Code of Conduct

@boofy that question came up, but it was decided that we need a separate policy for federation rather than trying to include it in the code of conduct for our members. So, look for upcoming convos on that topic in the Community WG.

That's good to know and was a question of mine. Thanks!

I was part of the working group but couldn't find a draft to comment on it at all before it was apparently approved. The process for participation if you couldn't be on the calls was really unclear, and as far as I can tell this draft was never actually submitted for any kind of discussion except during a call. At the very least, I never saw a link to an actual draft document before now.

Unfortunately, this has a LOT issues and I think it's going to need at least one more draft. The choice of "citizen code of conduct" as a base was somewhat inappropriate (it's relatively outdated, and a lot has been learned since it was drafted on what does and doesn't work for online CoCs, which don't really resemble ones for in-person events for what are mostly important reasons)

The reporting guidelines should really say something about confidentiality (like, whether, the reporters can expect any) and "create a thread in the community working group" is an exceptionally inappropriate way to handle potentially sensitive situations that can involve multiple people's personal information. It also says nothing about who will receive the reports, what will be done with them, when they can expect any response (if ever?) or when they might expect any action to be taken.

"Unwelcome sexual contact" being moved to a subset of harassment and only prohibited on repeat offenses is a red flag. It's also a red flag that much more space was devoted to the prohibition on product advertising than sexual harassment.

"Reconsideration" says to check the reporting guidelines for how they will be handled, but the reporting guidelines don't say anything about reconsideration, or describe the process initial complaints are handled by either.

"Not safe for work" is really unclear, this should explicitly say which things it covers (presumably sexually explicit content, nudity and violent imagery?).

There's more, but I'm trying to be somewhat brief here. I really wish the process for contributing to this had been clearer and at least some of it took place asynchronously -- requiring video calls for participation can be really alienating to marginalized people -- who are the people you really needed participating in this process.

I appreciate that this has been a long time in development, and that a lot of work went into this, but in my opinion this isn't ready for a vote.

Hey Melody,
A number of issues you raise definitely require more consideration, but a quick comment on the process that large parts of it did happen asynchronously. The most recent iteration was this check poll that closed eight days ago in the Community Working Group that called for comments to be left on the document, so there was a great deal of work contributed by many people that was integrated into the document in the synchronous work session that @emido @manuelabosch @mattnoyes @tomelliott @jakebeamish and I had this morning.

As to a number of the items you raise, this code of conduct is a starting point that requires the development of a bunch of different policies that will define the processes mentioned here. There is definitely still much work to be done, and my intention is that, once we have a CoC adopted, I will be redirecting energy towards the formation of the Ops Team that will tackle that next round of work.

So, my sense is that, having been run through the agreed-upon process, this is a starting point that can then be amended as we continue the next levels of work. Happy to discuss further...

Hi Melody. I couldn't make the call either but was able to leave some comments the draft. We should definitely look into how pre-propsal drafts are spread out there (maybe as a separate thread inside the Governance WG or Community WG) as it seemed like considerable effort to get contributions on the pre-prosasal draft was done but that you didn't see them is of concern.

You raise some valid points. The best thing to do from a process point of view may be to Vote BLOCK on the proposal which will raise the threshold for it to pass and allow for other members to consider their votes along with your concerns.

Possible that this could lead to an additional revision period for the proposal.

If not there the bulk of the policy/process work still has to be done for how the CoC is actually implemented that will be done by the Community WG which is open to all members to join.

Unfortunately, the issues with the reporting guidelines are critically serious. If there's not a well-developed process for handling reports, the code of conduct is actually worse than not having one. In order for the CoC to matter, it has to be enforceable, and the members need to trust the anti-abuse team. When nobody knows what that process is, whether reports will be confidential, who will be handling them, or what they can expect to happen if they do, that's impossible.

Nobody active in this space, who needs codes of conduct to participate in a group like this, will read this code of conduct and be comforted. Our reputation might already be sunk for people who care about this kind of thing by having gone so long without one, but taking over a year to develop one and then releasing one that misses really important details about what makes a code of conduct matter will be worse than taking longer to approve one.

If this goes forward as-is, I will have to vote to block, but I'm concerned that there weren't more obvious calls for participation and a clear link to the document being worked on (now that it's been pointed out, i can see it in the poll, but the tiny "here" link in the poll really should not have been the only link to the doc ever shared)

At work, so will try to respond further to the operational elements you've raised soon, but do want to point out process-wise that that check poll was a specific call to action to review the document in its title, that the document was linked in several additional polls as the process progressed (you may review the past decisions in the working group).

