Governance #10

tacaswell · 2020-07-27T00:40:23Z

This is the governance and named roles text from #7 pulled out into its own PR.

story645 · 2020-07-27T05:47:07Z

governance.md

+- Matplotlib's social media presence
+
+This communication is primarily broadcast out and intended to engage with
+end-users.


change end -users to broader community. Like a good chunk of our social media engagement is core devs on other projects.

story645 · 2020-07-27T05:53:08Z

governance.md

+
+In addition to communication to end-users we need to maintain
+communication channels to developers of down-stream libraries and
+other expert users.


the discourse is not solely for expert users and like we specifically want folks at all levels 'cause the discourse is also for the folks who won't go near SO.

The difference between the two community roles as I see it are:

communications-have a conversation with folks [general] about what's going on w/ the project & what folks can do w/ it & the like - this is more marketing/outreach

developer relations - have conversations with specific folk about what they're trying to do w/ mpl & want to do - this is more a problem solving/helping folks get acclimated role

story645 · 2020-07-27T05:55:45Z

governance.md

+communication channels to developers of down-stream libraries and
+other expert users.
+
+- interaction with 3rd party extensions / publicity


I think this belongs in comms 'cause like that's the type of stuff the person wearing that hat is good at while like this role doesn't need it. For example I can think of multiple devs who have shown that they're fantastic at all the things on this list except publicity. I think 3rd party extensions go under both 'cause again it's all community.

governance.md

tacaswell · 2020-07-27T14:53:38Z

Maybe the text in the Community Manager description is too detailed / fine-grained?

On a bit more reflection I realized that this wording bakes an inconsistency into the document (change the DPL by majority vote of SC, governance by 2/3 majority note of SC) so moving them out to their own file.

governance.md

story645 · 2020-07-27T15:40:00Z

Maybe the text in the Community Manager description is too detailed / fine-grained?

Yes/No in that it's got more of an umbrella role than the others, but I think it's fine to pull the description back to something like

evangelism, outreach, and support.
maintains the venues for conversation w/ the community

jklymak · 2020-07-27T15:46:41Z

I'm having some trouble placing this in the framework of other organizations I'm familiar with. The steering council seems like an executive, but no roles are discussed in the context of the steering council. So there is no Treasurer, Secretary, or any other responsibility designated to the steering council. Similarly, it would be unusual to define the president, and then the group that he is president of - the president is typically just the leader of the executive.

I think the current setup is a bit unclear who has ultimate authority. I think it should be the steering council, and that the PL's authority should devolve from the steering council. I appreciate that is a change, but I think it would be more robust to problems with future PLs. I know you claim forking is the way to change the PL, but if a future PL takes the project down some crazy route, its not particularly acceptable for them to take the matplotlib name with them, and it doesn't seem like it should be allowed to happen if there is a reasonable Steering Council in place.

tacaswell · 2020-07-27T17:34:12Z

That is a fair point, however I would prefer to keep this discussion about just the Deputy Lead roles (and renaming BDFL -> PL) rather than completely re-do the governance structure. I can remove all of the changes to the SC if that would make it clearer.

Historically we have been a BDFL / single lead + rough consensus style project which in some sense is, as long as you trust the PL (and I trust me and hope y'all also trust me), the simplest and clearest ultimate authority ;).

I think one of the worst failure modes here are a complicated governance document that we don't actually follow. For better-or-worse, we have an existing (and seemingly functioning) process that I worry if we try to make too big of a change we will revert back to (and evidence of this being a problem I cite my failure to fully stand up the SC yet). There have been a couple of projects (notably skimage and cpython) that have moved from single leader -> committee that we can use as a model. My preference here is that we actually stand up a full-sized (bigger than 3) SC and then writing them selves into full existence / authority.

jklymak · 2020-07-27T17:49:04Z

... and then writing them selves into full existence / authority.

Sure, if they have that authority.

The problem is not whether we trust you, its whether you get hit by a bus tomorrow, then what happens?

tacaswell · 2020-07-27T18:11:26Z

Sure, if they have that authority.

They do because the document has a method of changing its self (via 2/3 vote of current SC)

jklymak · 2020-07-27T18:31:56Z

They do because the document has a method of changing its self (via 2/3 vote of current SC)

... which the PL can veto?

The full Steering Council must vote and at least 2/3 of the votes must be positive to carry out the proposed action (fractions of a vote rounded up to the nearest integer). Since the PL holds ultimate authority in The Project, the PL has authority to act alone in accepting or rejecting changes or overriding Steering Council decisions.

dopplershift

Comments on a few minor things.

