git/list[1] front-page[2] threads[3] people[4] search[5] about
 

Re: [PATCH] Documentation: propose community voting process

From
Emily Shaffer <nasamuffin@google.com>
Date
Oct 21, 2026, 20:05 UTC
Message-ID
<CAJoAoZnhc6Tjm9m9dmTA6UD18HzkxAv2_ifB5x74ectqX8YieA@mail.gmail.com>
In-Reply-To
<20261021200341.378978-1-emilyshaffer@google.com>
On Wed, Oct 21, 2026 at 1:03 PM Emily Shaffer <emilyshaffer@google.com> wrote:
Show 132 quoted lines
>
> Not all discussions can be settled by community consensus. For truly
> contentious decisions we need a way to democratically make a decision
> that represents the majority will of the project. This doc outlines the
> methodology we use in the rare case where we need to vote to make a
> decision.
>
> Signed-off-by: Emily Shaffer <emilyshaffer@google.com>
> ---
> Hi folks,
>
> Thanks for your patience with this taking much longer than I expected or
> wanted it to.
>
> As we discussed in the Contributor Summit this past September[1], there's
> interest in taking a formal community vote to decide whether and how to accept
> AI contributions. The topic is of course a very contentious one, and I don't
> think that we're going to come to general consensus the way we have on most
> other concerns in the past.
>
> With this patch I'm trying to outline only the process for collecting the vote,
> NOT a proposal for the AI contribution policy itself. I'm interested in making
> sure that the Git community is able to be heard fairly, and that we don't end
> up dealing with tampering during the voting process (e.g. non-contributors
> dogpiling to shift the vote one way or another
>
> I expect the approval of the voting process itself to be fairly non-contentious
> (or else I haven't managed to write a fair process!) so for approving the
> voting process, I suggest we use a similar approach to the one we did with the
> Code of Conduct back in 2019[2][3]: discussing the process itself, ultimately
> coming to general consensus, and collecting a list of Acked-by: to put onto the
> commit outlining the voting process. And by the way, I do think that we should
> check in the voting process; it seems likely to me that we will want to revise
> the AI policy again given the pace of development, and it may be useful for
> other discussions that we've had trouble resolving in the past.
>
> BTW, there are some parts of this proposal that I still don't have a good
> answer for, so look for my self-review immediately to follow, highlighting open
> questions.
>
> [1]: https://lore.kernel.org/git/summit-2026.94e33e9ddf234334.06@ttaylorr.com/
> [2]: https://lore.kernel.org/git/20190924064454.GA30419@sigill.intra.peff.net/
> [3]: https://github.com/git/git/commit/5cdf2301d4a8b439f89fd2cdfed13736ccd64a30
>
> ---
>
>  Documentation/CommunityVoting.adoc | 100 +++++++++++++++++++++++++++++
>  Documentation/DecisionMaking.adoc  |   6 ++
>  Documentation/Makefile             |   1 +
>  3 files changed, 107 insertions(+)
>  create mode 100644 Documentation/CommunityVoting.adoc
>
> diff --git a/Documentation/CommunityVoting.adoc b/Documentation/CommunityVoting.adoc
> new file mode 100644
> index 0000000000..155acc9777
> --- /dev/null
> +++ b/Documentation/CommunityVoting.adoc
> @@ -0,0 +1,100 @@
> +Voting in the Git Project
> +=========================
> +
> +For the most part, technical concerns in Git should be resolved by community
> +consensus - either on the mailing list or in one of the annual-ish Contributor
> +Summits. Voting should be used as a last resort when the community really and
> +truly cannot come to a compromise on a policy decision. As such, the initiation
> +of a vote should be generally agreed on by the mailto:git@sfconservancy.org[Git
> +PLC] or primary maintainer.
> +
> +The voting discussion should be held to a single (huge) thread, to make it easy
> +to refer back to the decision process when looking at the successful proposal
> +after the process concludes. The thread may be structured like so:
> +
> +--------
> +Formal vote: Commit subject line style guide
> +  Re: Formal vote: Commit subject line style guide
> +  ...
> +  [PROPOSAL] Require emojis in commit subjects
> +    Re: [PROPOSAL] Require emojis in commit subjects
> +    [PROPOSAL v2] Require emojis in commit subjects
> +      Re: [PROPOSAL v2] Require emojis in commit subjects
> +      ...
> +          [PROPOSAL v4] Require emojis in commit subjects
> +  [PROPOSAL] Require Camel Case In Commit Subjects
> +    Re: [PROPOSAL] Require Camel Case In Commit Subjects
> +    [PROPOSAL v2] Require Camel Case In Commit Subjects
> +  [PROPOSAL] forbid capitalization in commit subjects
> +  ...
> +--------
> +
> +A formal vote has three phases:
> +
> +1. Proposal collection and refinement
> +2. Voting window
> +3. Implementation
> +
> +Phase 1: Proposal collection and refinement
> +-------------------------------------------
> +
> +Proposals should be formatted as a patch to Git's documentation, for easier
> +review and implementation once landed. Anyone who is eligible to vote is
> +eligible to submit a proposal. To avoid noise on the final ballot, a proposal
> +will only be added to the ballot if it has been seconded by another eligible
> +voter, indicated by a `Seconded-by:` footer in the patch's commit message.
> +
> +After the first draft of the proposal has been sent, the community is
> +encouraged to discuss and polish the proposal, just like any other code review.
> +Keep in mind that these proposals still need to be voted on, so it is likely
> +counterproductive to use the discussion period to try to completely change the
> +nature of the proposal or the mind of the author.
> +
> +This phase should last around 4 weeks in order to provide plenty of time for
> +community members to write proposals and respond to feedback, and make space
> +for contributors' individual timezone and availability differences. The
> +maintainer or PLC member who sends the kickoff message should also include the
> +closing date for proposal collection.
> +
> +It may also be useful to invite expert feedback during this time, e.g. from
> +Software Freedom Conservancy's legal counsel, depending on the topic under
> +consideration.
> +
> +Phase 2: Voting window
> +----------------------
> +
> +The final version of each seconded proposal will be collected into a ballot at
> +the end of Phase 1. When the ballot is prepared, eligible voters will be able
> +to vote by mail for two weeks, to make space for contributors' individual
> +timezone and availability differences. During this time the proposals may no
> +longer be altered; the ballot should refer to a specific version of the
> +proposal (e.g. `[PROPOSAL v4] Require emojis in commit subjects`) and any later
> +version of the proposal will be ignored.
> +
> +Votes will be decided using <mechanism TBD>.
There are a few appealing options.
* Debian uses devotee (https://salsa.debian.org/debian/devotee), which
appears to require hosting. Finding neutral hosting may prove tricky.
* Konstantin let me know that The Linux Foundation's Technical
Advisory Board has used CIVS (https://civs1.civs.us/) in the past. He
says he wasn't involved in setting it up but doesn't think it looks
too challenging. CIVS and devotee use the same tallying algorithm.
* jj used voting on a GitHub poll in the past. That would just give us
simple majority, and it would require GitHub accounts, which may be a
barrier.

