VOTE: Changes to Governance Model for pandas

[Dr-Irv]

There is a proposal to change the governance model for pandas. The proposed governance model is described in 2 pull requests:

• Governance proposal: Steering Committee and Teams #17 describes the creation of a Steering Committee and the concept of Teams
• Teams Responsibilities, permissions, criteria #18 describes the responsibilities of the different Teams

According to the current governance model, 80% of the core team needs to vote on this issue, with at least 2/3 of the votes being in favor. That means we need the votes of 17 individuals, with a positive vote of 14 members of the core team guaranteeing approval.

The rendered governance documents can be found at:

• Main Governance Document: https://github.com/pandas-dev/pandas-governance/blob/3b494821873a50fd1bdc8609445fa7d411848fe7/governance.md
• Description of Individual Teams: https://github.com/pandas-dev/pandas-governance/blob/c0e5c586ed9b6dd95d180eff8a6e482571268fa9/teams.md

Voting will close on August 31, 2024.

Cast your vote in a comment below.

• +1: approve.
• 0: abstain.
• -1: disapprove

@pandas-dev/pandas-core

[WillAyd]

+1

[mroeschke]

+1

[alimcmaster1]

+1

[wesm]

+1

…

On Wed, Aug 7, 2024 at 2:37 PM Ali McMaster ***@***.***> wrote: +1 — Reply to this email directly, view it on GitHub <#20 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AACQO56O7VREP7KY6ZQBEBTZQJZQRAVCNFSM6AAAAABMEYW2FWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDENZUGIYTENJSG4> . You are receiving this because you are on a team that was mentioned.Message ID: ***@***.***>

[fangchenli]

+1

[jbrockmendel]

Abstain

[lukemanley]

+1

[rhshadrach]

+1

[datapythonista]

-1

• Not a fan of the steward comcept and giving non core devs the same voting rights as core devs
• I think the new team definitions are unrealistically complex. In oururrent simpler team structure most teams don't do much in practice, and the ones that work (like finance) is usually a single person doing all the work. We should go into making that simpler and closer to reality, not the opposite
• Feels like a 5+ member steering commitee is inefficient. I'd rather have a 3 people committee that just helps when consensus isn't possible

[jreback]

+1

[attack68]

+1

Was torn between 0 and +1, actually, thinking similar to Marc that this may be too complex and impractical. In the end I view having the momentum to implement something positive and reacting to or iterating any failures later is more favourable than losing all the momentum and risking not adopting any changes at all.

[jorisvandenbossche]

+1

[TomAugspurger]

+0 changed to +1 #20 (comment)

[twoertwein]

0

[lithomas1]

0

[MarcoGorelli]

+1

A few things I'd like to comment on:

• Regarding Marc's comments: my understanding is that stewards only get equal voting rights for changes to the governance model, and not to pdeps or other matters. I think I'm OK with that
• I'm also not sure about some teams, but I don't that having them is harmful
• This wouldn't set the governance model in stone. Things can be revised, and I think this is a step in the right direction (for example, I'd like to propose shortening the voting period, but I'll do that separately)

I think this is a positive though, and a good move away from the current BDFL model: Wes hasn't been active in years, and based on his vote he probably doesn't have interest in acting as BDFL anymore anyway

Note from Joris: further responses and discussion -> #17 (comment)

[simonjayhawkins]

Anyway, I think people are happy with this and it's virtually approved. I

just wanted to clarify that I don't think your first bullet point is a minor thing. IIUC from Irv's post 80% of the core team needs to vote on this issue, with at least 2/3 of the votes being in favor. We have 4 abstensions and 1 against. Another abstention or negative vote could swing the outcome. Joris has asked people to vote even if they abstain. I'm reluctant to do so!

…

On Tue, 27 Aug 2024 at 15:53, Joris Van den Bossche < ***@***.***> wrote: Marc, could you move your comment to #17 <#17> ? (I will reply there as well) — Reply to this email directly, view it on GitHub <#20 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ADEMUXOGTWF3YZHTAY45LQDZTSHIBAVCNFSM6AAAAABMEYW2FWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGMJSG44TCOBVGU> . You are receiving this because you are on a team that was mentioned.Message ID: ***@***.***>

[jorisvandenbossche]

Procedural note: it seems there is some contradiction between what I said in the email about abstentions (just counting towards quorum, as we do for PDEPs) versus what Irv wrote in the top post. I am currently clarifying that with him.

(and for more comments or responses on the actual proposal body itself -> #17)

[Dr-Irv]

+1

[noatamir]

0

[bashtage]

• I think 5 is a reasonable number for the steerco. There will likely be times where one or more people cannot be present and if it was only 3, then you have a very small committee.
• There is something wrong with the subsequent election mechanism. It sounds like if the original 5 never step down, then there is no obvious way to infuse new blood. It seems to say that an election will only be called if there is a resignation and new volunteers. Presumably there should be a new election whenever there are new volunteers irrespective of whether the steerco is willing to continue.
• The official list of teams seems very complex. Is there a minimum team size (other than 1)?

(note Joris: I answered to the second bullet point at #17 (comment))

[bashtage]

While I have reservations about the specific implementation. I think the overall idea is needed. Replacing my 0 (abstention) as asked for.

One thing that mich simplify the teams issue would be to let the teams be designated by the steerco as needed. This is often how non-exec boards operate - they have the power to nominate subcommittees and sometimes delegate authority to them. Subcommittes can often include people who are not on the non-exec board due to special skills or knowledge.

+1

[simonjayhawkins]

-1

The current governance documents allow for changes to be submitted via a GitHub pull request, refined through public comment and review, aiming for community consensus. If this negative vote swings the outcome, the proposal can be modified to address concerns raised by both positive and negative voters, and then voted on again. It might even be beneficial to revisit the proposal, address the ambiguities, and ensure that the governance model is both robust and clear. There should be no risk of losing momentum.

[attack68]

One day left until this officially closes. Is it worth reaching out to anyone that hasn't yet voted @phofl @topper-123?

<style> </style>

10.00%	25.00%	65.00%
2	5	13
-1	0	1
datapythonista	jrbrockmendel	willAyd
simonjayhawkins	tomaugspurger	mroeschke
	twoertwein	alimcmaster1
	lithomas1	wesm
	notamir	fangchenli
		lukemanley
		rhshadrach
		jreback
		attack68
		jorisvandenbossche
		marcogorelli
		dr-irv
		bashtage

[noatamir]

Amending my vote to +1

Recognising that there is no perfect governance model,
I agree with Marco that this is a positive move away from the BDFL model and that it can be improved based on what is helpful and unhelpful in practice.

I know it reads complex at first sight, but I do like how explicit it is. I think this is a great improvement in transparency.

[phofl]

This was a hard one :(

+1

Overall, I think this moves us to a better state than what we had before (but barely, I would have voted 0 if abstaining was still an option). I thought about this for the last few weeks and I really dislike the following sentence:

The Steering Committee will itself decide when a circumstance is exceptional.

I strongly hope that we can work on clarifying this better even if the proposal is accepted in the current form.

Additionally, I am not a fan of

All members of each Team, except the Code of Conduct Team, are eligible to vote.

I'd generally be happy with giving newish contributors more agency but making them eligible to vote on the steering council adds weight to that that I am definitely not a fan of.

[TomAugspurger]

Revising my vote to a +1.

Given my limited participation recently I don't want to push this too much one way or the other, but overall I do think the proposal is a good one.

[jbrockmendel]

IIUC the other 0s changing to 1s mean my 0 is no longer at risk of making the vote fail, so I'm keeping at 0. IFF that understanding is incorrect then count me as +1.

1. In the future we should find a way to abstain that isn't a de-facto -1.

2. My concern about the [gestures vaguely] governance stuff is that it empowers people who are willing to spend their time in governance meetings and who are willing to wade through or write walls of text.

3. My preferred form of governance is an informal do-ocracy: the people who actually do the work make the decisions. Both as a matter or principle and because they are more likely to actually understand what's going on.

4. I agree with others who have pointed out that the SC deciding what constitutes an exceptional circumstance is bad design.

(note Joris: I responded to some of those points in #17 (comment))

[jorisvandenbossche]

With the voting period ended, summarizing the vote:

-1	0	1
2	3	15
10%	15%	75%
--	--	--
datapythonista	jrbrockmendel	willAyd
simonjayhawkins	twoertwein	mroeschke
	lithomas1	alimcmaster1
		fangchenli
		lukemanley
		rhshadrach
		jreback
		attack68
		jorisvandenbossche
		marcogorelli
		dr-irv
		bashtage
		noatamir
		phofl
		tomaugspurger

And in addition a special +1 vote of (the now almost officially retired BDFL) Wes.

With this, the quorum is reached (20/21 voted, >80%) and the required majority of 2/3 is passed, and the proposal is accepted!

Thanks all for voting and engaging in the discussion. With the governance working group we will propose some next steps to transition to the new governance, and also I want to ensure that we are aware of some remaining concerns and will try to follow-up on that.

[jorisvandenbossche]

With the vote finalized, closing this issue.