Talk:Code of Conduct/Archive 4

Process completed, CoC Commitee constituted

The Code of Conduct Committee bootstrapping process has been completed. Some intermediate updates were posted in phab:T159923. Starting today, the Code of Conduct for Wikimedia technical spaces is enforced by the new Committe formed by Amir Sarabadani (Ladsgroup), Lucie-Aimée Kaffee (Frimelle), Nuria Ruiz (Nurieta), Sébastien Santoro (Dereckson), and Tony Thomas (01tonythomas). Congratulations to them, to the additional five auxiliary members (Απεργός, Léna, Florianschmidtwelzow, Huji, Matanya), and to everybody who contributed to this process! Qgil-WMF (talk) 08:19, 20 May 2017 (UTC)

Drafting phase data

Here is some history about the drafting phase, captured for posterity. To put these numbers into perspective, as of today Special:Statistics reports 1,420 active users in MediaWiki.org in the past 30 days.

Code of Conduct

• 449 edits
• First edit: 2015-07-16 (1 year 6 months 22 days ago) by Mattflaschen-WMF (after a session scheduled in the Wikimania 2015 Hackathon, original session notes).
• 40 unique editors (1 IP address).
• 88 watchers
• Pageviews Oct 2015 - Feb 2017: 10,680 (628/month) -- Oct 2015 - Feb 2017 only due to page renames.

Talk:Code of Conduct

• 2,718 edits
• First edit: 2015-08-07 (1 year 6 months ago) by Bawolff (a suggestion for improvement).
• 142 unique editors (6 IP addresses)
• 147 sections (subsections not included)
• 21 voting rounds
• 88 watchers
• Pageviews: 38,137 (2,243/month) -- Oct 2015 - Feb 2017 only due to page renames.

Qgil-WMF (talk) 11:24, 8 March 2017 (UTC)

To add some context, for posterity, of the 451 edits to Code of Conduct, 313 of those edits, about 69%, came from three editors, all of whom work for the Wikimedia Foundation.

For "Talk:Code of Conduct" (this page), of its 2,730 edits, two Wikimedia Foundation staffers contributed 943 edits, or about 35% of edits. These two editors have over 200 edits each ahead of the next highest contributor by edit count. --MZMcBride (talk) 14:03, 13 March 2017 (UTC)

MZMcBride How does that differ from other policy discussions in this community (MediWiki) and other communities of the Wikimedia Movement? Is it uncommon for one or two editors to "drive" the discussions along, and make (most of) the changes (even the changes proposed by other editors) on the policy page? Chico Venancio (talk) 14:27, 13 March 2017 (UTC)

Hi Chico Venancio. You can use the tool I linked to look at the editing profile of other pages. For example, w:en:Wikipedia:Civility: <http://vs.aka-online.de/cgi-bin/wppagehiststat.pl?lang=en.wikipedia&page=Wikipedia:Civility>. If you're interested in similarish local pages, you could look at <http://vs.aka-online.de/cgi-bin/wppagehiststat.pl?lang=www.mediawiki&page=Bug+management%2FPhabricator+etiquette> and friends, I suppose. --MZMcBride (talk) 23:06, 13 March 2017 (UTC)

It would also be nice to compare with other policies to see how many required that people not agreeing with a pre-determined outcome be blocked. Nemo 06:44, 6 April 2017 (UTC)

Comparison of different models for review?

This code looks good, including fine elements from some of its sources. Thanks to those who worked on it. Sj (talk) 13:37, 8 March 2017 (UTC)

Has there been consideration of any sort of A/B testing of different models for reviewing different cases? (There's an argument against that you want all cases to be treated 'equally', but even with a fixed group of reviewers that isn't the case; it is unclear whether a fixed, short set of rules is more or less equal than a fixed set of reviewers with their own internal inclinations beyond those rules).

I ask because some of the most scalable models for conduct have some sort of ad-hoc, rotating, or randomized selection of a small group to make a decision. If you support appeals you might want to make the appeal process to a fixed group more carefully chosen; but testing a more ad-hoc model for the first pass might increase speed of response and reduce bureaucracy. Sj (talk) 13:37, 8 March 2017 (UTC)

New members in the Technical Collaboration's Community Health group

The Technical Collaboration team has a Community Health group who handles reports on behavior in Wikimedia technical spaces. It is also the group that is now tasked with the bootstrapping of the first Code of Conduct Committee (more about this later today). Andre Klapper, Rachel Farrand, and Quim Gil have been doing this work for several years. In order to respond better to the role that the CoC assigns to this group, we have decided to have five members and one auxiliary member, to mirror better the procedures of the Committee. Developer advocate Srishti Sethi and community liaison Chris Koerner are joining the group, while Nick Wilson (Community Liaison) has been appointed as auxiliary member in case of absence or conflict of interest by any of the group members. These additions bring other perspectives about technical contributors in various areas, with various backgrounds and types of experience. Thanks to them we will be able to put more time, experience, and opinions in the search and selection of candidates for the first Code of Conduct Committee in the next weeks.Qgil-WMF (talk) 09:47, 9 March 2017 (UTC)

Making the Code of Conduct more visible

What would be a good way to make the Code of Conduct more visible to members of the community? For example, we could put a link in the sidebar or footer. Mattflaschen-WMF (talk) 16:24, 20 March 2017 (UTC)

See also After Code of Conduct is approved, Etherpad footer should link to it. (T136744). Mattflaschen-WMF (talk) 16:27, 20 March 2017 (UTC)

IMHO footer - feels similar to Privacy Policy and the Disclaimers. --AKlapper (WMF) (talk) 21:42, 20 March 2017 (UTC)

We should notify the CoC to anyone creating a new account to be used on gerrit/phabricator/wikitech etc. We should also make it very clear which spaces this CoC applies to. GLavagetto (WMF) (talk) 07:13, 24 March 2017 (UTC)

The spaces are listed in the first sentence. Basically, any tech event in the Wikimedia community, plus "Wikimedia technical presentations in other events", plus the explicit list of online spaces given (note there is no "such as" on the online list). Mattflaschen-WMF (talk) 06:27, 7 April 2017 (UTC)

