that team was unusually well balanced, I'd say close to fifty percent women. It was by far the most women I had worked with in a development group, certainly still is.

JG:  Have you seen discriminatory behavior towards women in your job? I'm assuming this might be difficult to answer because in your core group you are not working with any women? So let's expand it out to include any groups where you have worked with women.

CW:  Even in the group that I work with which is all men, you can see some choices in how people talk about either technology or programming where you can see gender bias happening even without interacting with women directly. A common terminology or phrase would be equating an inexperienced computer user with a woman -- it's a mother, it's a grandmother, it's a girlfriend, it's a wife, it is rarely a brother, an uncle. You can see it in a group where we are not directly interacting with women -- when there are women involved -- I have to say in a larger project like the Linux kernel development project the majority of the people you have never met, it's an international community, I'm American, it's certainly not clear to me always the gender of the person I'm interacting with.

JG:  But do you think there's an assumption on your part that it's usually male?

CW:  Yeah, it's hard not to make that assumption just based on the numbers. One thing that's interesting is that you could very well be communicating with a woman and thinking that you're communicating with a man and never really know. And how that affects the way you communicate, you're just communicating using your basic assumptions and they could be completely off base. So in that sense you're not doing anything differently than you would if you understood the gender differences. One of the things I've heard from some of the women who do work in the Linux kernel is that it's kind of an aggressive environment and that turns off a variety of people. One group of people who is turned off by that is women but other groups are culturally based, certainly some Asian cultures have had a really difficult time getting involved in Linux kernel development and I think it's in part due to the aggressive way people interact. It's hard to distinguish the difference between somebody saying that's a bad idea or that's a dumb idea from you're dumb, or take it as a personal attack. And I know that some of the women that I have worked with have felt that where it's aggressive and hostile and unpleasant.

JG:  What I hear you saying is that the environment is sort of aggressive generally, but do you think there is any specific aggression that targets women for being in the environment?

CW:  I don't think so. Maybe I should preface that with, not in an overt fashion. Any of the gender biases that a person is carrying around with them where they're already assuming that it's the wife or the girlfriend or the grandmother or the mother who are clueless about computers some of that language will come through I think regardless of the gender of the person that you're interacting with. I can imagine that if you're a woman reading an email with that language that can feel very offensive in a way that a man receiving that same kind of email wouldn't even notice it. I don't know if that answers your questions but there's definitely a subtlety in there -- take it from a different angle, open source in general and certainly linux specifically, is often referred to as a meritocracy, you are awarded based on the manner of your work. Given the vaguely anonymous sense of just an email address I'm not sure how aggressively people attack based on just predisposed concepts of gender. So a male developer or a female developer could assert the same idea and I am not sure that they would be received any differently by the community based on the gender. I think that one of the things that I've heard from some of the women that I've worked with is that the feedback they receive is offensive enough, it's not just offensive because they are being discriminated against as a woman, it's just abrasive and not an enjoyable environment to work in. Having said that I'm sure you could find any number of men who really disrespect the women in particular as developers. But my feeling, and certainly my point of view is really focused on the technology, improving the technology -- in the ideal you would take away the person, the ego and all of that stuff and really be focused on the technology and practically it's hard to do that. I mean we're all humans we get emotionally involved in the work we are doing and there are arguments and disagreements that are not technical all the time in our community; it's politics it's people. I think conceptionally what comes first is improving the technology and from my point of view it's really hard to distinguish if the work that somebody has done appears different based on their gender.

JG:  I find the notion of technical meritocracy really interesting and it's one of the aspects that I was initially drawn to in technology but I've found that it's really difficult to decontextualize it. We don't enter a gender-blind world, we are all socialized creatures and this obviously takes place prior to us submitting any technical work. So what it has come down to for me is examining the ways participation is actually occurring, questioning if people really do have equal access to the modes of participation prior to getting to the higher technical levels and being able to submit work to projects. These kinds of questions have led me to reject all notions of meritocracies. Specifically I would say that we're just not there yet in terms of traditional ideas about who should be doing the brunt of domestic work and this notion has a very clear and cascading impact on the ability of women to have free time to contribute to open source code projects without getting paid, it becomes a double-whammy of unpaid labor. So meritocracy sounds ideal, it sounds even gender-blind, and it's often touted as such, but clearly it's not always what transpires because not eveyrone's work is reflected because not everyone is able to equally participate, this means specifically in open source there is a very large male-based conversation that is happening out there with the idea that it's a "meritocracy" that just happens to have 1.5% women.