These seem like straightforward modifications that would be good to get done. Discussion of eliminating PL and moving to a committee would best be done in another issue, or even better, against concrete language adjusting the document. None of this moves us further away from that model, and in fact could be argued is small step towards it.

deputy_project_leads.md

governance.md

- BDFL is an inside joke that has gotten outside of the in-group where it was funny. I feel dumb using it to describe my self, but the look on my in-laws face when I used the term was the final straw. - Deputy Lead Developers aka "named roles" is a way to recognize that people are leading development along different axes of process. This is motivated by several aspects: 1. empower people to make decisions 2. provide a title people can put on their resume's

No fundamental changes, just wording and a bit of re-arrangement.

- adjust BDFL -> LD - add emphasis to financial / fund raising aspect of council - assert the LD counts in the council size

- introduce 2yr terms. We want people to be able to gracefully roll out of responsibility. - introduce A/B staggered classes. To make sure we maintain some continuity

Given that our Steering Council is smaller than Jupyters, do not imply Institutional Partners can install Council Members.

Given that we have a smaller council, expect all to participate. If someone has gone inactive, then they should be removed first.

tacaswell · 2020-08-04T22:51:49Z

@efiring @dopplershift Can you please approve and merge this? Once it hits master it will be "in force".

dopplershift · 2020-08-04T23:01:00Z

deputy_project_leads.md

+**Communications**
+
+We need to maintain communication channels with the wider world.  To
+this end the CM is responsible for maintaining.  This communication is


"To this end the CM is responsible for maintaining." maintaining what?

dopplershift · 2020-08-04T23:01:45Z

deputy_project_leads.md

+primarily broadcast out and intended to engage with the wider
+community.  The
+
+- the [Ann] mailing list


This is the only one not capitalized.

dopplershift · 2020-08-04T23:01:57Z

deputy_project_leads.md

+other expert users.
+
+- interaction with 3rd party extensions / publicity
+- recruiting new contributors


Again, inconsistent capitalization.

people.md

dopplershift

I'm not going to hold this up for the minor things I noticed above.

tacaswell · 2020-08-04T23:43:25Z

Fixed the spelling, we might want to tighten up the whole CM section to be closer in length to the others.

efiring

There are still two of Ryan's changes outstanding, and I have added a few requests for clarification and one for deletion of a sentence.

governance.md

efiring · 2020-08-05T00:38:07Z

governance.md

+  Services for the benefit of the Project and Community.
+- Make decisions when regular community discussion doesn’t produce consensus on
+  an issue in a reasonable time frame.
+- Granting or revoking commit rights.


-> "Grant or revoke..."

governance.md

efiring · 2020-08-05T00:49:27Z

governance.md

+
+The Council reserves the right to eject current Members, other than the PL, if
+they are deemed to be actively harmful to the project’s well-being, and attempts
+at communication and conflict resolution have failed, from The Project.


Move "from The Project" up -> "eject current Members, other than the PL, from The Project, if they..."
But is this correct? Is the ejection from the Project, or from the Council? What would the former mean, and why would that ejection seat be reserved for Council members?

I think this was meant to eject people from the SC but not from the project overall, but I am not sure the middle option (stay in project but ejected from SC?), changing to be fully eject from the project.

Is it even defined for a person to be in, or out of, The Project? I don't think we define project membership as such; there are Contributors, Developers, etc. but not Members, I think.

The Council reserves the right to eject anyone, including the Council
Members, from the project's online spaces if they are deemed to be
actively harmful to the project’s well-being, and attempts at
communication and conflict resolution have failed, by a super majority vote.

We can ban people from github / discourse / mailing lists / gitter if they are being disruptive.

efiring · 2020-08-05T00:50:46Z

governance.md


 ### Conflict of interest

-It is expected that the BDFL and Council Members will be employed at a wide
+It is expected that the PL, DPLs, and Council Members will be employed at a wide


Below, plural is "conflicts of interest", not "conflict of interests".

I think I got all of these?

governance.md

To ensure the independence of SC members, none of the SC members can be in the same reporting tree through employment or contracting.

This brings the description to the same level of detail as the other roles and does not encode particular communication forums / mechanisms directly in the governance text.

tacaswell · 2020-08-05T03:25:12Z

Well, I was going to do follow ups, but it's merged so 🎉

story645 · 2020-08-05T05:01:45Z

deputy_project_leads.md

-From time to time Matplotlib as an organization has in-person
-meetings.
+the future and may want to enlist further assistants.  The CM is responsible
+for Matplotlib's evangelism, outreach, and user support as well as mainitaining


for community outreach and support, as well as for publicizing what's new and useful in the project.

