Vote in new maintainer volunteers for image-spec

[caniszczyk]

We've had a great set of active volunteers to become maintainers on the image spec:

• @jonjohnsonjr [Google]: Audit of active maintainers + call for new maintainers tob#95 (comment)
• @tych0 [Cisco]: Audit of active maintainers + call for new maintainers tob#95 (comment)
• @vrothberg [Red Hat]: Audit of active maintainers + call for new maintainers tob#95 (comment)
• @imjasonh [Red Hat]: Audit of active maintainers + call for new maintainers tob#95 (comment)
• @SteveLasker [Microsoft]: Audit of active maintainers + call for new maintainers tob#95 (comment)

The next step in the process after the call for volunteers is to initiate a PR and vote by the existing @opencontainers/image-spec-maintainers to add them as official maintainers, this will be done on a one by one basis: https://github.com/opencontainers/image-spec/blob/main/MAINTAINERS

[vbatts]

is there a period we wait on this vote?

[caniszczyk]

I plan on kicking the vote off tomorrow for folks, been a bit buried the last day or so

…

On Thu, Apr 15, 2021 at 9:07 PM Vincent Batts ***@***.***> wrote: is there a period we wait on this vote? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#838 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAAPSIP4LPMBNQIWHLCB6BDTI6LXBANCNFSM426DIUTA> .

-- Cheers, Chris Aniszczyk http://aniszczyk.org

[caniszczyk]

For the new image-spec interested maintainers, can you please talk to an existing @opencontainers/image-spec-maintainers and have them support you becoming a maintainer, once that is done, send a PR to kick off a vote, similar to this PR in another spec: opencontainers/distribution-spec#269

This will be a bit easier and have existing maintainers buy in then me doing it for each individual interested maintainer :)

[cyphar]

It might be a good idea to first send PRs removing maintainers who are no longer involved in the project -- so we can lower the LGTM threshold for new maintainers? (I'm not sure we have a 2/3 super majority of active maintainers so we might even need actual TOB involvement to break the deadlock.)

[dlorenc]

It might be a good idea to first send PRs removing maintainers who are no longer involved in the project -- so we can lower the LGTM threshold for new maintainers? (I'm not sure we have a 2/3 super majority of active maintainers so we might even need actual TOB involvement to break the deadlock.)

Heh - that was my guess as well as a casual observer :) Removing first might be needed to actually break the deadlock.

[sarahnovotny]

agreed that we have a challenge with so many inactive maintainers. that said, unblocking by making the group smaller and then voting will pose different issues. I don't see an issue with involving the TOB (this time) as a way to broaden the group selecting the new maintainers.

[dlorenc]

unblocking by making the group smaller and then voting will pose different issues.

Could you clarify any more here? I can't think of any issues but that usually means I'm missing something :)

Another benefit to having people step down first is that it might also give a clearer understanding of what the final group would look like without having to guess at who may or may not step down, sort of like the "atomic swap" here: #831

[sarahnovotny]

Could you clarify any more here?

small groups voting about who becomes a maintainer without clear criteria for the role (i might be mis-remembering this part.) doesn't project confidence of impartiality. a larger group will at least have /more/ different (individual or corporate) agendas to be set aside or be biased by.

This is particularly a challenge with Distribution and Image specs being distinct and intertwined.. There seem to be continued discussions about when the image spec will need a new semver major.

[caniszczyk]

FYI the reason I called for the vote of maintainers first is that it will be very easy to deduce if existing maintainers are active, if they don't either way, the signal is clear for the existing active maintainers and @opencontainers/tob to step in

If any of the existing @opencontainers/image-spec-maintainers wanted to step down, they are free to signify that: #839 - but I believe voting is a strong signal

[cyphar]

@sarahnovotny The issue is that if we are in a situation where it seems that more than half of maintainers are chronically inactive, and you need a two-thirds majority to do important business required for the project (governance changes, releases, and maintainer changes), keeping the old maintainers on the list and trying to expand it means that you need far more new maintainers in order to overturn the deadlock (and even more if you want to restore something close to a two-thirds threshold for the active maintainers).

I don't think we should just remove people without asking for their input first, but given that this is actually having an impact on the project's ability to function, I don't think having a larger group of people (many of whom are inactive) is a positive.

[sarahnovotny]

I don't think having a larger group of people (many of whom are inactive) is a positive.

I agree with that in principle. and, i see this as an unusual situation needing an out of process reset of some sort. I'd like to find a way that process looks, feels, and IS unbiased (as much as is humanly possible).

Chris' outlined plan (quoted below) is reasonable... i.e. if the Image Spec maintainers can't get 2/3 of the group together for the votes, then the TOB is brought in to help reset the maintainers list

FYI the reason I called for the vote of maintainers first is that it will be very easy to deduce if existing maintainers are active, if they don't either way, the signal is clear for the existing active maintainers and @opencontainers/tob to step in.

[cyphar]

Yeah, reading Chris's comment I think it makes more sense to first try things through the usual procedure and see how things go.

[dlorenc]

Is there a timeline on calling this issue "resolved"? A few PRs got merged but a bunch are still outstanding.