Volunteer coordination and outreach/ECT July 2013 quarterly review

From mediawiki.org

Agenda for the Engineering Community Team's review of Q2 (March through June 2013). Notes from the meeting are on the talk page.

What we've been doing[edit]

Nurturing volunteers[edit]

  • Mentorship and gardening
    • We got ~20 OPW/GSoC interns (compare to 9 last summer), ~30% of them women or genderqueer (compare to: 0)
    • We're up from 11 to 13 non-WMF maintainers of MediaWiki core; other stats are in flux
    • We've helped in shepherding & discussion around controversial contributors
    • We met with Analytics, Ops, and UX to lay groundwork for volunteer onramp in next year
    • We developed information architecture proposal re new tech contributors, decided to park it temporarily in favor of analytics + concentrating on mentorship efforts in the near future
    • We connected volunteers with grant opportunities (IEG/Participation Support)
  • QA
    • We ramped up training, mentorship, list, events
    • The current trainee is working on automated testing for VE, which is a good way to have an apprentice work on a high-priority project
  • Events
    • We moved from the consistency of "the wheel" to a more opportunistic & flexible approach
    • We co-organized the Amsterdam hackathon: 150 participants, great survey results

Tech communications[edit]


  • Andre made incremental code improvements -- in the Bugzilla front page, weekly report email, etc.
  • Andre led social improvements: clarified admin policy, blogs BZ tips, runs office hours, meets with dev teams

Lessons learned[edit]

  • LevelUp was useful for helping people discover what they wanted
    • But without constant metamentorship from ECT, most people disengage because other things are more urgent
    • We'll stick to OPW/GSoC which is clearly under our remit, and leave professional development stuff to contributors & their managers
    • Code review backlog is less urgent than previously (Subversion era backlog blocked release)
  • Volunteer product management/advising
    • LCA is innovating with the Community Liaisons and showing very interesting results
    • It's easier to funnel product management volunteers into community liaising than into feature backlog work
    • ECT (Sumana) has not focused on metamentoring the product management volunteers who do show up, and will start feeding them to Maggie
  • Mentorship programs
    • More gatekeeping and task preparation and more frequent communication lead to better outcomes
    • Running OPW correlates to more women ending up applying for GSoC
  • Lack of awareness
    • Nearly no volunteers know about the grants
    • Key tech contributors don't read parts of the monthly report that affect them


Key goals for the next quarter:

  • We teach several (6) volunteers how to contribute to and fix automated tests, and they do so regularly, improving our capacity for faster deployments.
  • To improve awareness of Wikimedia's engineering needs, we systematically reach out to FLOSS projects we rely on, and grow Tech Ambassadors membership.

Also in the next several months: Starting community metrics, and lending Guillaume to VE communications for July.

Plus our ongoing activities: summer mentorship, tech communications, bug management, and volunteer outreach & events.


We need your help to best serve the movement. Questions in bold are the most important; we need your answers for the coming quarter.

We want to work well with LCA/Community liaisons.

  • Which projects will be getting liaisons after VE push ends? Should ECT pick up the slack on others?
  • Are we duplicating effort at all?
  • Can we ask the liaisons to help us grow tech ambassadors membership by recommending it to likely community members?
    • A - Yes, and they are doing it (lightly) now. I will further emphasize it, however. - Philippe

We want to help increase our deployment frequency.

  • Is automated testing the best way to help?
  • Is targeting very popular gadgets & bots like HotCat, Twinkle, and SineBot a good idea? What about user scripts?

We're identifying which upstream open source projects we ought to be making closer relationships with.

  • What should we prioritize?
  • Is there already an inventory?

We aim to help more Toolserver tools move to Tool Labs (including a doc sprint in late July.)

  • What should the specific goal be for December? Proportion of bots/tools moved to Labs?
  • Are there WMF internal tools that we need to move ASAP?