(I hate the word evangelism)

@story645 Your review came after I merged the PR, but please go ahead and put in a PR with this fix and any other fine-tuning you see.

efiring · 2020-08-05T05:57:04Z

Sorry if I jumped the gun, but I'll blame @dopplershift--I inferred from his review that it was time to just get this in.

dopplershift · 2020-08-05T07:35:10Z

I'll take the blame. I didn't want to grammar this to death, especially for things that are not really governing (like verbiage describing DPL roles), since we've spent so long getting things going governance-wise.

story645 · 2020-08-05T07:51:07Z

I just didn't see the merge, it's all fine.

tacaswell changed the title ~~Goverance~~ Governance Jul 27, 2020

story645 reviewed Jul 27, 2020

View reviewed changes

story645 mentioned this pull request Jul 27, 2020

overhaul governance #7

Closed

timhoffm reviewed Jul 27, 2020

View reviewed changes

governance.md Outdated Show resolved Hide resolved

governance.md Outdated Show resolved Hide resolved

governance.md Outdated Show resolved Hide resolved

jklymak reviewed Jul 27, 2020

View reviewed changes

governance.md Outdated Show resolved Hide resolved

jklymak reviewed Jul 27, 2020

View reviewed changes

governance.md Outdated Show resolved Hide resolved

jklymak reviewed Jul 27, 2020

View reviewed changes

governance.md Outdated Show resolved Hide resolved

dopplershift reviewed Jul 28, 2020

View reviewed changes

deputy_project_leads.md Outdated Show resolved Hide resolved

deputy_project_leads.md Outdated Show resolved Hide resolved

governance.md Outdated Show resolved Hide resolved

story645 and others added 15 commits July 28, 2020 09:49

onboarding oriented governance docs

fb0de6d

DOC: google+ is dead

2c5e9bb

GOV: re-write BDFL -> LD section

29d0d8e

No fundamental changes, just wording and a bit of re-arrangement.

GOV: tweak Steering Council a bit

0045e03

- adjust BDFL -> LD - add emphasis to financial / fund raising aspect of council - assert the LD counts in the council size

GOV: substantially re-write the council member section

8791ab9

- introduce 2yr terms. We want people to be able to gracefully roll out of responsibility. - introduce A/B staggered classes. To make sure we maintain some continuity

GOV: update wording in CoI section

d991ef6

GOV: slight wording change to Council communications

c4fbe0a

GOV: clear out rest of BDFL -> LD

e4c1ece

GOV: tweak wording for institutional partners

6fb94ee

Given that our Steering Council is smaller than Jupyters, do not imply Institutional Partners can install Council Members.

GOV: tweak rules to change governance

50a7dc2

Given that we have a smaller council, expect all to participate. If someone has gone inactive, then they should be removed first.

GOV: add sketch of DLD titles

92ad746

GOV: make the get a commit bit wording consistent

000c844

GOV: add domain project leads

419a2a2

GOV: change "Lead Developer" -> "Project Leader"

ba4b43e

DOC: fix grammar and wording

f43145e

GOV: Name initial Deputy Project Leads

dd55d30

tacaswell force-pushed the goverance branch from bd569f8 to dd55d30 Compare August 4, 2020 22:54

dopplershift reviewed Aug 4, 2020

View reviewed changes

people.md Outdated Show resolved Hide resolved

dopplershift approved these changes Aug 4, 2020

View reviewed changes

dopplershift mentioned this pull request Aug 4, 2020

GOV: add Tim Hoffmann to the steering council #13

Merged

DOC: fix spelling

ce1e9cf

efiring requested changes Aug 5, 2020

View reviewed changes

tacaswell added 5 commits August 4, 2020 22:24

DOC: typo and minor wording changes

7254bdf

GOV: remove "fork if you do not like it" suggestion

ae0579c

GOV: clarify that the SC can eject contributors from the project

734ed9f

GOV: add no-reports condition on SC membership

efabe3f

To ensure the independence of SC members, none of the SC members can be in the same reporting tree through employment or contracting.

GOV: shorten the description of the community manager

2b40e72

This brings the description to the same level of detail as the other roles and does not encode particular communication forums / mechanisms directly in the governance text.

tacaswell requested review from dopplershift, efiring, story645 and timhoffm August 5, 2020 02:34

efiring approved these changes Aug 5, 2020

View reviewed changes

efiring merged commit f9b3bc5 into master Aug 5, 2020

tacaswell deleted the goverance branch August 5, 2020 03:24

story645 reviewed Aug 5, 2020

View reviewed changes

Governance #10

Governance #10

Uh oh!

Conversation

tacaswell commented Jul 27, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

tacaswell commented Jul 27, 2020

Uh oh!

Uh oh!

Uh oh!

Uh oh!

story645 commented Jul 27, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

jklymak commented Jul 27, 2020

Uh oh!

tacaswell commented Jul 27, 2020

Uh oh!

jklymak commented Jul 27, 2020

Uh oh!

tacaswell commented Jul 27, 2020

Uh oh!

jklymak commented Jul 27, 2020

Uh oh!

dopplershift left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

tacaswell commented Aug 4, 2020

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

dopplershift left a comment

Choose a reason for hiding this comment

Uh oh!

tacaswell commented Aug 4, 2020

Uh oh!

efiring left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

tacaswell commented Aug 5, 2020

Uh oh!

Choose a reason for hiding this comment

tacaswell commented Jul 27, 2020 •

edited

Loading

story645 commented Jul 27, 2020 •

edited

Loading