Moderation Team Accountability Issues

Meta Moderation
1

Given bits and pieces I have cought up, I had expressed my concerns in due process via Please don't anonymize decision makers · Issue #14 · NixOS/moderation · GitHub, however, unfortunately, I’m not really satisfied with the depths of consideration that had been returned on behalf of the moderation team.

If “bits and pieces” seems vague to you, it truly is, and hence I want to request your comments about the accountability of the moderation team.

Do we have an issue? Or maybe not (yet)? Do we even need to take corrective or preventive action as a community?

I can’t give a good answer, really, but the proposed constitution mechanism outlined in Create team-rotation.md by mweinelt · Pull Request #10 · NixOS/moderation · GitHub isn’t necessary inspiring confidence in the moderation team’s long term accountability.

Please write down what you think about this topic.

3 Likes
2

Oh this is interesting to me, in particular in light of the critical exchange that I helped spark regarding the newly adopted Code of Conduct. :sweat_smile:

I am not really sure how I feel about this one though. I kinda can see both sides here.

Yet, what I find more important than individual accountability is whether there exists an effective (!!!) mechanism to get rid of the moderation team and exchange it with new people if need arises.

Mitigation of abuse of authority by a collective body doesn’t necessarily require pinning it down to individual actors I believe. You might just exchange the whole team if necessary.

But I’d love to hear more perspectives on this.

3

I’m happy with the process as it is currently.

3 Likes
4

This mechanism exists. The Moderation team was created via the RFC process and is subordinate to it. If an RFC is accepted that changes the membership of the moderation team, the membership will be changed.

Ryan

9 Likes
5

I’m proud of this team, happy it exists, and I think it is doing a good job. The introduction of effective moderation always has detractors, which is fine and good. However, effective moderation is a necessary part of an effective community. I trust the moderation team to do a good job, and ultimately the NixOS Foundation to step in if it ultimately needs to, which I’m sure it won’t.

Overall, I’ve heard extremely positive feedback about the moderation team, and I can imagine folks who have skirted the edge on being a positive contributor to the ecosystem may feel threatened by it. That is normal, too. If parts of the community are unhappy with moderation or other processes / outcomes of the project, it is always possible to exercise the four freedoms of open source and create the project and community they want.

28 Likes
6

I also want to join Graham on this wholeheartedly.

Moderation team is doing a difficult job while doing also their own job (contributors, work, etc.) and I am deeply respectful of the work they are doing.

Not everything is perfect and things can be frustrated, but, I believe this team is a team with whom you can discuss, talk over things, disagree on things and that is a very important property, therefore, I can totally imagine what kind of stories, situations, can lead moderation team to do difficult decisions and I support them.

Thank you again to the moderation team for doing all of this work, it’s extremely hard to convey the importance of this work w.r.t. to its visibility.

22 Likes
7

Thank you for your opinions!

So far that doesn’t sound like my concerns about accountability are shared by the forespeakers.

This is interesting, because abstractly speaking an RFC is probably the only built-in / left over mechanism of accountability that there is.

Leaving the constraints of reality aside (day job / part time / etc), I wonder wether it’s a fundamentally good idea to not govern the community through the ideas of checks and balances, but reactionary in light of the emerging need for moderation. And then yet again reactionary, when we see it (possibly, even likely) fail in the future.

I can see in the RFC process and non-codified in the NixOS Foundation a last resort remedy, but even after individual preferences, these hurdles to effective accountability are high and an imbalanced system remains.

In my opinion, Actual Accountability is an integral part of the sutainability of the systems design. There are sure better ways to promote it than my concrete Issue’s request.

Happy to hear some ideas in that direction, too.

2 Likes
8

So if I get this correctly then your point is that “hurdles to effective accountability are high”. Now the question is of course: How would a more effective accountability look like for you?

Because, if I understand correctly here, too, the counterargument is that too low a standard for accountability is prone to abuse and exposes people to personal risk. And moderation does require a certain protection, first to find people to do it, and second to be effective.

And we cannot deny that there is at least a process in place.

1 Like
9

I’ve conveyed an idea to the team on different channels, but also want to disclose for more general scrutiny by the interested audience. It looks like:

  1. make a moderation entry log each time the team changes (just as for the ban log) and state the current team composition. → This allows the general public to audit diversity of opinions within the Mod Team and to better assess how much of their trust each individual contributor places into the people involved. It also allows to find tendencies and correlations and thus harden or dismantle an initial suspicion about bias (which I can’t quite get rid of based on the overall context).

  2. Slightly expand on the ban context, without creating too much overhead that would be otherwise unreasonable to the team. Limit to 200 chars each of the following mandatory log contents:

  • Accusation (anonymous)
  • Counterpledge (by the accused)
  • Decision (by the mod team)

The latter is very cheap to implement and not an unreasonable process clarification but very effective in preventing power abuse (or even only suspected power abuse), which the current model is prone to.

5 Likes
10

Point 1 is already effectively implemented through https://github.com/NixOS/nixos-homepage/blob/23620893fec32b3e87c84c01353fecde0314f6f2/community/teams/moderation.tt

3 Likes
13

I think the very conduct of this thread points out (yet again) that there is (so far) no unequivocal, visible and serious commitment to accountability and checks and balances, so calling for transparency might almost seem visionary, at this point.

