Governance Office Hours Meeting Minutes [archives up to 2022 April 22]

Meta Governance
61

We blew past our timeline with the holidays, but the idea is for Fernando and Brian to present a draft for deliberation soon. I don’t think we re-calibrated the timeline, though.

1 Like
62

The weekly hourly office hours call is open to the community members who care about governance issues. The call is held 9-10AM PST on Tuesdays.

Meeting Minutes – February 4, 2020

  • Attendees
    • Fernando
    • Darian
    • Tim G.
    • Brian
    • Luciano
    • Jason
  • Quick update from Jason about the conference committee
  • Brought up discussion of, “What is the failure mode to cover crises that arise from deep breaks of trust?”
  • Continued discussion of the organizational and executive bodies (primarily in Fernando and Brian’s brainstorming document)
63

The weekly hourly office hours call is open to the community members who care about governance issues. The call is held 9-10AM PST on Tuesdays.

Meeting Minutes – February 11, 2020

  • Attendees
    • Brian
    • Darian
    • Tim G.
  • Brian updated about how he and Fernando are thinking about governing the smaller sub-teams in the project.
    • As small teams that organize themselves grow, they devolve into the tyranny of structurelessness, so not defining how that governance works needs to be defined.
    • Otherwise, it would just be punting the governance problem down to the sub-groups.
    • The sub-groups should have authority by default unless otherwise stated by the main governance model.
  • Questions the group governance model has to answer
    • How do they make decisions?
    • What decisions are in scope?
    • Q: How do they elect their delegate/rep to the executive body?
      A: The delegates for sub-groups are selected by electors from each sub-group. A list of electors (a voter roll) should publicly exist. These people vote on who represents their project as a delegate. For software projects, this pretty much correlates with the list of committers. With non-software sub-groups. the analogue still holds.
    • How do they report?
  • The executive body (E body) should be proportional to the number of voters per sub-group. A huge sub-group should have more representation in the E body compared to a small fledgeling group.
1 Like
64

The weekly hourly office hours call is open to the community members who care about governance issues. The call is held 9-10AM PST on Tuesdays.

There will be no call on February 25. The next call will be on March 3.

Meeting Minutes – February 18, 2020

  • Attendees
    • Brian
    • Fernando
    • Tim G.
    • Luciano
    • Darian
    • Zach
  • One concern in the new model is to avoid a large number of representatives from which a smaller set is chosen to make the hard decisions and move the project forward. Current leaders have been in that position and don’t want to be, and also don’t want to ask anybody else to be.
    • Even in a large representative body, there emerges leaders, implicitly or explicitly.
    • The governance system can’t fix a wide breach of trust and good faith. At best we can govern in places where trust exists and people can continue to act in good-faith even when the project doesn’t press in their direction for a time.
  • A key difference between us and something like a nation-type government: a key ouptut from Jupyter is concrete products, that can adapt to external conditions, markets and needs, and that require a clear vision. The creation of effective products of this kind may not necessarily be what pure group-based decision making requires.
    • Legitimacy of decision-making can be enhanced by a clear process that constrains unhappiness about specific decisions.
    • Innovation can come out of a strong infrastructure decision that is made and invested into, e.g… JupyterLab was a strategic product decision, not based on full community consensus.
      • Risk is another factor in innovation. Open-Source distributes that risk across industry stakeholders that may not be able to foot the risk individually.
    • There’s a balance between having a strong central government and representing the voice of the individual. Examples given in some democratic systems the individual has a lot of power to veto things that could be good for the larger population. An opposite example is an authoritarian system, where the individual has no power, great things can be done, but at the cost at the cost of the individual.
  • Thoughts on quorum: if the quorum requirement is very low, that might encourage participation b/c there’s clarity that if only a few people show up, they will actually be empowered to make decisions.
    • Whatever the mechanism is, there has to be something in place to keep velocity.
  • The overall two-body model seems OK as a starting point, what needs fleshing out to have a first working draft? A few things…
    • Emeritus status and/or quorum tricks to balance participation and velocity?
    • Growth of representative body, proportionality with team sizes?
    • How to ensure decision making doesn’t stall?
      • Perhaps we don’t have to ensure that, rather just create the right incentives.
65

Last minute update: there will be no call tomorrow, March 3 as Brian and Fernando will be using the time for some heads down work. The next call will be on March 10.

