Yesterday I spoke at an OpenHatch Open Source Comes to Campus event on the other side of the country. The organizers set up a video conference and asked me, and two other people who work in open source, to talk about our careers. (If you are running an event like this, or teach a class or something, I'd probably be happy to do this for you as well.) Some things we mentioned:

 

• The three of us get paid pretty darn well. Have I mentioned that I just took a three-month sabbatical? And that I support my spouse, who writes scifi?
• Companies that make proprietary software often frustratingly limit what their employees can do. I can blog, speak at conferences, swap useful information, and so on without asking lawyers for every little thing. If you haven't dealt with the chilling effects of onerous nondisclosure agreements before, you may underestimate how annoying they are.
• Many processes we reflexively follow in open source are just good ideas regardless. If you're making software with other people, you'll want to systematically keep track of your shared TODO list (with a bug tracker), track who made what changes to the code and have easy rollback (source control), and talk in one central place with other people to make and document decisions (mailing list or similar). Some closed-source firms don't follow these best practices, and that boggles me and often lands them in The Daily WTF. But open source orgs generally get this right.
• Do not underestimate the value of meeting people face-to-face. The LWN calendar, Lanyrd, and Meetup can help you find these events. And there's often funding available if you need help getting to conferences, either formally (e.g., WisCon, PyCon, and Wikimania), or informally; try asking.
• If you are doing anything even remotely related to Python, try to get to PyCon, because the people are friendly and the sprints are invaluable. [This of course said by someone other than me, as this year will be my first PyCon!]

By the way, I should also link to this one-hour video on the realities of open source careers. Cynical in some ways, and I particularly disagree with the speaker on the matter of references; you should not assume that the hiring manager isn't going to directly call the references you provide, and ask them interesting questions. And the emphasis on unpaid work can go awry, and I started rolling my eyes at the oversimplifications halfway through (e.g., the idea that an employment contract means literally nothing). But the talk also has some truths that students don't hear often enough.