These all have their pros and cons; devotee and CIVS seem to me to be better suited to many options that may be somewhat similar to each other.

The other factor that I haven't dug into too much is how to only allow eligible voters to cast a vote. GitHub isn't a great fit for this, we'd need to collect GH IDs of every eligible voter and put them into a group or something, if I understand best how this would work. CIVS emails a link to each eligible voter - since I propose basing eligibility on mailing list participation and patch submissions I think there is no problem there. devotee also appears to be email based.

Show 26 quoted lines
> +
> +Phase 3: Implementation
> +-----------------------
> +
> +When the voting window ends, the winning proposal should be merged to `main`
> +without delay by the maintainer. The commit message should be appended to
> +include a link to the top of the voting discussion thread so that the
> +discussion can be revisited by history divers in the future, as well as a brief
> +summary of the result of the vote (% in favor out of total number of
> +proposals).
> +
> +Voting Eligibility
> +------------------
> +
> +Votes are welcome from contributors who have authored a commit which has been
> +merged to next, or who have provided 2 or more code reviews, and who have not
> +been banned from participation in the project. All eligible contributor votes
> +are weighed equally, regardless of the overall number of contributions they
> +have landed.
> +
> +To prevent dogpiling, contributions are only eligible for a given vote if they
> +landed before the beginning of the voting process (the kickoff email mentioned
> +above). That is, the code reviews must have been sent or the patch must have
> +been accepted before the initial email is sent. The intention is to ensure the
> +mailing list isn't spammed with low-quality reviews during the voting process
> +solely to increase the number of eligible voters on one side of the issue.
The specifics of this eligibility rule can change a little. I'm trying
to figure a balance between a few principles:
* Not all contributions are code (e.g. community moderation is
contribution but a moderator may never send a patch). I think in the
Git project we have less issue with that, because we don't have a
community moderator unless we're counting the mods in the low-traffic
Discord, and I believe all of those mods have code or review
contributions anyways.
* Avoiding dogpiling. This is my biggest fear, Git is a very
well-known project and there will be a lot of big opinions on the open
internet once people know we're voting on this policy. I want us to
measure the opinions of those who participate in the project, not the
opinions of random passers-by.
* Easy (scriptable) eligibility determination. If we start to explore
other types of contribution, like user support in non-mailing-list
platforms, it becomes really difficult to send a ballot to everyone
who's supposed to get one.
Show 29 quoted lines
> diff --git a/Documentation/DecisionMaking.adoc b/Documentation/DecisionMaking.adoc
> index b43c472ae5..c08fb7abd5 100644
> --- a/Documentation/DecisionMaking.adoc
> +++ b/Documentation/DecisionMaking.adoc
> @@ -72,3 +72,9 @@ We prefer that "official" discussion happens on the list so that the full
>  community has opportunity to engage in discussion. This also means that the