I would also very much appreciate it, if we can link this somehow during the phabricator signup phase. —TheDJ (Not WMF) (talk • contribs) 13:47, 19 April 2017 (UTC)

It needs to be added as a condition of use in the notice on each IRC channel it applies to, i.e. the table of technical channels linked in the CoC document. Implicitly all of the users of these channels have now agreed to comply with this policy. --Fæ (talk) 11:18, 26 April 2017 (UTC)

See also T165540 - Add CODE_OF_CONDUCT.md to Wikimedia projects. Qgil-WMF (talk) 13:34, 18 May 2017 (UTC)

I have Gerrit change 354612 to add a footer link on mediawiki.org and wikitech. I think we can get that deployed on or before 2017-05-25. --BDavis (WMF) (talk) 22:29, 23 May 2017 (UTC)

 Done this is now live. --BDavis (WMF) (talk) 23:38, 25 May 2017 (UTC)

Translation

Should the CoC pages be translatable? --Tgr (WMF) (talk) 20:34, 18 May 2017 (UTC)

This was asked today by one of the future Committee members. I'd say yes, of course! Although... a) please wait until the Committee is constituted on Saturday, so this can be their decision, and b) I would consider the possibility of adding a note explaining that the English version is the canonical one, in case of incorrect translations or outdated versions. I don't know whether there is already a best practice in place for translatable Wikimedia policies and similar docs. Qgil-WMF (talk) 20:39, 18 May 2017 (UTC)

The content of Code_of_Conduct#See_also should be probably be left out of the translation (there is not much to translate anyway). I am proposing a change to simplify it. Qgil-WMF (talk) 07:44, 19 May 2017 (UTC)

Yes, I think it should be. Re "b)", I agree it definitely needs such a note. Unfortunately (I understand not everyone speaks English), the English version is the only version that was approved, and there is no guarantee the meaning exactly matches a translation. This is the same as e.g. the Privacy policy (see part, "Esto es una traducción. Para la versión auténtica y vinculante de este documento por favor vea el documento original en inglés.")

I don't see any reason to exclude anything from the translation. Mattflaschen-WMF (talk) 04:44, 26 May 2017 (UTC)

Cleaning the "See also" section

Code_of_Conduct#See_also grew organizally as we were discussing the CoC draft. Thinking of someone landing for the first time on that page (to learn about the CoC or to report a problem), I think the current collection of links can be more confusing than informative. Here is a proposal to simplify it:

• Phabricator etiquette, applicable to any incidents in Wikimedia Phabricator.
• Friendly space policy, applicable to any incidents in Wikimedia events.
• Expected behavior, a spin-off of this Code of Conduct complementing the Unacceptable behavior section.
• m:Code of conduct, an extensive collection of related resources across Wikimedia.

We should check that all the links removed are mentioned in m:Code of conduct.

This section is not part of the Code of Conduct and it didn't go through the review that the canonical sections went. For this reason, I don't think we need to treat it as a CoC amendment in order to discuss it, agree on it, and edit accordingly. Qgil-WMF (talk) 08:01, 19 May 2017 (UTC)

Future of conduct-discussion@

During the CoC draft discussion period, conduct-discussion@wikimedia has been available for people preferring a private channel to send their feedback. Now that we will have a Committee in place, it doesn't make much sense that Kalliope (WMF) and myself keep being involved. If we keep this email address, it should point to the Committee members. Then again, we have already techconduct@wikimedia, which will reach the Committee members already, so maybe the simplest is to close conduct-discussion@. Thoughts? Qgil-WMF (talk) 08:06, 19 May 2017 (UTC)

Can a permanent auto-reply be setup to indicate the discussion is mostly over, and if people still have private/anonymous comments they'd like to leave about the CoC they can send it to the committee? Legoktm (talk) 18:00, 19 May 2017 (UTC)

Since the account was created, it has received only three genuine messages with private feedback (one of them actually misplaced, since it was not about the CoC). The last one was received on February 2016. All the rest is spam. I think it would be simpler to to just delete the account. Qgil-WMF (talk) 09:06, 22 May 2017 (UTC)

Suggest forwarding to techconduct@, since that is the email address given for anonymous amendments, and that is really the only text to discuss now. Mattflaschen-WMF (talk) 04:50, 26 May 2017 (UTC)

This solution makes sense. If nobody objects, I will proceed accordingly. Qgil-WMF (talk) 08:43, 26 May 2017 (UTC)

Chair of the Committee

The Code of Conduct Committee bootstrapping process is about to be completed (I just posted an update). A step not included in this phase that will need to be handled by the newly appointed Committee is the selection of one of its members as chair. Code_of_Conduct/Committee#Selection_of_new_members simply says: "After new Committee members are chosen, the Committee selects a chair." I'm leaving this note here to avoid ambiguities about who does what. :) Qgil-WMF (talk) 09:01, 19 May 2017 (UTC)

The committee has elected User:Dereckson with unanimous vote to be the chair :) Ladsgroup (talk) 21:31, 8 June 2017 (UTC)

Linking to Code of Conduct/Committee members

There is one minor change to the CoC pages that I wanted to do before the Committee was constituted but... I forgot. Code of Conduct/Committee members was created in the past weeks as we were bootstrapping the Committee, and currently is basically an orphan page. A couple of suggestions:

• In Code of Conduct/Committee, add a link "The Code of Conduct Committee is a team of five trusted individuals..."
• In the same page, under Conflict of interest, add a link "Auxiliary members will join as needed, to replace recused members."

After some consideration, I would not add the page to the navigation infobox, as this would give more prominence than needed to the people forming the Committee at a point of time.

Considering that this is a minor detail informing an obvious question (who integrates the Committee), I think it could be added without having to wait until the next review phase. Qgil-WMF (talk) 09:19, 22 May 2017 (UTC)

I agree in this specific case. I don't have any objection here. Such links shouldn't be added to try to define (change the meaning of things), though, without an amendment. Mattflaschen-WMF (talk) 05:19, 26 May 2017 (UTC)

I am also in support of both changes. Huji (talk) 04:46, 28 May 2017 (UTC)

 Done Thanks for the suggestion. Mattflaschen-WMF (talk) 03:55, 9 June 2017 (UTC)

Suggested Amendment: Include GitHub

For the amendments process: GitHub should be added as a covered space, for GitHub projects connected to the Wikimedia community. Although it's not a primary space, there are some repos only on GitHub, and even on the Gerrit ones, occasionally people will submit a pull request there (on the GitHub mirror).

