2014/Generational Relay: Passing the Open Source Torch
People leave Open Source projects, and that’s ok. Failing to plan for it isn’t. How one community is recovering from the loss of its first generation and preparing for the rise of its third.
Speaker: Eric Steele
Return to this session's details
Contributed notes(Add your notes here!)
Started with quote from "Learning from Diaspora".
Plone community lost a bunch of leaders and contributors at once - scary - but it's possible to go through this process with positive change.
Context from Plone community - it does lots of in-person sprints - "stare at your laptops in exotic places". Such as the castle of an Austrian prince who contributes to Plone (Castle Goldegg). In-person meetups and friendships between Plone contributors are important to Plone.
Did research on what has been written about community change before: Open Advice book. And read research papers about contributor models and measuring knowledge loss. Learned that developers can take 30 months to "spin up"; measure developer populations by "half-life" (Debian - 7.5 years).
OK, the crisis - "everyone is giving up, we're doomed!" "who's in charge here?" "oh crap, it's us." Look at why everyone is leaving - software sucks? people suck? boredom? economic shifts? (For Plone - economic shifts were important.) Types of leaving: "the denial", "the tumbleweed", "the rage quit".
The first thing to do: put a name to it. He wrote an inspirational call-to-action blog post ("Plone: the Second Decade") - pointing out that people were still there - still the same amount of talent, but spread over more people. Shifting from a few stars to a broader set of contributors, more diversity - pointing out the benefits of that.
People feeling guilty about not contributing = not constructive or helpful, caused worse results. Had to remove the stigma. "Thanks for helping" instead of "where have you been".
Plone people wrote a manifesto together about their new goals - transparency, community building, spending money on things that work (like development sprints).
Looked at roadmap and trimmed pet projects from people who had stopped contributing. Watch for rising contributors - Ohloh was helpful as a start. Say thank you.
Trolling constructively - bring up controversial ideas to goad contributors into excitement about the project. Show activity. Minor, related annoyances get brought up, fixed.
Spent lots of money on sprints! $25000/year. Work on strategically-important projects. Bring along knowledge-holders. Get promising new contributors to face-to-face interaction with more senior ones. Evangelism in geographical areas of growth, far-flung communities that don't get much contact with the bulk of the community.
Make new stories and document your stories. Like why Amsterdam-Schipol has a sign reminding people that katanas can't be brought onto planes.
Shifting title of prestige from "core developer" to "core contributor" - to more correctly emphasize people who do work like documentation and especially communication.
Mentioned research papers: