The Bait and Switch of Open Source

Katrina Owen

Description
Open Source sells itself as being about technical problems--delightfully thorny technical problems, at that. However, successful projects are filled with people, which introduces a whole different set of problems.

This talk tells a story about the many ways in which things went badly wrong because I didn’t treat people problems as first-class citizens in exercism.io.

What’s the take-away for the audience?
If you’ve never run a open source project before, this talk will let you approach your first big project knowing what you’re in for, and give you ideas for how to avoid some of the pain and pitfalls. If you’re running your own project, you can laugh and cry and commiserate, and get some ideas for tackling these types of problems strategically.

Abstract
The story of Exercism is one of chaos, confusion, and surprising successes. The site started out as a workflow optimization tool intended to be used by 25 people, but within a short amount of time, thousands of people were using the site, and hundreds of people were contributing to it. Which sounds wonderful, except that it wasn’t really clear who all of these people were, why they were there, and what they needed or wanted.

The questions that had me stumped were not about automated tests, continuous deployment, or architecture, but rather about mentorship, motivation, and communication.

  • How do you provide a user experience that fosters friendly and thoughtful conversations?
  • How do you model effective code reviews when the recipients have wildly differing skill levels (and need wildly different types of feedback)?
  • How do you encourage desired behaviors without killing intrinsic motivation?
  • How do you reject pull requests and feature suggestions without hurting people’s feelings?
  • How do you onboard new contributors, many of whom have never contributed to open source, and are completely overwhelmed and intimidated by the process?

I thought I was solving technical problems, but the truly difficult problems were all inherently squishy.

Edit proposal

Submissions

OSCON 2016 - Accepted [Edit]

Add submission