Have you seen women being relegated to the mediator in the highly technical circles?

CW:  So it's really different work environments because the other earlier jobs I had were all proprietary endeavors. And in those environments the women that I worked with that were doing technical development work, were doing similar technical development work as the men on the team, it would be hard to make a generalization from my experiences that those women were falling into a general category of doing more mediation or organization and less technical development. In the open source world, especially the Linux kernel development world there are really not very many women working on the project. The people who are involved in mediating discussions are technical experts in the area so it's not going to follow any kind of gender bias where the sort of stereotype of a women being more emotionally skilled and capable of sort of soothing the different parties involved -- the people who are doing the work in Linux tend to be technical experts in the area and they are focused primarily on the technical arguments and when it starts to become just a personal argument, I think the reality is that at some point you just have to sort of flip a coin and pick one and they're most likely going to pick a side of the argument of the person whose work or opinion they generally have more respect for. So the person doing the mediation is not so much there for their mediation skills, they're probably the recipient of the work of two parties arguing that their work is the better way to do things, take my work, no take my work. I definitely recall working on projects where the technical leads were men or the technical leads were women and they both felt the same because the interest was in the work.

The interesting thing about technology is that you're talking about a group of people who have a wide variety of social skills. Some developers have pretty poor social skills and that falls on both gender sides. Maybe it's just easier when you don't have social skills to focus on the technology I don't know.

JG:  Do you think there is equal opportunity for women to participate?

CW:  I'm clearly an idealist because I do think that's the case but it's not reflective in the day to day reality. And I don't know where that starts. Is the barrier to entry at the Linux kernel development community or is the barrier to entry starting as you're educated in highschool and being lead a direction away from computer science? That part I really don't know. I'm sure a lot of people have done a lot studies to see where women end up in various stages of education and obviously, I never took computer science courses in college so it's not a prerequisite, but that concept I don't know if the lack of women in the Linux kernel development is due to the project itself, or the technical nature of the project and all of the things that would draw you in to a work like that, all of the social parts. Sometimes it's hard when you're a kid, I certainly know stories of my mom having some difficulty in understanding some particular problem in math and being told that she didn't need to bother because she was a girl and she wasn't going to need to know math. Hopefully things have progressed since the time she was in school but there's still a huge amount of that sexism going on. I really don't know how that plays out by the time you get to the Linux kernel development community, how many points of discrimination or redirections based on gender that people experience before they are even a viable member of the community, I really don't know. Certainly though the women that I have worked with are like the rest of us, they like computers, they like the technology, but they have not been able to be a more magnetic force for more women coming in to the field. If anything, the women that I know that have worked on the Linux kernel, more of them don't work on it now.

JG:  Right, and the number is falling off now actually. Not just in Linux, not just in open source, but across the board in computer science. And because open source already has so few women the men entering the field are clearly outpacing the women.

Do you think programming enjoys a more privileged position in Linux or FLOSS than other contributions?

CW:  You know that's going to depend to a degree by project to project. But at the heart, the concept of free and open source development is about the developers.

JG:  Meaning the programmers?

CW:  Meaning the programmers, so it's always a struggle for open source projects to build communities around testers, people writing documentation. Users who are actively involved in generating feedback, clearly there are millions of users in open source, but the users who are not developers but are actively giving feedback to the developers, those kinds of communities. I guess privileged would be one way to put it but I don't think of it that way as much as just the focus is around development.

JG:  The reason I ask is that how one frames open source development will affect what is seen as a contribution to the work and ultimately change how women are either reflected in that or excluded from that. If you have more women in documentation, in project management, in translation than women who work in hardcore programming, women will occupy a parallel position for contribution.

CW:  Right, in that context I think it's fair to say that programming is the cream of the crop for contributing to the project.

JG:  So then do you think one must be a programmer in order to contribute?