>  mailing list archives contain a more-or-less complete history of project
>  discussions and decisions.
> +
> +Formal Voting
> +-------------
> +When a decision cannot be reached by general consensus or compromise, a formal
> +vote may be called. See link:CommunityVoting.adoc[Community Voting] for details
> +of this process.
> diff --git a/Documentation/Makefile b/Documentation/Makefile
> index f8dea4b395..314b8dc8f5 100644
> --- a/Documentation/Makefile
> +++ b/Documentation/Makefile
> @@ -113,6 +113,7 @@ SP_ARTICLES += $(API_DOCS)
>
>  TECH_DOCS += BreakingChanges
>  TECH_DOCS += DecisionMaking
> +TECH_DOCS += CommunityVoting
>  TECH_DOCS += ReviewingGuidelines
>  TECH_DOCS += MyFirstContribution
>  TECH_DOCS += MyFirstObjectWalk
> --
> 2.56.0.rc1.315.gc6ed9934b7-goog
>
Previous: Emily ShafferNext: Phillip Wood
Message 11 of 52 in “Notes from the Git Contributor's Summit, 2026”
  1. Taylor BlauOct 6, 2026
  2. 01/07 Security mailing list and security processTaylor Blau, Oct 6, 2026
  3. 02/07 Git 3.0Taylor Blau, Oct 6, 2026
  4. 03/07 DocumentationTaylor Blau, Oct 6, 2026
  5. Todd ZullingerOct 7, 2026
  6. Junio C HamanoOct 7, 2026
  7. 04/07 Outreachy sponsorshipTaylor Blau, Oct 6, 2026
  8. 05/07 What can we do next with pluggable ODB?Taylor Blau, Oct 6, 2026
  9. 06/07 AI contribution policyTaylor Blau, Oct 6, 2026
  10. Documentation: propose community voting processEmily Shaffer, Oct 21, 2026
  11. Emily ShafferOct 21, 2026
  12. Phillip WoodOct 22, 2026
  13. Junio C HamanoOct 23, 2026
  14. Phillip WoodOct 23, 2026
  15. Kristoffer HaugsbakkOct 23, 2026
  16. Junio C HamanoOct 23, 2026
  17. Kristoffer HaugsbakkOct 23, 2026
  18. Junio C HamanoOct 23, 2026
  19. Kristoffer HaugsbakkOct 23, 2026
  20. Tuomas AholaOct 24, 2026
  21. Weijie YuanOct 24, 2026
  22. Toon ClaesOct 30, 2026
  23. Weijie YuanOct 30, 2026
  24. Junio C HamanoOct 24, 2026
  25. Weijie YuanOct 24, 2026
  26. Jeff KingOct 26, 2026
  27. Junio C HamanoOct 26, 2026
  28. Patrick SteinhardtOct 26, 2026
  29. Tuomas AholaOct 26, 2026
  30. Emily ShafferOct 30, 2026
  31. Junio C HamanoOct 30, 2026
  32. Jeff KingOct 30, 2026
  33. Patrick SteinhardtNov 2, 2026
  34. Jeff KingNov 4, 2026
  35. Phillip WoodOct 26, 2026
  36. brian m. carlsonOct 22, 2026
  37. Junio C HamanoOct 21, 2026
  38. Junio C HamanoOct 21, 2026
  39. Justin ToblerOct 22, 2026
  40. Junio C HamanoOct 23, 2026
  41. Emily ShafferOct 30, 2026
  42. Kristoffer HaugsbakkOct 22, 2026
  43. Patrick SteinhardtOct 22, 2026
  44. brian m. carlsonOct 22, 2026
  45. Junio C HamanoOct 22, 2026
  46. Emily ShafferOct 30, 2026
  47. Kristoffer HaugsbakkOct 23, 2026
  48. Emily ShafferOct 30, 2026
  49. Junio C HamanoOct 30, 2026
  50. Jonathan CorbetOct 30, 2026
  51. Stéphane GlonduNov 9, 2026
  52. 07/07 Protocol v2 for pushesTaylor Blau, Oct 6, 2026

Read the whole thread, see it on lore, or plain text.

$ cat FOOTERMessages come from the public archive at lore.kernel.org/git, fetched every hour. The front page is chosen and written each morning by an AI editor and can be wrong; the threads themselves are the record. About and API. For agents: an MCP server at https://gitlist.dev/mcp, and any thread, story or person page as Markdown by adding .md to its URL (or sending Accept: text/markdown). Details in /llms.txt.