1 Like
66

The weekly hourly office hours call is open to the community members who care about governance issues. The call is held 9-10AM PST on Tuesdays.

Meeting Minutes – March 10, 2020

  • Attendees
    • Brian
    • Tim G.
    • Darian
    • Chris H.
    • Jason
    • Luciano
  • We read through a draft of the governance proposal that Fernando and Brian have been working on and added some comments in-line as well as discussing in person.
    • The Organizational Body for institutional stakeholders
    • The Executive Body is the decision-making body for individuals
    • The Distinguished Contributors population exists purely for recognition
  • What is the “strategic planning” project area? It is the sort of implicit power work that goes on with members of the project who already have authority. So we need to attempt to capture this explicitly instead.
  • There is one delegate per subproject or working group in the Executive Body, we discussed the trade-offs between proportional representation vs. having a “Senate” model.
    • A purely formulaic approach that calculates delegate count as a proportion of contributors was considered but contributor count ≠ importance, necessarily.
  • What is the role of JEPs?
    • The work done in the past year on codifying JEPs as part of our process does not need to be reinvented.
    • The Executive Body would be the decision maker on JEPs.
  • In all bodies a consensus seeking mode would be used, with a low quorum during the voting phase.
  • We need a decision making process that can trickle down to individual sub-projects regardless of whether they are 3 people or 20+.
  • If all sub-project decisions are consensus driven and allow for an objector to call a vote, does that slow down rapid-fire pull request iterations?
    • One thing that needs to be possible is a post-merge vote, especially if we want low-friction pull requests.
    • This model should allow for anyone to call a vote post-merge.
    • Each sub-project has a list of voters (for example voters may be comitters, but this is not required).
  • Sub-project governance should basically be modeled on the larger project’s governance, with certain variables being modifiable (e.g., how big is quorum for this sub-project, voting periods)
    • Voting/reporting should be standardized and automated across projects
67

The weekly hourly office hours call is open to the community members who care about governance issues. The call is held 9-10AM PST on Tuesdays.

Meeting Minutes – March 17, 2020

  • Attendees
    • Fernando
    • Darian
    • Tim G.
    • Luciano
    • Jason
    • Sylvain
  • Discussed question of proportional representation on the executive body vs fixed at one-per-project, which creates issues with the relative sizes of different projects.
    • Orgs vs Projects vs Sub-projects
      • eg. JupyterLab has 39 repos within it. Are we looking at 50+ people on the executive body?
        • No, GitHub orgs are the level of ‘project’ and repos are ‘sub-projects’. Only projects get representatives on the executive body.
  • Question on how to encode top-level leadership and vision with a more explicit model?
    • What replaces the BDFL part of our current governance? The strategic vision responsibility does not yet seem encoded in the new model.
  • Idea: a special, small (3-person) team in the Executive Body that is flagged for strategic leadership/vision that encodes what Fernando/Brian and others have done informally for years with grants/etc.
  • The existing steering council covers a variety of areas, but isn’t successful in all the areas it’s responsible for right now. We talked through some of the implications of what is being done by the steering council now that will be broken down into other groups.
    • Strategic Vision ⇒ Strategic committee of Executive body (max 3 people)
    • Technical/project level decisions ⇒ Executive Body
    • Institutional Partners ⇒ Organization
    • Prolific contributors ⇒ Distinguished Contributors
68

Is it possible to read through the draft? I have standing meetings that prevent attending the Governance Office Hours.

69

Hi Carol! I hope you are well!

The document is shared in the Jupyter Governance group, of which you’re a member. If you open up Google docs and search for: Draft Governance Proposal it should be accessible.

1 Like
70

Thanks Darian for the response and reminder. I had completely forgotten.

I hope you are doing well too. Stay safe.

1 Like
71

Thanks, Carol! You too!

72

The weekly hourly office hours call is open to the community members who care about governance issues. The call is held 9-10AM PST on Tuesdays.

