Hang around the free and open source software community for any length of time, and you can’t help seeing examples of burnout. A colleague takes on too much, and suddenly they’re working harder for fewer results.
They have a hard time concentrating on their work. They neglect their personal life. When challenged, they become defense and unusually hostile. Eventually, they withdraw — and, sometimes, they don’t come back.
Burnout isn’t unique to the Linux community, of course. However, at times, the problem can seem almost epidemic in the community, and people seem reluctant to talk about it publicly.
Both Ubuntu community manager Jono Bacon and Ubuntu volunteer and journalist Amber Graner find that, when they deliver talks based on Herbert Freudenberger and Gail North’s The Burnout Cycle, afterwards people approach them privately to talk about their own experiences with burnout.
Similarly, kernel hacker and Ada Initiative co-founder Valerie Aurora talks about sitting with a dozen women and technology activists and discovering that all of them were burned out, recovering from burnout, or else had been burned out in the past.
No one seems immune, not even Linus Torvalds. Although he begins by saying, “I’ve never really had a burnout event,” he goes on to recollect a situation that sounds very much like the early stages of burnout:
“We had really big fights back in 2002 or so (“Linus doesn’t scale“) where I was dropping patches left and right, and things really weren’t working. It was very painful for everybody, and very much for me, too. Nobody really likes criticism, and there was a lot of flaming going around — and because it wasn’t a strictly technical problem, you couldn’t point to a patch and say, ‘hey, look, that patch improves timings by 15%’ or anything like that: there was no technical solution. The solution ended up being better tools, and a work flow that allowed much more distributed management.”
What causes burnout, particularly in the free software community? What can be done to prevent it, both individually and on a community level? These are questions that an increasing number of free software organizers are struggling to answer as burnout starts to be recognized as a problem that needs to be addressed.
Sources of Burnout
The organization of free software makes community members especially prone to stress. As Bacon points out, when contributors are scattered around the globe and some are volunteers, each has to monitor their work-levels themselves.
Yet when someone somewhere is likely to be working on a project at any hour, setting limits is difficult. As with a real-time simulation game, there is no obvious moment to quit. In fact, because of the instant responses that are the norm on the Internet, others may become annoyed when others are not immediately available.
The stress may be increased because the first generations of community members are now well into middle-age, and some are starting to have trouble working the hours to which they are accustomed, either because of reduced stamina or family obligations.
Graner points out, too, that some community members may add to their stress by taking on more work to prove themselves. She observes, for example, that non-developers in Ubuntu may feel less part of the project than developers, or take on extra responsibilities in the hopes of having their expenses paid so that they can attend the Ubuntu Developer Summit.
“They think that if they don’t take on more and more and be that super community person, people won’t think they’re doing enough,” Graner says.
However, as Torvalds points out, burnout is not just about stress. “I personally tend to get really fired up and love the occasional flame-fest,” he says. “That can be stressful, but it’s also often really invigorating, and I think that if you don’t have those kinds of occasional flare-ups, your project is dying, or just don’t care enough.”
However, Torvalds adds, “But constant stress can just wear you down. For me, it’s pretty much always been about some work flow issue, where just the way I do something doesn’t really work well any more, and the stress is about just not having enough energy (or hours in the day) to do what I need to be doing. So that is why, for the kernel, I feel like the big stressful events have been about work flow issues.”
Bacon perceives burnout in much the same way, defining it as “when you carry over stress from the previous day, and it builds until you can’t get rid of it.”
By contrast, former Fedora chair and community organizer Paul Frields sees burnout as originating in group interactions, as occurring:
“When people have mismatched sets of expectations. For example, maybe it’s an expectation that other team members deliver the same level of output regardless of their capability, time, or personal situation. Or maybe the expectation is for everyone to love your radical new idea right off the bat. If those expectations aren’t met, and you continue to grind away on it, chances are excellent that you’re going to burn out.”
Still another source of burnout for women in particular is their under-representation in the community. Depending on the project, women typically make up one to five percent of the community. Not only do women have to endure sexist remarks, pornographic presentations, and outright hostility, but they often feel the need to prove themselves — often, to established women as much as the male majority.
“It’s similar to being in the military,” says Graner, a veteran from the first Gulf War. “You need to do ten percent more than everyone else to be seen as good as they are.”
For women actively trying to alter the culture, the stress is even greater. “There are simply too few women in open source to do all the work,” Aurora says. “A one percent community that’s already struggling is just a recipe for burnout. You’re already in a precarious situation where you’re getting a lot of messages saying that you don’t belong, and you’re adding on top of that hours of volunteerism for activism. You feel bad for not doing the programming, and for having doubts at all.”
Nor is the situation improved by the fact that, until recently, one woman has tended to be the figurehead for women’s activism at any one time. “You become the lightning rod for criticism and death threats,” Aurora says. “This is a huge cost. Every time someone becomes a leader for women in open source, their career suffers.”
To further complicate matters for men and women alike, burnout is a condition that everybody has trouble seeing or admitting to. “They can see the signs in everyone else, but can’t see the reality in their own reflection looking back at them,” Graner says. “And sometimes they believe that to use the term ‘burnout’ is negative — like they can’t tell anyone they’ve burned out.”