CW:  No definitely not but that's the struggle. It's not that those other communities don't have value, I think it's been difficult to promote that value. Programmers are terrible at documentation or understanding how users experience or what users experience is, they really need people to be working with them to document projects and get the feedback about how things are working, that kind of feedback is invaluable. But ultimately it's kind of this odd balance where the projects are about the developer developing software. The software does not exist in a vacuum, the software really only exits to serve its users. The users benefit the most from having software that has the features that they care about and having it be usable. Whether that means having intuitive interfaces or whether that means that it's well documented or translated to their local language all of those things are critical aspects. Developers are pretty focused on low level implementation details of the software and they're most excited about making subtle changes to the code base. But a user will never see that or be interested in that so it's bridging the gap between people who are really just eking out the subtle implementation details of the software project and people who are either using it or interested in helping direct it in some way or interested in participating in documentation or any number of ways you can get involved in the technical project without writing the code, those have historically been tough communities to build.

JG:  Yeah, the dialog I always have in my head is something around the imbalance of power inherent in a community like that. On the one hand open source can be so exciting and seductive, used for all of these disparate reasons and in all of these different communities but the problem I have is that at it's core it's only able to be manipulated and understood by this very small, rarefied group of mostly white men. So I'm very concerned about the power imbalance and think it's needs to be shifted somehow especially as Linux is clearly gaining momentum.

CW:  Right, I mean as a kernel developer I'd say we're sort of desperate for user feedback, desperate for use cases, and desperate for people to give us that insight on what we should be doing to improve it. But, we're still the people who are going to do the development on it. We're not really interested in driving it in the direction where nobody wants the project to go in, no user. There's also the difficulty in the high level technical project like the linux kernel -- there's always sort of this push and pull where we're trying to get user feedback but often the user feedback we get is not in a form that we can understand as requirements or changes that are relevant to the linux kernel. So if we get a request to provide feature "A", and we look at that and say that's crazy that should never be in the kernel, go away, basically. Teasing out some of the subtleties and understanding what the user is really asking for and figuring out what that might mean to a project like the linux kernel and where you would make changes is a valuable skill and useful part of participating in the conversation. It's otherwise difficult to get that interaction because user groups aren't talking in terms of the implementation details, they're talking in terms of the tangible pieces of the project that they interact with and how they could behave differently so there's almost like a language barrier. With people that are doing documentation or tracking bugs we're speaking in similar enough terms so that way the power dynamic is not just us and them.

JG:  How do you think we can bring more women in to open source?

CW:  That, I have no idea. It's a really difficult question. To me the reason it is difficult to answer is that I don't understand why we don't have them already apart from the normal situation that women represent a small percentage in science and technology in general. Why we have a disproportionately small number of women involved in the Linux kernel compared to that small, but maybe ten-fold larger amount of women involved in the proprietary development efforts, why that proportion is so much different in the Linux kernel, I don't understand it well enough to understand how we could change it.

JG:  Perhaps you're an anomaly here, but do you think this is something that is even thought about?

CW:  It is and I've certainly spoken with other male Linux developers about it. Some of the women Linux developers I've spoken to have struggled with it and have developed a wealth of communities that are specific to bringing more women into Linux development. So the men have sort of been on the peripheral where we know about the LinuxChix project and we know about women who are actively interested in improving the situation which maybe influences some of the conversations we have, but I've never seen anybody really have the right insight that's made a difference. I would say of the Linux kernel development community the contributor base has grown consistently over the last ten years and I would not be surpirsed if the percentage of growth for the male contributions is outpacing the percentage of the growth of the female contributions to the point, as you mentioned, that it will be tailing off and trailing away. Which at a certain level is totally depressing, at another level when you're just dealing with email addresses and you're not that emotionally involved with the people you are working with because a lot of them you don't even know, maybe that's a part of why people are not motivated to changing the situation. Or maybe it's just the stereotypical maladjusted male geek who feels more comfortable in a room full of men, I don't know. But I've met a lot of the male developers and the vast majority are surprisingly well adjusted people so something doesn't really resonate as to why we can't attract more women to kernel development. And I've always wondered if it's starting earlier than kernel development, if the paths are divergent from grades school, high school, college, if you're a women who goes into IT in college do you wind up being more interested in a different type of programming something other than what's needed for kernel development, I really don't know, it's sort of the only rational explanation I can come up with other than we're a bunch of pig-headed guys who can't do anything except offend the few women we have to the point where they'll never come back.

JG:  What advice would you give to a young woman interested in becoming a linux kernel developer?