Additionally, I and several other folks tooted on the instance encouraging participation at multiple points in the process as it progressed (example), so, while there's always more to be done in terms of getting the word out, I stand by the outreach that was done to collect feedback as a good faith best effort.

Here is a very preliminary draft of reporting guidelines copied from Django's. Please feel free to edit: https://docs.google.com/document/d/1pZTXsGuWNTe5X6z_DytJZTbjOyJUT4HlvPPnfmbuPEw/edit

About process, if you look at the code of conduct thread on Loomio, you can see the series of polls and decisions going back a year. If you search the hashtag #codeofconduct in social.coop you can find many of the posts urging people to participate. I think, though, that the question of engagement raised here underscores the need for a more direct, proactive approach along the lines of the Member-to-Member network I referred to here: https://www.loomio.org/d/9X1PfHGg/reaching-out-to-less-active-members

For now, it seems best to focus on editing the CoC and preparing good reporting guidelines.

Hi Melody, can you recommend a good code that is suited to online communities?

I agree with this: "create a thread in the community working group" is an exceptionally inappropriate way to handle potentially sensitive situations that can involve multiple people's personal information." I think we should drop the sentence about creating a thread, and replace it with a link to the Reporting Guidelines.

"Unwelcome sexual contact" being moved to a subset of harassment and only prohibited on repeat offences is a red flag. It's also a red flag that much more space was devoted to the prohibition on product advertising than sexual harassment.
-- This seems not exactly right as a description, but the points are right: we only mention "unwelcome sexual attention" not "non-consensual sexual contact" or assault. We should add this. I don't see the only prohibited on repeat offences... The CWG is empowered to act on the first incident and we say, "Following serious or multiple infractions..." they can raise the punishment.

"Reconsideration" says to check the reporting guidelines for how they will be handled, but the reporting guidelines don't say anything about reconsideration, or describe the process initial complaints are handled by either."
-- The reporting guidelines are just a rough version, but they include a Reconsideration section that needs to be concretized. Again, it would be great to have better language and procedures.

Threads don't nest more than one level deep, so I'm going to address multiple asks here:

In terms of a good code for online communities, it really varies a lot depending on the type of community, I don't have a cookie-cutter suggestion for a platform cooperative spanning 4+ platforms, and this was always going to require some custom work. I am not sure which I'd have suggested we use as a base offhand, but I'm also not recommending we start over now.

The "only on repeat offenses" issue is because it's a subset under harassment now, and harassment is defined as "continuing to interact with a user after having been asked to cease contact" (the "after being asked bit, being key") -- the implication is that unwelcome sexual attention is not prohibited until you've asked them to stop at least once. On principle, this is something of a consent issue,

Re: the reporting guidelines, the document doesn't link to them, so I assumed the extent of the reporting guidelines was section 4 of the CoC labeled "Reporting Guidelines". I am now looking at the separate document you linked that has a much better-defined process, I think we should aim to provide a summary & a link in the CoC, or aim for an easier-to-describe process and save most of the nuts & bolts details for more "internal" documentation.

I'm not sure how I missed any prior drafts of this that were circulating, or other asynchronous decisions, but I am not denying that a good faith attempt was made to call for participation. I'm clearly not looped into the right communication circle somehow to have missed things, but I was trying to pay attention to this and participate, and it's only now that there's a new thread discussing it that I finally got a chance to see the working draft. Not laying any blame here, perhaps I'm struggling somewhat with Loomio's UX, but I do wish I had been involved earlier.

Tacitly, this code of conduct refers to instances of Mastodon. I assume it should also refer to conduct in Loomio, in the wiki - and in any future platform? Do folks agree? Should this be made explicit?

It is meant to apply to everything, but I agree it's contradictory/confusing that it for instance, refers to using the mastodon reporting tools to do reports and then says that it applies in all community spaces in the scope section.

Thanks Melody, I had missed the 'scope' section, but noticed references to 'instance', which made me think in Mastodon-specific terms. Perhaps occurrences of 'instance' might be checked for clarity?

That should be made explicit, maybe say "social.coop and related spaces (Loomio, Riot/Matrix, etc.)"?

I think that the points re: enforcement are quite well made. I do know from experience that one can have issues with CoC if the handling process + points of contact aren't clear.