Very unfortunately.

3 Likes
14

I don’t know what you hope to accomplish by repeatedly asserting that there are no checks and balances. The team has been very clear about this: they are subordinate to the RFC process. The community can use RFCs to hold them accountable and impose any checks and balances the community can get through the RFC process. If you have a specific proposal, make an RFC for it.

If you do make such an RFC, and if it is accepted, and the mod team for some reason refuses to abide by it, then it’s time to say oh no, there is no accountability. But the rules of the game seem pretty clear to me right now, with no evidence that they won’t be followed.

13 Likes
16

I don’t know what you hope to accomplish

Quite obviously: that there will be over time an emerging insight that accountability is in the best self interest of the team.

The success of a policy-driven RFC is quite low, in my opinion, due to the incentive structures of RFCs. Realistically, accountability seems to be an issue that can only resolve out of the team itself, when it is realized that it is not enough to mobilize a few with the team’s decision, but the vast majority of silent onlookers, and in doubt keep a cool hand and head.

And, despide the staked word, I hope nobody takes this advocacy (naivly) negatively, that’d be a missed opportunity to improve, and not in accordance with my stance.

It would be refreshing to hear any sort of commitment to improve, if a clear and serious commitment to accountability is already asking for too much. Tiny steps are good steps, too.

2 Likes
19

I’ve been overjoyed at the adoption of a code of conduct, but personally, after having had very bad interactions with member of the moderation team, where they unequivocally treated me against the terms laid out in this code of conduct, and without any sort of transparency as what happened to the member that did treat me this way, I also have concerns.

With this is mind, I do have to agree that there is an accountability issue. As much as I am glad that we finally have a code of conduct, as much as I don’t think there has been any lines crossed in the adoption of this code of conduct that shouldn’t have been crossed, I still have to point out that there are voices on both sides here calling for more transparency into the process.

At my most cynical, I fear without this transparency, the code of conduct will be little more than for show. For the community to be able to hold the moderation team accountable, we need more insight into their process.

At the same this, we need this insight in a way that doesn’t put any potential victims at risk.

And even if this accountability technically exists, having to read through several RFCs to find out the details isn’t sufficient, without discoverability they don’t exist.

And yes, the lack of insight into whether or not a person who treated me intolerably is still part of the moderation team does heavily discourage my continued participation in the project. Since said member asked what I even do for NixOS, I’d like to put this into context.

In October, I reviewed 160 nixpkgs PRs and opened 43 nixpkgs PRs, of which 37 where merged, and 3 still are on hold, as I have little intention of finishing them.

In November, I reviewed 4 PRs and opened 5 PRs, so far, mostly just for my own software releases.

I’m personally probably permanently burned after this bad interaction, but hopefully you can pull yourself together so the next person isn’t.

7 Likes
21

Thinking that creation of the moderation team has changed the processes or the style of moderation is mistaken. The RFC, when compared to reality before it, effectively said that the people doing the moderation are given specific places for announcing team changes and bans, and otherwise continue whatever they were doing before.

I do agree that even the briefest summary of what moderators found the key unacceptable component of world-visible behaviour — have we ever had a ban purely for unacceptable PMs? — would be a useful addition to posting a general summary of moderation principles (call it CoC ot whatever), maybe even a more useful thing than CoC on its own (although it is useful to say that moderators agree that the general policy of enforcement aligns with some text).

3 Likes
22

Then Moderation Team: please provide the secret reason for the ban.

The Moderation Team can ban anyone without publishing who did the decision and why but anyone oppugn them has to open an RFC?

And now, in this thread, The Moderation Team just hidden all posts they don’t like without a reason?

My account here is not linked to any personal account because I don’t want to be a target of GOVERNMENT CENSORSHIP but it seems this can also save me from the NixOS Moderation Team.

I understand I should assume good faith. But trusts is hard to establish. If a controversial is made and then clarified later, then I can trust the team. How can I trust someone who keep banning people in black box?

4 Likes
24

It’s not a secret: Update moderation-log.md · NixOS/moderation@e9d67b7 · GitHub

Quoting myself here from elsewhere: “Posts that get flagged by Discourse community members with a sufficient trust level get automatically hidden at first, although the moderators eventually have the final say (there is a review queue). Please don’t make any bad faith assumptions.”

We are currently hiding all posts that are off-topic or inflammatory.

6 Likes
25

Yes, I actually have to agree that my last post was somewhat off-topic. :+1:

27

Glad to know that the reason is not a secret. It’s a big step toward a more transparent administration and the Moderation Team gets more trust from me. Unfortunately the reason is published late and only by a team member in a reply. This can’t dispel all misgivings though so I still hope to see more details.

1 Like
28

The reason is not published by default because it’s a public log and people that got banned might get into trouble in their personal life/carrier i.e. if they maybe in a few years later they try to apply for a job. We always tell the person in question in private why they got banned. In this case we made an exception because the person explicitly allowed this.

19 Likes