Such denial is especially common when those in leadership roles burn out, either because they see themselves as essential or are more used to giving help than being in need of help themselves. But, in all cases, this denial only aggravates the situation by making people more reluctant to take the steps they need to recover.
Dealing with Burnout
On a personal level, Torvalds suggests, the key to recovering from burning out is to:
“Learn to let go. If not of the whole project, then at least letting go of trying to control it entirely. I’ve done both. With the kernel, I may be the to-level maintainer, but I simply trust others to do the right thing. There are still a few areas that I’m fairly closely involved with, but even there I’m more than happy to be overridden by people I trust.
Or, just let go of the project entirely. I did that with git [The distributed version control system]: I really enjoyed doing it, but I also felt like I couldn’t really afford to be the full-time maintainer that the project needed, and I was more than happy to find a great maintainer (Junio Hamano). I really felt it was my project, but, at the same time, I also felt that the best thing for it would be to have somebody else maintain it.”
Of course, as Torvalds adds, “Sometimes people seem to have problems with letting go. Me included.”
To counter the natural resistance to letting go, Frields suggests that “you need to be willing to engage in self-examination, to check your balance and your ability to give time to the things that fulfill you. And, more than being willing, you have to consciously take the time to do it.”
Bacon is even more specific. Based on his own experience with burnout about a year after he joined Canonical, he has given considerable thought about how to create a balanced life that will might make himself and others more resistant to burnout.
Not being single is one of the surest safeguards against burnout for Bacon, but he observes that even singles can take an evening to get away from the community and enjoy the company of friends. He suggests, too, developing other interests (for Bacon, one such interest is playing music with his band Severed Fifth), regular exercise, and a healthier, low calorie diet.
Part of that diet, he suggests, is a reduced level of caffeine, which many members of the community are literally addicted to; Bacon himself describes withdrawal from his six cans of Coke a night, with all the vomiting and shaking it involved as “one of the most wretched experiences of my life” and lists limiting caffeine among the changes he made in his life to reduce the chances of burnout.
But burnout can also be controlled on a community level by creating a culture in which people believe, in Graner’s words, that “If you’re not operating at one hundred percent, then you’re letting the team down,” and that regular breaks from work are encouraged.
Graner further suggests that free software work, “needs to be a team effort so no one person is responsible for it all.” Based on her experience in the army, she advocates everyone in a project learning each other’s job. Such rotation has the advantage of reducing the tendency for anyone to think of themselves as indispensable, and provides variety that can help to lessen any feeling of burnout. It also means that, should anyone burn out, other project members can take over their responsibilities with a minimum of adjustment.
Another suggestion that Graner makes is that project roles be clearly defined — something that rarely happens in a distributed project staffed partly by volunteers. That way, people might be less apt to take on new responsibilities.
To these suggestions, Aurora adds that burnout can also be alleviated by “personal expressions of support from multiple people — sending that email that says ‘I think you’re doing a really great job and you’re right’ makes a big difference.” In fact, Aurora explains that “any form of validation” can help:
“It seem trivial, but part of any form of burnout is feeling that what you’re doing is trivial and not appreciated. I think that the Internet is good at sending the feedback that what you’re doing isn’t appreciated. Apparently, bad people enjoy sending flames more than nice people enjoy sending thank-yous, and the removal of human faces and voice tones mean that misunderstandings are common.”
Aurora cites her own experience of having her work developing anti-harassment policies for conferences, undertaken at a time of near-burnout being met with so many encouraging emails that she could “almost cry with happiness.” The encouragement was a much-need validation at a crucial time.
All these responses to burnout can be enhanced by community leaders. Bacon suggests that a manager with “a day to day sense of engagement with [their] community” is in a better position to notice signs of burnout than anyone.
Bacon suggests a frank but supportive talk about burnout, and whether it requires a leave of absence or a reduction in responsibilities. This talk should be in person if possible, by phone if not, and never via email or chat, whose lack of nonverbal cues can create misunderstandings, especially to someone who is already feeling inadequate about their work. In this talk, the community leader should make clear that they are not reprimanding, but giving feedback and suggestions that are in everybody’s best interest.
If necessary, the perception of the talk can be softened by encouraging the person with whom they are talking to perform the same kind of intervention if the community leader ever shows signs of burnout.
Staging a Comeback
Just as burnout has no single cause, so no single remedy can prevent or cure it. That means that whether victims of burnout can overcome their problems or will simply disappear from the community is hard to predict. The prognosis is especially hard because the community is just starting to discuss burnout and how to prevent it. Nor should anyone expect a quick recovery.
Recalling her own burnout, Graner says, “The burnout didn’t happen overnight, and neither will the comeback. There were times when I thought that if I didn’t come back and do all the same things that I had been doing, I would have failed. But nobody thought that except me. That was me blaming me. I had to tell myself, ‘No, you’re not failing. You’re being responsible now.'”
The good news is that those who return from burnout frequently have a greater awareness of what went wrong and how to prevent burnout form happening again. “Once you have fully burned out, you can see when it’s coming back,” Graner says. Such awareness, more than anything else, may be ultimately the strongest single weapon against burnout as increasing numbers of people return to put what they have learned from their experience to practical use.