Having said that - I have decided to retain my agreement with the proposal and would encourage @meltheadorable et al to bring a proposed amendment to the working group as soon as possible. I have no doubt they will have the support of social.coop members.

Thanks to everyone again for their hard work (and hard questions!). :)

I intend to hold my block for now. I genuinely think that when the problems are large enough, having a broken code of conduct is worse than not having one. I'm not expecting perfection, but the lack of a well-specified reporting & handling process is a fundamental flaw. It's not a "fix it later" kind of situation.

I'm not going to nitpick like, the language around "authenticity" (which i think is occasionally problematic) or the list of protected classes, I wouldn't be voting block if I had only minor or superficial concerns, but almost literally the entire point of a code of conduct is to put in place the reporting and handling process that has been left somewhere between mostly unspecified and actively counterproductive in this draft.

I don't oppose your block, but I want to clarify two points:

1. I agree 200% with the whole emphasis that a truly reliable, thought-out reporting/handling process is absolutely essential.

2. I happen to disagree with the idea that such a process is part of a "Code of Conduct".

I think the two parts (the Code and the enforcement) go together strongly enough that you can't build them entirely independently. But I think they are still different things. I'd evaluate a CoC based on how amenable or not it is to a real-world reporting/handling process, and I wouldn't consider it acceptable to have a CoC without the process to go with it.

You might disagree, but it really is essential to a functioning code of conduct. Elements of an effective code of conduct (best practices come to by long struggles against ineffective ones) include, but aren't necessarily limited to:

1. Information about reporting, including contact information
2. Information about how it might be enforced
3. A clear separation between behavior that is considered completely unacceptable and more general guidelines about how to interact with the platform/etc
4. Some specific examples of things that aren't acceptable

Additionally, it's becoming increasingly important for many projects to have

1. Some assurances about how much and what kind of confidentiality to expect around reports
2. Clarification that the code of conduct applies to project leadership and a responsibility to enforce their violations as well.
3. A way to close the "didn't happen here" loophole to prohibit participation based on sufficiently toxic outside behavior
4. Some information about whose interests are going to be prioritized in the event that there are seriously conflicting wants or needs.

Technically, the draft under consideration has some form of all the strictly required elements, but the reporting and enforcement are broken because they either make suggestions that are toxic to social harmony (like encouraging people to join the community working group on loomio and publicly posting about the violation there) or not functional for most of the spaces it's supposed to cover (use the mastodon reporting feature). Even if the mastodon was the only thing covered by the code of conduct, there's one more issue with that being the only reporting mechanism: there may be very imperfect overlap between the group of people that the instance trusts to handle abuse complaints and the group of people who have (possibly incidental) access to mastodon's reporting infrastructure. Especially in a cooperative where server and administration access may need to be spread broadly to share the work.

Beyond that, a cooperative often needs to err on the side of more structure than usual, rather than less -- we can't lean on formal hierarchy to direct power, and if we don't substitute good process to do that work, informal hierarchy will establish itself instead. For something as important as the process for handling anti-abuse complaints, harassment, doxxing, etc. that can't be allowed to happen. The code of conduct does not necessarily need to describe the entire internal process for reporting and enforcement, but it does need to explain how to initiate the process and provide some summary of expectations for what a person reporting can expect from outside of it -- in order to describe that the internal process needs to be there first.

Okay, I'm not sure if it got lost in the shuffle (haven't been that active here), but I encourage everyone to take/adapt/use stuff from https://wiki.snowdrift.coop/community/conduct

We read all the CoC's and wrote ours to incorporate all the best ideas specifically for an online co-op community. We're working out the best uses of flagging in Discourse for enforcement still, but the CoC is complete. For that matter, any feedback/concerns are welcome, but I feel it's got value over what I've seen otherwise.

Cheers

Flagging is interesting - how might it work here? Our idea up to now has been "contact [designated person] in the Community Working Group" -- how would flagging work?

flagging requires some actual tool to work that way. Discourse is closest (it allows a flag to hide a post, sends a note, then if the post is edited, it can be reposted and the flag cleared). Broadly, there's a huge topic around all of this, and I just posted a note to our forum about how this relates to Restorative Justice.