This was an accidental oversight. Mattflaschen-WMF (talk) 04:35, 26 May 2017 (UTC)

Some projects also use github as primary source of development, some repositories of WMDE's Wikidata team and almost all of ORES is stored in github and mirrored in diffusion. I think CoC should apply to those too but how we can clearly define such projects, I don't know. Ladsgroup (talk) 21:36, 8 June 2017 (UTC)

Projects maintained by Wikimedia organizations? --Tgr (WMF) (talk) 21:44, 8 June 2017 (UTC)

Organization in the technical GitHub sense, or just meaning affiliates/employees/members of a Wikimedia organization (in the normal sense of the word) maintain the project? I would interpret that to include other organizations like Wikimedia Germany, but perhaps that should be clarified. Mattflaschen-WMF (talk) 04:01, 9 June 2017 (UTC)

This needs to be worded with extreme care so as not to extend to repositories that are connected to different parts Wikimedia community (local projects, mainly). My suggestion would be to keep it at most for https://github.com/wikimedia --Strainu (talk) 11:50, 30 August 2017 (UTC)

Projects maintained by affiliates/employees/members of a Wikimedia organization (in their official capacity), yes. (Ideally such organizations should use Github orgs to make their project ownerships transparent, but whether someone forgot to do that should not change whether the CoC applies.) --Tgr 05:44, 5 September 2017 (UTC)

It should include any GitHub repositories that are part of the Wikimedia ("Wikimedia" in the sense of the movement, not any particular organization) technical community, whether the repository is developed by volunteers or staff (or both). This would certainly include anything in https://github.com/wikimedia , https://github.com/wmde , etc. But I agree with Tgr it should go beyond that, so things that are truly Wikimedia technical work (but perhaps not in a formal GitHub organization) are not accidentally omitted. Like everything else, it should be space-based. If a repository is covered, all activity in that repository is covered. Mattflaschen-WMF (talk) 01:51, 13 October 2017 (UTC)

Feedback from Vienna hackathon

Hi,

Some CoC committee members met at the Vienna hackathon. This was an opportunity to see each other in real life, and to discuss how we can organize workflows to handle cases. We also discussed our communication tools, see this and this slide.

A point of action is to create a private wiki to archive the cases and coordinate procedures.

During this hackathon, some concerns have been voiced about ideas to improve the CoC, but too late to discuss these in the initial round. The CoC is a living tool, it can still be amended and improved. So if you wish to organize a consultation in your community/team/group about ideas to improve the CoC, it's still time, and it will be still time in the future.

--Dereckson (talk) 09:01, 26 May 2017 (UTC)

One thing that I'm greatly interested in for now is to add a link in mediawiki repos about the CoC. It's being discussed in phabricator. Ladsgroup (talk) 10:11, 26 May 2017 (UTC)

(For reference, that's T165540 --Tgr (WMF) (talk) 21:43, 8 June 2017 (UTC))

Blog post

You can find an overview of the Code of Conduct at the new blog post. Mattflaschen-WMF (talk) 04:39, 9 June 2017 (UTC)

Suggested Amendment: No impersonation

We should prevent impersonation, including trying to make people confuse you with others.

Suggested text (to add to Unacceptable behavior):

"Impersonation, including inappropriate patterns of behavior that cause people to confuse one’s identity with others". Mattflaschen-WMF (talk) 23:13, 24 July 2017 (UTC)

What real-world problem is this intended to solve? --MZMcBride (talk) 13:47, 30 August 2017 (UTC)

It's not a real-world problem, thanks for highlighting that fact. The amendments we see proposed recently seem to be worryingly extending the reach of the CoC, extending the "WMF policed zone" way beyond Wikimedia projects, let alone Wikimedia technical forums.

I suggest starting informally with WMF employees, who are routinely seen using official and volunteer accounts in the same forum. This is not only confusing, but gives a false impression of consensus or who is acting as an "unpaid volunteer" or a "paid volunteer". You don't need a CoC amendment for WMF employees to behave well online. Thanks --Fæ (talk) 14:21, 30 August 2017 (UTC)

I'm not sure what you mean by "real-world" as it can be used to mean, "something that has happened" or "something that is plausible". I assume the latter in my reply. Let's say someone created an account on-wiki with a name like "Ckoerner (WMF)" (note the lowercase "k") and went around trying to convince people that Discovery was going to start using AltaVista for the Wikimedia search. The user name looks legitimate to folks unfamiliar, is similar to my own username for those who are familiar, and the intent is to sow confusion, disruption, and possibly tarnish whatever reputation I have. :) Folks could easily confuse this account with my own. Not only would I be harassed by this, but the community as a whole. I would have to spend time and energy working with admins to block the account and revert changes, chase after any conversations on mailing lists and talk pages to clarify that it wasn't me, allay fears that we're not doing what the account suggests, etc. The same could apply to any account, but as I'm replying my staff role I used example familiar to me. My interpretation of this amendment would make clear that such behavior is unacceptable by members of our community. CKoerner (WMF) (talk) 19:42, 30 August 2017 (UTC)

It should be impossible to create "Ckoerner (WMF)" - the AntiSpoof extension is supposed to prevent that. Legoktm (talk) 07:48, 31 August 2017 (UTC)

Since I created my volunteer account there have been two other accounts created with similar names. One remains an unblocked account. Ckoerner (talk) 19:44, 30 August 2017 (UTC)

Ok, so what does this have to do with a code of conduct? It's a simple case of abusive username. --Nemo 19:48, 30 August 2017 (UTC)

We don't have a global username policy, nor one in this community, that would cover abusive usernames. The only thing I was able to find was a single line mention on English Wikipedia that says basically "don't do that", but that does not address the behavioral concerns, nor any process for handling harassment as a result, nor is it applicable here. :) From a technical perspective AntiSpoof should prevent such accounts from being created, but that only handles on-wiki names, not other technical spaces, which the Code of Conduct would cover. CKoerner (WMF) (talk) 20:10, 30 August 2017 (UTC)