CW:  I think most of my advice would be around technical advice, it would be that there are some useful things to study, these are the things that will help you understand -- if you understand how a computer works, if you understand these specific languages and how languages work, the C programming language, study some of the specific data structures around them -- in my mind those are the keys, that plus a passionate interest. For me those are the skills that you need, which again is why there is a disconnect because for me there is no logical reason why those skills are only available and interesting to 99% men and 1% women, that doesn't really make sense. Certainly I would encourage women who are interested in open source development to be prepared to deal with a sort of aggressive way of interacting. It's sort of an accepted way of being which I think is a little unusual in a professional environment. Usually in a professional environment you display a little bit more candor and respect for somebody else. Linux is actually a highly professional environment, most of the people doing it at this point are doing it professionally, more so than hobbists and academics, but the way they interact in the development community can be pretty abrasive which doesn't reflect the professional nature of most of the people doing the work.

JG:  Would your advice be the same for a man interested in Linux? You're not stating it that way because you think women are not or can not be aggressive, right?

CW:  Yes, I don't think I would make a distinction in gender for advice. But I don't know what I would say to the person interested in how we can draw more women in, it's kind of perplexing, I don't know what we can do to draw more women in. The reason I focus on the aggressive piece is that in this context it's part of the culture. Some of the women I've spoken to have found that style very offensive more so than some of the men, enough so that they're not going to want to participate. The key though is that you're excited and engaged and enthusiastic about the work that you are doing to a point that you can sort of put up with it. But if social interactions are going to be disturbing you in the wrong way that's going to overshadow any of your technical interests in the project. You can only interact with the code for so long, eventually you're going to have to engage with the community and if you don't like that part of interaction you're not going to belong to that community that's all there is to it. It's an area where we focus in general, trying to bring new developers in and make the environment a little bit more hospitable to new comers. In part that's because people have complained that it's difficult to get involved because it's abrasive and unpleasant and in part because we knew we had really valuable contributions coming from whole cultures that felt very uncomfortable participating in our forums, primarily Asian cultures. So we've made some conscious efforts to tone down that abrasiveness as a way to try to eliminate that barrier to entry to new developers because ultimately Linux as a project is only going to be successful if we bring in new developers, when it's the same old people we'll either lose interest or get old and fade away or whatever, I mean the project will die if we don't keep bringing in new people and there's just no logical reason that a healthy portion of those new developers wouldn't be women.

JG:  What excites you the most about technology, what keeps you in it?

CW:  It depends on the day. What excites me most at the core is just the technology. I'm a geek, I like to play with things and learn how they work and there's always something new I can learn working on Linux. Within a particular area sometimes I can get bored but the Linux kernel is a huge project and there are a lot of different areas to work on but also I enjoy the people. It's really thrilling to be involved in a collaborative effort, there's a rush that you get from bouncing ideas off of people and listening to their ideas and creating something new out of that collaboration, that part is really hard to beat. It's also not what I've been doing in my day to day job for maybe the past two years so some of the things that I end up doing now are interesting in a whole new set of ways. The fundamental tenet in all of these things I find interesting is doing something new, and something you can learn and a way to grow and keep yourself fresh and invigorated by that learning process. For open source development specifically it's all about improving the code, making it more elegant and collaborating with other people to do that, it's a really cool experience, it's so cool that it's hard to believe people would not just be flocking to the project like crazy wanting to be involved. We do get a lot of people who are interested at first pass, but a lot of people are short term contributors. They're trained, they think it's cool, then for whatever reason they realize that to really participate would mean dedicating a certain amount of time to it that they're not prepared to dedicate to, or they just had a fleeting interest, or they get involved and find out that they don't really care for the community, there's a whole variety of reasons, that's healthy that we have this influx of people coming and going, the goal would be to make sure that a growing percentage of people who stay are new comers, otherwise we'll lose our inertia.

JG:  What might be next for you?

CW:  In a certain real sense, more of the same. I haven't done as much work in the community directly in the past year, so in the near term future it would be nice to be able to spend more time doing that. In the long term I enjoy guiding people and bringing them into a project. So in a sense it's growing more in a leadership role and less as a technical participant. It's hard to let go of the technology, it's really a fun part, on the flip side you can have greater impact by bringing people in, so it's a tough one, something I wrestle with. I don't want to come too dangerously close to being a people manager, I prefer the technology side.


Return to top