Anyway, it all depends on the tool. Discourse has its flagging system (which we've tweaked and where I have a bunch of notes about feature requests needed to make it truly optimal). GNU Social type systems I'm not familiar with the moderation options. The ideal would be to get such flagging built-in (flagger asked to identify the problem, ideally offer suggestions, flagged posts get hidden, get reposted once fixed, mods get involved only if this doesn't go smoothly). I don't know what that would take, but it's a lot bigger than just finalizing the CoC itself.

Also, to be clear, CoCs for online text (and sometimes image) communities are not the same as in-person conference CoC's, and while I have thoughts about the latter, I've only focused on the former.

Aside from a well-designed flagging tool, here's the broad issue:

A) do NOT make all flagging be highly serious extreme violations — if addressing something requires privately getting a moderator, then people will hesitate to go that far, that needs to be available as last resort
B) offer both anonymous/non-anonymous and public/private options.

The person noticing a problem should be able to use good human judgment about whether this is appropriate to discuss privately or publicly or needs moderator help or not etc. Sometimes, public discussion of the issue is a good learning experience for everyone. Sometimes, making it public puts people on the defensive around saving-face or other social concerns and private is better. Sometimes the person doesn't want to interact at all with the violator so anonymous and/or moderator-help makes sense.

The goal is to provide a clear framework and what-to-do options while encouraging people NOT to feel helpless and NOT to always look to authorities to solve every case.

Thanks. Aside from flagging, which is a lot bigger than finalizing the CoC, can you suggest any other specific changes? The Snowdrift code, which I like, doesn't seem that different substantively. I mean, I could live with either one.

I don't know if it's worth getting into the weeds, but we put ours together not to reinvent wheels but because others weren't good enough in our view.

Obviously, my perspective is you should just go with the Snowdrift.coop one, perhaps tweaked appropriately. I know some go with the stuff you picked just because it's already more widely used.

I know others disagree on this, but I think the calling out of specific list of concerns (gender, ethnicity, etc.) is problematic. It always fails to mention some classes (e.g. occupation) no matter how long you make the list. We aimed to reference the topic of bigotry and prejudice which should be accepted as wrong in any scenario.

Our CoC is also more hard-nosed about good communication. "No condescension" is a higher standard than most CoCs because most have this idea that good people will never violate the CoC. In ours, the assumption is that because the point is restorative addressing of problems, we want to allow people to call out tensions from bad communication even when they aren't something really racist or anything.

There are a lot of other reasons I prefer ours, but instead of ramble here, I'd be happy to discuss any specific thoughts or questions from anyone. If it (or any of it) is to be really considered here (which I'd be happy to hear), it would make sense to have others weigh in and overall be sure everyone feels good about it. I just don't want to be someone jumping in and redirecting work that already has been going on, even though I think getting this right and optimized early is a big deal.

[side note: we had quitter.se account and it's down now, we may likely move to Snowdrift.coop using social.coop and from our perspective, it would be ideal if the same CoC were used, but that's our bias].

Addendum: if I thought the existing proposal was setting things up to be tragic, I would push back harder. I think it's just not as good and won't cover as many cases as well and won't do as good at pushing everyone to be their best. But I think it's adequate at least.

I'm having a hard time finding things in this LONG topic here, but trying to put this in the right place.

I appreciated notes I saw from @meltheadorable and drafted a first improved version (work-in-progress of course) of a better updated conduct-enforcement page for Snowdrift.coop along with some extra clarity about what is and isn't enforceable. I also added an item to our CoC about respecting the core mission and co-op values.

https://wiki.snowdrift.coop/community/conduct now links to https://wiki.snowdrift.coop/community/conduct-enforcement

I'm sure there's room for critique, but anyway people here might be interested in seeing the updates. As always, I offer this as one way of sharing ideas that can be used to whatever productive extent in developing the social.coop CoC system.

Cheers

Hi - I am new to this community and still determining if I would like to join. That depends in part on this CoC. I do feel as though it is clear that "unwelcome sexual attention" is not limited to repeat offenses and is actionable upon the first offense. Please correct me if I'm wrong.

For me, I don't see this CoC covering casesas follows: When a community member reponds to someone on another instance to say things that are a) ad hominem attacks b) mean spirited-type behavior c) outright mudslinging and unfriendliness. If I were going to join this community, I'd want guarantees that if another member were to try to back up their political point of view by outright unfriendliness and mean spirited language, the CoC would be clear that they should be reprimanded. This doesn't cover that for me, so it's hard as an outsider to want to join this community. I'd be willing to talk more about this. Thank you.