So, again, this is about the hypothetical problem that, because abusive usernames are considered too obviously abusive to even bother writing specific policies about them, something bad might happen without anybody bothering to intervene, even though no such case of neglect is known? --Nemo 20:15, 30 August 2017 (UTC)

In the past years my team has received complaints related to impersonation in Wikimedia technical spaces. We tried to intervene yet nothing could be actually done because we didn't have any policy or similar in place to solve the problem. In more general terms, impersonation is a frequent form of online harassment and I don't see why Wikimedia technical spaces would be immune to it. What is being discussed here is whether impersonation should be considered "unacceptable behavior". What is the benefit of not including impersonation as unacceptable behavior in the CoC? Will we be protecting potential victims of impersonation better by including this concept in the CoC or by not including it? Qgil-WMF (talk) 20:41, 30 August 2017 (UTC)

Yes, it would be harmful because it would encourage to handle in non-standard ways what is routine work on our wikis. By the same reasoning, we could end up adding wiki vandalism to the CoC and then people would go complain to the committee instead of clicking "undo" or "rollback". In the cases you mention where "nothing could actually be done", have you tried using standard wiki processes such as talking with the user and reporting to administrators or (for cross-wiki abuse) stewards? Nemo 20:48, 30 August 2017 (UTC)

Nemo, I think you're overlooking the fact that the "technical spaces" covered by the Code of Conduct does not perfectly coincide with "our wikis". It is entirely possible for a person to to be involved in the "technical spaces" without ever touching a wiki. How well do you think our "standard wiki processes" work for dealing with problems that occur exclusively in off wiki? Whatamidoing (WMF) (talk) 21:48, 30 August 2017 (UTC)

The complaints mentioned in my previous comments were not about MediaWiki usernames and therefore "routine work on our wikis" could not be applied. I also want to note that such routine work in relation to abusive usernames seems to be limited to deleting such usernames and preventing the creation of new ones. It's oversighters work, maintenance work. The perspective of harassment is (as far as I can see) missing. By adding impersonation as unacceptable behavior in this CoC we are improving the support and protection to those being harassed through this method, also through MediaWiki usernames. Qgil-WMF (talk) 07:25, 31 August 2017 (UTC)

How is this oversighters' work? The required action is a quick block and revert by one of the global admins or stewards or one of the ~200 local admins. If there's insufficient policy clarity on impersonation, get consensus for importing the Commons username policy here. And in other technical spaces, how does impersonation even work? IIRC Gerrit and Wikitech don't have open account registration, and IRC and the mailing lists have their own ways of dealing with this issue, and Etherpad is so removed from identities in general so as to make the concept of impersonation meaningless.

We want to create an open, welcoming community. Making the CoCC the one-stop location for problems in general does not accomplish this, and expanding the domain of the completely-unaccountable secretive uber-powered unchallengeable-without-ban Committee comes with real consequences. --Yair rand (talk) 02:33, 4 September 2017 (UTC)

Having a "one-stop location" for conduct issues is a feature, not a bug (I agree non-conduct problems should be reported elsewhere, but impersonation is clearly a conduct issue). It makes it easy to report, and that's a good thing. There are also some other incorrect or unproven assumptions here. Wikitech does have open account registration. I also specifically disagree with your assumption that IRC and mailing lists deal effectively with this issue. Mattflaschen-WMF (talk) 04:17, 13 October 2017 (UTC)

@Nemo bis: The comparison to vandalism is invalid for multiple reasons.

• There are existing project pages here that deal with vandalism, including but not limited to Project:Deletion. There is no policy dealing with impersonation.
• Some things can be reported in more than one place. E.g. unacceptable behavior at events can be reported to an event organizer and/or to the Committee. The same could go for impersonation (local admins or Committee, or both, as necessary).
• Dealing with impersonation is not "routine work on our wikis" for two reasons. It's not that routine (thankfully) and (as Whatamidoing (WMF) already noted) it's not just on the wikis. Mattflaschen-WMF (talk) 04:17, 13 October 2017 (UTC)

[unindent] So there seems to be two sub-cases here:

1. Someone unintentionally creates a username very similar to another username. The user does not intentionally use this username to mislead people, it appears to be purely accidental. This does happen from time to time. I don't think Code of conduct comittee needs to be involved in such cases - its something that can probably be worked out between the people involved, or mediated by an uninvolved party.
2. Someone intentionally creates a username similar to a different username for the express purpose of misleading people. Usually this kind of behaviour is obvious because they don't just have similar usernames, they also try and talk/behave like the other user. I think this sort of thing can be handled the same way people who spam the wiki (or spam phabricator, or spam gerrit) are handled - block and move on. At least, naively it doesn't seem like a complex user dispute requring "judgement" by the committee (Of course its always possible that username impersonation is happening in the context of a larger dispute in which case the CoC should consider it, etc).

Thus I tend to agree with Nemo here. Furthermore, I worry about delegating things I would consider "everyday" vandalism to the committee due to the committee's commitment to confidentiality. While I understand the arguments for confidentiality in cases of serious harrasment, I am unconvinced of the appropriateness of confidentiality in such day-to-day incidents. Bawolff (talk) 18:29, 31 August 2017 (UTC)

I am slightly concerned at the phrase all I can see is a single line - that is quite sufficient. In fact it is more than enough since it is obvious that impersonation is in general a bad act.

Rich Farmbrough 23:21, 31 August 2017 (UTC).

@Bawolff: Note that the Committee only reacts to reports, and will resolve on reports deemed to be related with harassment or disrespect. Adding the use case of impersonation in the CoC does not mean that now the Committee will be encouraged to fiddle in the regular processes that already exist to deal with abusive usernames. It doesn't mean either that cases of unintentional similarity of usernames will be handled by the Committee. It means that people who feel disrespected or harassed because of an alleged case of impersonation will be able to report the problem to someone who will not only look at the fact (a username created that might or might not be deleted) but also the root of the problem (someone harassing someone else, and whether there is an intention behind). Qgil-WMF (talk) 10:31, 1 September 2017 (UTC)

Maybe I'd understand the opposition better with an example:

Imagine that someone is intentionally impersonating you, and that this impersonation is actually causing a problem for you. Under that circumstance, which of these options sounds more like your position?