Meeting Minutes – March 24, 2020

  • Attendees
    • Tim G.
    • Jason
    • Darian
  • Discussed what the composition and election to the Distinguished Contributors section of the governance model should be:
    • The primary questions are who selects and how
      • We should add members once annually.
        • If we cap a total number per year we keep some value on this
      • Tagging on some ‘cost’ to honoring people adds a natural barrier to adding too many people.
    • We can unbound for the first year, to get important contributors included.
      • The first raft of people should be the existing SC members.
    • These members should be nominated by an existing member, vetted by a subcommittee of the distinguished contributors, and elected by the whole group once annually.
  • Executive Body section of the document should link out to a list of all projects.
73

The weekly hourly office hours call is open to the community members who care about governance issues. The call is held 9-10AM PST on Tuesdays.

Meeting Minutes – March 31, 2020

  • Attendees
    • Brian
    • Tim G.
    • Chris H.
    • Luciano
    • Darian
  • We spent the meeting looking at drafts of the Distinguished Contributors section of the model and the Executive Body section.
  • It was basically a working call to edit and not a typical weekly call.
75

The weekly hourly office hours call is open to the community members who care about governance issues. The call is held 9-10AM PST on Tuesdays.

Meeting Minutes – April 7, 2020

  • Attendees
    • Brian
    • Tim G.
    • Chris H.
    • Darian
    • Jason
  • We looked at the Distinguished Contributor draft:
    • The minimum contribution for a Distinguished Contributor should be three years.
    • A nomination should be accompanied with a CV/biography of the nominee written by the DC who nominates.
    • All DCs should have a bio that is publicly accessible and maintained by the DC (preferably, but not codified in the governance)
    • All requirements for being a member ought to be met at the time of nomination (as opposed to at the time of the election)
  • We looked at the Executive Body draft:
    • This body might better be described as a Technical Steering Council
  • The “vision and leadership” gap in the governance model needs more discussion
76

The weekly hourly office hours call is open to the community members who care about governance issues. The call is held 9-10AM PST on Tuesdays.

Meeting Minutes – April 21, 2020

  • Attendees
    • Brian
    • Tim G.
    • Luciano
    • Darian
  • We read a draft of the “Board of Directors” (BoD, formerly Organizational Body) document.
    • A new concept of a hired Organizational Director that is managed by the Board
    • How are the members of the Board of Directors chosen?
    • How much time does a member of the Board of Directors need to put in?
      • Order of magnitude: 10 hours per week
    • There should be specific classes of directors drawn from different stakeholder constituencies (e.g., institutions, companies, subprojects, community)
    • BoD may send money to a project that has a JEP approved by the TSC.
    • How to bootstrap the initial set of BoD’s depends on how we elect the BoD
      • If we have open elections, we can just do that.
      • If we work off a mentorship model we’ll need to think about how to seed that group.
  • Conflict of interest and ethics rules need to be enshrined in the governance document
    • case study: analytics trackers on our web properties being debated by the BoD
77

The weekly hourly office hours call is open to the community members who care about governance issues. The call is held 9-10AM PST on Tuesdays.

