2022-08-10 Maintainers Meeting notes


Aug 10, 2022


  • @Edward Zarecor

  • @Ned Batchelder

  • @Sarina Canelake

  • @Colin Brash

  • @Jeremy Bowman

  • @Braden MacDonald

  • @Felipe Montoya

  • @Azan Bin Zahid (Deactivated)

  • @Waheed Ahmed (Deactivated)

  • @Feanil Patel

  • @Piotr Surowiec

  • @Umar Ali

  • @Maria Grimaldi


  • Understand the purpose of the maintainers program and the pilot

  • Tools we have

  • Agree to a plan for adopting OEP-55 recommendations for Open edX maintenance

    • Proposal for phase 1

      • Consistent READMEs

      • catalog-info.yaml file exists with maintainers specified

      • Consistent application of repo standards

        • Requirements bot is running

        • Maintainers are getting PR messages / are taggable

        • Auto-deletion of branches is enabled

        • etc.

      • Reachability of maintainers – discussion with proposal

        • Issues on in all repos

        • Maintainers are monitoring – what’s possible?

          • Issues

          • Discourse

          • Slack

 Discussion topics












Overview of current edX/2U ownership tracking

  • Arch-BOM owns this at 2U

  • edX has spreadsheets to track, with tooling to update

  • edX tracks end-to-end, including code maintenance to operational monitoring

    • OEP-55 can’t cover all of that


What do we want to include in the pilot?



Overview of existing tooling


  • Grimoire

  • Backstage

  • catalog-info.yaml


Linter/checker for rules?


  • 2U has “repo-health dashboard”

  • We could use a compare/contrast of the tooling

    • Grimoire vs backstage vs etc.


Intersection of maintainership and core repos?


  • Can we filter out non-core to reduce noise?

  • Maybe not for pilot, but part of next phase?

  • The pilot includes deprecated repos (ecommerce and friends)

    • These are currently “core” but would need a way to be marked “non-core” as part of deprecation.

  • Can we filter for pilot repos?


Things to get done for the pilot



  • Export of repo-health data for pilot (or public) repos

    • Group to look at data to see what is valuable


Maintainer reachability


  • What channels should maintainers monitor?

    • Issues / Discourse / Slack

  • Slack integration

  • Slack is for coordination/collaboration

  • Discourse is for:

    • getting help

    • conversations that last and can be found.

  • How do we solve the problem of poorly categorized questions?

    • Few people will properly identify the repo at fault

    • Feanil is looking at Discourse categorization

      • Pilot will include improved Discourse->Slack integration

      • Subscribe Slack channels to subsets of Discourse content

      • …and maybe the other direction also.

    • Categories for coarse-grained components?

      • Certificates, commerce, etc?

      • More categories might lead to more incorrect categorizations

  • Clearly state what we are watching

  • More community people helping with “level 1 support”?









 Action items

Catalog and provide guidance for common tasks with our current tools
Create a version of the repo health dashboard for public pilot consumption
Propose a list of course grained categories for discourse.