1. You should have the right to voluntarily report that problem under the Code of Conduct, if you choose to.
2. You should be prohibited from voluntarily reporting that problem under the Code of Conduct, even if you wanted to. The committee is prohibited from acting upon complaints about impersonation. We might all agree that impersonation is bad, and that it is a serious security risk when it takes the form of w:Pretexting, but the Code of Conduct doesn't cover that, and every victim will need to figure out some other process(es) for addressing it (probably a different process for each separate forum or tool).
3. Maybe you should be able to report it under the Code of Conduct, but it is desirable (why?) for there to be some uncertainty around whether this behavior is covered by the Code of Conduct.

Based upon this conversation, that last item appears to be what we've got, and I don't like it.

Bawolff, I am certain that if I found someone impersonating you (or anyone on your team), then reporting it under the Code of Conduct would be the last thing on my mind. My first action would be a face-to-face video call with your teammates and any staff member who could convince Darian that I really was me, and not part of the hacking. But this question isn't about what I would do. It's not even about what you would actually do. It's about what you should be permitted to do, if you personally and voluntarily chose to do it. In the end, my question is: Do you want victims to have an uncontested ability to voluntarily report this kind of problem, at their own discretion, under the Code of Conduct, or not? WhatamIdoing (talk) 16:31, 2 September 2017 (UTC)

Also: This conversation seems to have gotten fixated on the idea of a similar username. So let me give you an example that has nothing to do with usernames, but which an acquaintance in university administration reports as a surprisingly common reason for academics to file harassment complaints against students: Imagine that you are participating in a technical event. Imagine that you learn there that some person completely unknown to you is claiming to be your spouse or romantic partner. That's "impersonation". This is not a problem that can be solved by blocking accounts. I assume that staff, if attending in their paid capacity, could report it to HR to get some assistance. Should our volunteers also have an easily identifiable source of official assistance? WhatamIdoing (talk) 17:44, 4 September 2017 (UTC)

The CoC already forbids harassment and sustained disruption, and also unspecified other types of clearly inappropriate behavior, so presumably the concern is that someone might impersonate someone else in such a way that it is not harassment nor disruptive not clearly inappropriate, but nevertheless harmful to the community. That seems like an empty set to me. Some examples would help.

Even if something is already covered by the CoC, it might be worth calling it out specifically (after all "inappropriate behavior" already covers everything) if we think that it's not something everyone would recognize as inappropriate so we want to clearly warn away would-be perpetrators, or if it's not something generally covered by codes of conduct so we want to assure readers that they are going to be protected against that kind of behavior if they join the Wikimedia community. Impersonation doesn't seem like either case to me.

As for claiming to be somebody's partner (relative, etc) that's not actually impersonation in itself, as you do not pretend to be someone else, you just lie about your relationship to them. In any case it's pretty clearly inappropriate and covered by the current CoC. --Tgr 06:59, 5 September 2017 (UTC)

Yes, pretending to be someone is a case of impersonation. You don't have to use the exact words "I'm Anne, the Princess Royal" to impersonate her; you could equally say "I'm the Queen's daughter". WhatamIdoing (talk) 22:31, 5 September 2017 (UTC)

@Bawolff: It's expected that some Code of Conduct violations can be dealt with informally (rather than through the Committee), if it makes sense. The policy even spells that out in the reporting section. It's still valuable to note that it's unacceptable, to provide a reference point. But a true one-time accident wouldn't be a violation. However, I would argue that the user in your second example is particularly harmful, and a good use case for the policy. What they've done is worse than just putting some juvenile vandalism, or spamming their company (everyday vandalism). They're actively attempting to impersonate someone to misuse or harm their reputation, and I think that's exactly the kind of unacceptable interpersonal behavior this policy should cover. Mattflaschen-WMF (talk) 04:17, 13 October 2017 (UTC)

• Comment: If we already say "impersonation is bad". I do not see anything wrong with adding that fact to the CoC. Doc James (talk) 16:12, 2 September 2017 (UTC)

To be clear, is the proposal about impersonating a specific person, or just generally pretending to be someone else? For example, is it impersonation to post under a random real-sounding name (random as in not belonging to anyone specific) to a mailing list? I would suppose it's not, but in that case I really have trouble interpreting "inappropriate patterns of behavior that cause people to confuse one’s identity with others". Isn't that exactly what "impersonation" means? --Tgr 06:59, 5 September 2017 (UTC)

It's about impersonating a specific person, e.g., pretending to be your wife, your boss, or you. It could reasonably be extended to impersonating a role, e.g., pretending to be a WMF employee when you're not (think criminal impersonation, not entertainment). It does not include accidental confusion, such as two people who happen to have the same name, or two users who happen to have similar usernames. WhatamIdoing (talk) 22:31, 5 September 2017 (UTC)

Both, depending on the details. Clearly it covers impersonating a specific person. Choosing a "random real-sounding name" in good-faith is fine. Maybe someone wants to volunteer under a psuedonym, so they choose John Smith. That's fine. If it turns out there is already a John Smith (or something confusingly similar like Jon Smith) in the community, that's probably a good-faith mistake. It's a one-time thing, so not an "inappropriate pattern[] of behavior", and not impersonation (unless it was intentional). However, if they then pick Robert Cooper as a psuedonym, and it "accidentally" turns out Bob Cooper is also a long-standing user, it starts to become a pattern (especially if this continues). Mattflaschen-WMF (talk) 04:17, 13 October 2017 (UTC)

I'll respond to some topics from above here. Almost everyone agrees that it is a real problem, and some have pointed to policies that cover all or part of the problem on other wikis (e.g. username spoofing). Yair rand also stated that "We want to create an open, welcoming community.", which I agree with, and I hope everyone does.

Yet, there is no policy here about it. (At least, no one has pointed to one, and AFAICT there is none.). It is claimed that putting it in the CoC would make it "non-standard". But it's just the opposite. No one can point to any existing MediaWiki.org standard. We are trying to establish a clear, written, standard here, based on known best practices. Yair rand says we need to "get consensus" to import a policy. Well, that's exactly what we're doing here (except there is no reason our policy needs to be a copy of Commons, we can write one that makes sense for here).