Meeting Minutes – April 28, 2020

  • Attendees
    • Brian
    • Tim G.
    • Luciano
    • Darian
    • Chris H.
    • Jason
  • Briefly discussed the JEP process to ‘sunset’ the notebook.
  • A new focus for the “Organizational Body” - acts more like a “board of directors” (BoD)
    • Q: what would be the relationship between the BOD and the sub-projects?
      • A: there would be a lot of autonomy for the sub-projects, any decisions that are made within the sub-project organizations follow their own governance models
      • For decisions that exist outside of any one particular github organization, that decision needs to be made with a JEP, and the final decision will be from the technical steering committee (TSC)
      • This is close to what we already do with Jupyter - a difference is that the new steering committee will be made up of people from each Jupyter sub-project
      • Subprojects can petition for funding and resources
    • Suppose “voila” wasn’t in the Jupyter org, but wanted to be
      • Board of Directors would need to do this along with TSC
    • Board of directors would run organizational parts of Jupyter
      • “They create and delegate”
      • Create groups for fundraising
      • Rally support from sub-projects for strategic goals
      • Can try to set an agenda and a vision, but cannot impose that agenda themselves
    • How do the board of directors execute on their agenda and interest?
      • Some members of the board of directors are coming from industry and they’ll be able to use the resolutions that come out of the board of directors to direct how members of their own institutions/companies act
      • They can: fundraise, hire, establish working groups (with both volunteers and employees). Write JEPS and participate as any other community member.
    • How can the board of directors and the TSC relate to one another?
      • Need to come up with a way to have a strong connection between the two organizations
      • Historically there have been challenges in maintaining trust between members and projects of the Jupyter community
      • Ensure that membership on the board of directors are (partially) trusted individuals that the TSC respect
        • Should be members of the open source community, representatives from the people that we explicitly want to help with Jupyter’s mission
      • What happens if the board of directors doesn’t get stuff done?
        • E.g. if a bunch of CoC violations happen and they aren’t dealt with
      • Shouldn’t be too prescriptive about process in the board of directors
        • Instead, this group should define the process for itself. It can come up with its own internal rules, but there shouldn’t be too many rules that the governance documents give to it.
      • Who will the “Organizational Director” be? What will they do?
        • It could depend on who they are, and what is the broader context of the Jupyter steering council
        • Could be “vision-setting”, could be more organizational and operational.
        • May change over time as new people are appointed.
    • What are some actionable steps forward
      • Write out voting procedures
      • Write out a JEP process
      • Have a high-level sketch of the pieces of the governance document so we know what’s needed to have a complete doc
      • Arrive at a consensus about who are the members of the board of directors
        • Allocation %s from different groups etc
        • How long to they stay on
    • The primary power of the TSC is to approve JEPS and the primary power of the BoD is to create working groups.
1 Like
78

The weekly hourly office hours call is open to the community members who care about governance issues. The call is held 9-10AM PST on Tuesdays.

Meeting Minutes – May 5, 2020

  • Attendees
    • Darian
    • Tim G.
    • Sharan
    • Steve
    • Chris H.
    • Jason
  • Are there actionable, bite-sized, iterative improvements we can do to improve governance?
  • Open a PR in governance repo to reduce the quorum required, and decrease time allowed to vote on governance changes after which they will be closed or implemented.
  • We spent the call drafting this PR proposing a time limit on votes: https://github.com/jupyter/governance/pull/77
79

The weekly hourly office hours call is open to the community members who care about governance issues. The call is held 9-10AM PST on Tuesdays.

Meeting Minutes – May 12, 2020

  • Attendees
    • Darian
    • Tim G.
    • Sharan
    • Steve
    • Chris H.
  • Follow up on pull request from last week: https://github.com/jupyter/governance/pull/77
    • We assigned individuals to reach out to SC members who have not voted.
  • We discussed what a future incremental update to governance could be:
    • Better defining the process of a vote:
      • A PR should be opened in “draft” mode to allow for conversation
      • All discussion of the language should only happen in the drafting phase
      • The PR should enter the review / voting phase
      • Once a proposal is up for voting, it should not be modified (as it was in #77 above)
      • The top line comment in a PR should always reflect the executive summary of the changes.
      • Issues this could help with:
        • When people vote, it should be clear what they’re voting on, and the content of the proposed change shouldn’t be changed after they vote
        • When people haven’t been closely following a topic, they should have quick access to information about the current state of conversation, and more importantly to the current thing they’re voting on
        • People who open up proposed changes should have guidelines to make their proposal as smooth and efficient as possible to help resolve the above issue
    • We could defined “abstain” votes as per this comment: https://github.com/jupyter/governance/pull/77#issuecomment-624683297
80

Thanks for the update :smiley:

  • All discussion of the language should only happen in the drafting phase

So if I am interpreting this correctly, a vote will pertain to only the wording presented. If changes are needed, the opened vote would be withdrawn and a new vote opened.

1 Like
81

Hi Carol! Yes, exactly what you said: if the wording needs to change, we can either close the PR entirely or put it back into “draft” mode and zero out all the votes and ask everyone to consider the new language.

This is a change from the way we might comment on a code pull request, and the idea is to make sure that what happened in the current open vote on time limits for votes where I merged in a language change into the pull request after voting had begun does not happen again. This seems like a modest change that we can propose in a separate PR. The goal of our meeting this week was to think of similar modest incremental change we can iterate on concurrent with the larger work that is going on.

1 Like