I don't understand the argument that it's a real-world problem, but shouldn't be in here (or should use the catch-all clause). It's common enough that there are existing policies on other wikis. It's not some obscure thing and it's not minor. We just need to cover it here clearly. The clearer we are about real-world problems, the more useful this list is as a formal and informal reference point.

I specifically want to emphasize that this policy is needed for all the spaces, not just the wikis. Username abuse is a valid example, but as noted, it's not the only valid example. Mattflaschen-WMF (talk) 04:17, 13 October 2017 (UTC)

I agree that in non-wiki spaces such a policy is needed. Part of my concern here is we're conflating many different needs into a single point. I think the different types of impersonation differ in severity very signficantly, and its a mistake to group them all together under a single point. I personally believe its important to be specific when it comes to forbidden behaviour, and also to treat things differently depending on context. If we go too broad, we run the risk of the policy turning into a vauge blanket statement of "Don't be an asshole". Bawolff (talk) 17:56, 13 October 2017 (UTC)

Bawolff, do you have a feeling that "I can report something that I perceive as a problem to the CoC committee" almost automatically means "Somebody will get punished"? Sure, the level of urgency and severity varies for this problem. But don't we think that the CoC committee could say, "Thanks for the report. Yes, your real name is practically the same as this other person's real name. We can see that might cause some confusion. But, you know, this isn't the end of the world. Have you considered just putting a note on your talk page about it, to say 'I'm Brian. You might be looking for Brion'?"

I think I'm not understanding why anyone would want to prohibit people from being allowed to talk to the committee about their (perceived) problems. Whatamidoing (WMF) (talk) 17:07, 19 October 2017 (UTC)

No I imagine that the committee would give complaints all due consideration and act appropriately given whatever the complaint is (The committee is new however and much of its proceedings are secret. We have yet to see how the committee will actually act, and if its judgements are wise. Time will tell). That said, I don't believe one size fits all for all things, and small issues (of which some but not all of the things mentioned in this section are) should be handled by local means. Bawolff (talk) 18:57, 20 October 2017 (UTC)

I see the effect of including this kind of complaint as clarifying and simplifying the process for (alleged) victims: Whatever form of harassment you're experiencing, the committee should hear about it and will try to help in some fashion (including helping indirectly, e.g., by recommending that you ask for help in a different forum).

I see the effect of disallowing this kind of complaint as very odd in some circumstances (e.g., impersonation at an in-person event) and adding complexity and burden to the (alleged) victim, because it turns "send e-mail message complaining that someone is creating problems by claiming to be me/my boss/my wife" into a long process of figuring out exactly where and how to file the complaint, which is a really significant burden for newcomers. I've been impersonated in the past, and I'm not the only Wikipedia editor who has had that experience. It's okay to expect someone like me to know that accounts such as User:WhatamIdoingg can be reported at WP:UAA or WP:SPI, that other problems go to ANI, that cross-wiki abuse can be taken to the global sysops or stewards, and that other problem is in a Yahoo! forum, so there's nothing to be done about it, etc. But you and I are the top 1% here. We know the back pages, and we can tap a friendly admin if we need help. It's IMO not okay to expect every single normal person or newcomer to be able to do what we can do.

I see the effect of silence about this kind of complaint as introducing undesirable uncertainty: first, the committee may or may not feel empowered to handle it; second, the complainants might see a discussion like this and conclude that the Code of Conduct only covers the kinds of harassment that don't involve impersonation; and third, if such a complaint is filed, and reported out, that there will be a couple of people here claiming that the committee was overstepping its authority to receive/process/act upon any such complaint.

I'm not actually seeing any advantages to leaving impersonation out of the list. Can you name any advantages to treating impersonation as a separate special-case thing, that might or might not be covered by the normal Code of Conduct? WhatamIdoing (talk) 19:22, 24 October 2017 (UTC)

This proposal seems to be stuck, with neither side having been able to bring up stronger arguments than "feels unnecessary" or "might be useful in some non-specific hypothetical scenario". So maybe just hold a straw poll and archive it or refer it to the committee based on the results? --Tgr (talk) 00:37, 26 November 2017 (UTC)

Discrimination in section against discrimination

"particularly against marginalized and otherwise underrepresented groups"? This means that discrimination against the majority is bad, but not as bad as discrimination against the minority. 1984, anyone? This part of the Code of Conduct needs to be reworded. KATMAKROFAN (talk) 02:02, 25 September 2017 (UTC)

Absolutely true. Orwell's Animal Farm also applies here: "All animals are equal, but some are more equal than others." Yaron Koren (talk) 16:10, 25 September 2017 (UTC)

Definitely, as there is death penalty for being straight in parts of the world and men are being catcalled on daily basis in streets of NYC Ladsgroup (talk) 09:54, 7 October 2017 (UTC)

That comment was uncalled for and contributes to a hostile environment. There are legitimate arguments you could have made in favor of the delineation, and instead you have made the strongest argument against: Openly mocking the circumstances of non-underrepresented groups, coming straight from a member of the Committee.

When coming here to contribute to Mediawiki, please leave your local culture's grudges and conflicts at the door. We are not here to RIGHTGREATWRONGS. --Yair rand (not logged in) 08:31, 8 October 2017 (UTC)

The emphasis is intentional, to encourage considering under-represented groups. However, the discrimination clause doesn't allow harming any group, so no change is needed. I would replace Yair rand's statement with "We are not here to ignore context and pretend we're an island". Ignoring the real world is not how you build a welcoming and successful community.

Ladsgroup didn't mock any person or group. He used sarcasm to criticize a fallacious argument. Mattflaschen-WMF (talk) 07:13, 13 October 2017 (UTC)

I don't think you're contradicting anyone else's statement here - you're saying that the WMF needs to discriminate against certain people, in order to counter discrimination in the real world against others. Which is either justified and common sense or disturbing and Orwellian, depending on how you look at it (I tend to view it as the latter). Yaron Koren (talk) 13:27, 13 October 2017 (UTC)

I feel like some folks are reading this part of the Code of Conduct and are worried that "Oh! They're going to discriminate against ME!". That shows a lack of understanding of the inherent power/privilege between the groups and individuals involved in any case that could potentially involve this part of the CoC - or in society in general. That's OK to not understand, people make mistakes and we're all learning. This concept is sometimes referred to as "reverse discrimination". It is, by all accounts, a myth.

Discrimination can be leveraged at 'non-marganlized' folks. It's just called discrimination - and covered by any healthy social norms - like what you'd find in a Code of Conduct. It's when you consider the systemic relationship of power between the group of underrepresented folks and those that are not, that the importance of this statement become apparent. We can't assume there's a fair and equal playing field. We don't exist in a vacuum in which all groups have equal power. This line of the CoC acknowledges that and makes it clear that we don't discriminate against marginalized and otherwise underrepresented groups here. Ckoerner (talk) 22:17, 13 October 2017 (UTC)

You may want to revise your language. You seem to be trying to make a point that "reverse" discrimination is a useless term because it's just another kind of discrimination. But in saying it's a myth, you may be confusing others into thinking you're advocating the (IMO blatantly false) position that it's impossible to discriminate against members of the group generally considered to hold the societal power. Anomie (talk) 14:17, 16 October 2017 (UTC)

I think in order to be fair the policy needs to be clearer and more explicit. For instance, Jews are not discriminated against much anymore and have prominent positions in technology, should they be considered '"non-marganlized" folks'? This article suggests that straight black men are the new white people, so are they considered fair game for discrimination? If the degree of protection from discrimination is going to based on a contributors racial background, we should work out which racial groups get what kind of protections and come up with some kind of procedure to verify whatever racial identity they're claiming. In order to enforce this provision consistently, we need to establish a clear racial hierarchy and determine to what degree different genders, sexual orientations, religions, etc make discrimination (un)acceptable. Datadash (talk) 19:58, 11 November 2017 (UTC)

No, I disagree with your argumentation. --AKlapper (WMF) (talk) 12:51, 12 November 2017 (UTC)

OK. Good to know! Would you like to explain why you disagree? Datadash (talk) 02:21, 17 November 2017 (UTC)

@Mattflaschen-WMF: Which fallacy specificly do you believe that Yair rand committed? From my reading of this conversation I do not believe he committed a fallacy so much as you and he have a difference of opinion, probably rooted in different cultural/moral values. If we're speaking of fallacies though, ladsgroup's statement is arguably committing the naturalistic fallacy (Just because there does exist groups that are marginalized, does not neccessarily imply that we ought to be especially careful about marginalized groups in the code of conduct. Facts about the world generally cannot imply by themselves what we should do) [Of course I'm being a little unfair here in that I don't think his statement was intended to be a full argument on why we should have the discrimination statement in the code of conduct] [Edit: I'm striking this part of my statement as I think it might be unfair. In the context of the conversation its not clear enough what argument ladsgroup was specificly making, whether he made his statement to imply specificly that the descrimination clause should be in the CoC or if he was just asserting a factual statement about the world that not everyone is equal]. Bawolff (talk) 01:09, 14 October 2017 (UTC)

---

I feel like people are talking past each other here. Moral beliefs are kind of like religious beliefs in that they vary a lot between different folks and its basically impossible to say one belief is right, and the other is wrong. At best you can evaluate an argument in the context of foundational moral values, informed by factual statements about the world (which can be true or false). To that end, I think both sides have legitament arguments here (hopefully I do not misrepresent anyone's opinion here. I've tried to start each argument with what I believe is the base assumption that people are arguing from):

For the against side:

• Based on the foundational belief of isonomy/Equality before the law: All people should be judged equally. Justice should be object as possible and not "relative". Any sort of special treatment of any group based on who they are (as opposed to the context of their crime) is morally repugnant. While it is certainly true and very unfortunate that there exists certain groups who are marginalized, advising people to especially consider marginalized groups and thus polluting the equality of decision making is a greater evil then advising people to consider things without any reference to a person's background.
• Based on the belief that justice should be as objective as possible, and that such a clause is inherently subjective: While the code of conduct needs to be flexible as to the context of various offenses, it should still nonetheless be as objective as possible; When evaluating a CoC claim in a given context, different people should ideally come to the same conclusion regardless of their personal views. The definition of a marginalized group, usually actually means a "popular" marginalized group. After all, the code of conduct is inherently an attempt to marginalize people who are "assholes"; nobody would suggest the code of conduct should try to prevent marginalization of mean, disruptive people. Nor would probably anyone suggest we need to especially consider how terrorists, pedophiles, etc are marginalized in Wikimedia projects. But at the same time, there probably exists groups which are unfairly marginalized but we just don't realize they are marginalized, or we are mistaken in our belief that such a group deserves to be marginalized - and whether or not a group deserves to be a legitimately marginalized group probably varries depending on who you ask. Given that there is no real definition of marginalized in CoC other than a group treated unfairly that doesn't deserve to be treated unfairly (With whether or not they deserve to be treated unfairly left up to the eye of the beholder) this is inherently unfair to groups that are treated unfairly but that the general public has not yet realized are being treated unfairly. The argument would go, that such inherent subjectivity has no place in a policy designed to govern how we behave.
• Argument from symbolism - Symbols have power over people. They are statements about the type of people we are and the type of people we want to be. Having symbolic associations that people disagree with can make them uncomfortable to belong to our community - even if the symbol is theoretical and not a practicial concern. This argument would go: we're not really worried that such a clause would actually cause people to discriminate against us (where I assume the person making the argument is not in a marginalized group) but we don't like the symbolism of suggesting that group membership should have any bearing on the code of conduct. Its easy to dissmiss this argument as irrelavent since the symbolic meaning doesn't hurt anyone, but I think symbols are a lot more influential than most people give them credit for. Of course the opposite of this argument could also be made - that the marginalized group clause is a symbol and without it people will feel less comfortable.

Arguments for:

• Argument from dispensing the best justice for the most people (I suppose this is semi-utilitarism?): The fairest thing to do is the action that increases the sum of the overall justice in our community: We have to acknowladge that we live in an imperfect world where groups are marginalized. We may wish this wasn't the case, but it is actually the case. Furthermore, this marginalization often limits affected people's access to justice. To this end, to maximize overall justice well treating all groups the same when it comes to infractions, we also need to specially consider groups that are more likely to be discrimated against and may be less able to seek recompense for the discrimination.

In many ways, the for and against break down is really a comparision of - do we optimize for consequences in the context of current society or do we optimize for global objectivity outside of current societal context. I believe there are no right answers, only trade-offs, and that which trade-off you like best is probably going to vary by person. Hopefully i represented these arguments faithfully and fairly. I tried in any case, but I'm only human and I'm not sure I agree with all of them so its possible I may have missed the mark. That's all the arguments on this point I recall hearing, both here and in other places, but its probable there are other arguments also.Bawolff (talk) 01:58, 14 October 2017 (UTC)

Thanks Bawolff. I do agree I think folks are talking past one another. I do appreciate the attempt to bridge the understanding. I am trying to understand the folks who take umbrage with this section, even if I don’t personally agree. I’m assuming good faith that folks are genuinely concerned. I hope my reply is perceived as equally genuine. Ckoerner (talk) 14:00, 16 October 2017 (UTC)

Personally, I'm confused by how you dismiss reverse discrimination as "a myth" in the process of defending a literal case of reverse discrimination. Yaron Koren (talk) 14:09, 16 October 2017 (UTC)

One other dispute I've seen related to this statement is more practical than ideological. The CoC explicitly exempts targeted outreach (e.g. outreach where people not meeting certain criteria are denied participation) from falling under the general ban on discrimination when it's outreach to marginalized and otherwise underrepresented groups. That implies that targeted outreach to groups that aren't marginalized or otherwise underrepresented is indeed banned as discrimination, otherwise the CoC would just say some variation on "targeted outreach is allowed and encouraged" or "targeted outreach is allowed, and targeted outreach to such groups is encouraged".

There's one popular outreach program that no one wants to ban but that limits participation to a group that does not seem to be "marginalized" or "underrepresented". The counter-arguments, as far as I can make them out, include that it's only discrimination if the organizer "intends" to discriminate; that it's only discrimination if it promotes "harassment and disrespect"; that no one actually wants to ban that program so it's "obviously" not against the CoC; that the program has "positive outcomes" so it can't be considered discriminatory; that anyone with sufficient time, money, and/or level of academic achievement could theoretically become a member of the group so limiting to that group isn't discrimination; that "to such groups" is supposed to be ignored when interpreting the statement; and that any "sane" committee wouldn't ban it regardless of the actual text of the CoC so the wording was non-problematic. Anomie (talk) 15:15, 16 October 2017 (UTC)

For the old discussion see here and here. I'm not sure that's really relevant here; whether the text should say "Targeted outreach to marginalized groups is allowed and encouraged" or just "Targeted outreach is allowed and encouraged" should probably be its own discussion. --Tgr (talk)

Err, this is the section discussing that statement. Anomie (talk) 14:35, 27 November 2017 (UTC)

I mean, while no one proposed a specific amendment here, presumably it would involve removing "particularly against marginalized and otherwise underrepresented groups"; and the much more direct fix for the outreach issue is the wording I mentioned above, which is not related to the topic of this section (discrimination). (Also, as someone interested in that discussion, I'd rather have it in a section that does not start with comparing everyone who disagrees to Stalin.) --Tgr (talk) 06:12, 28 November 2017 (UTC)

Thanks Bawolff for bringing this discussion back into constructive territory. I would kinda hope no one is suggesting we have anti-utilitarian moral policies :) People who think the current wording is bad should at least make an attempt to show it's bad on a utilitarian ground. Some quick points:

• Rules about marginalized groups in particular do not seem to be part of most codes of conduct (e.g. CC, Citizen CoC, OpenStack, Elastic, Django, Ubuntu, Go). Counterexamples: Rust, Open CoC. (Just a result of some random googling - a more systematic review would be nice.)
• Any sane legal system punishes crimes which intend to further marginalize a group harsher than "normal crimes"; that's the intent behind hate crimes being their own category (typically with significantly more jail type than "plain" libel/assault/whatever). This does not mean that different groups are not equal before the law per se (hate crimes are defined by the type of group signifier they are based on, not the specific group) but in practice most hate crimes are committed against marginalized minorities. So such groups do require special consideration. Maybe the intent with calling out marginalized groups in particular in the CoC was something similar, although there might be better ways to put that.
• One way to interpret the current text is that discrimination against marginalized groups will be treated more harshly. Another way is that we should pay extra attention for such discrimination as we are less likely to notice it due to unintentional bias (that's pretty much what it means to be marginalized). I'd hope the latter is fairly uncontroversial; maybe we could replace the current text with something saying that more unambiguously?

--Tgr (talk) 01:31, 26 November 2017 (UTC)

Relation between Phabricator Etiquette and Code of Conduct

Since early 2014 we have Bug management/Phabricator etiquette. Since 2017 we also have a Code of Conduct for technical spaces. This has led me to some questions in phab:T167786 how these two documents influence each other. Your thoughts are welcome in phab:T167786 (so the discussion is centralized in a single place). Thanks in advance! --17:40, 18 October 2017 (UTC)

Based on comments in phab:T167786#3735498 I plan to add one sentence at the very end of the Etiquette for clarification: "In case the disregard can be seen as unacceptable behavior defined by the Code of Conduct, the Code of Conduct committee will be informed." Again, comments are welcome in phab:T167786. --AKlapper (WMF) (talk) 19:23, 4 November 2017 (UTC)

Resolved in phab:T167786#3753457 --AKlapper (WMF) (talk) 21:47, 12 November 2017 (UTC)

Flowify?

Any thoughts about switching this page to Flow, like most other mw.org pages? The discussions are hard to follow, manual archiving is a bit cumbersome, plus the page to propose amendments should probably be technologically inclusive, which wikitext really isn't. --Tgr (talk) 00:35, 26 November 2017 (UTC)

OTOH, some people find Flow harder to follow, since it currently lacks sane indenting and doesn't work well with a "show me the diff of everything added since my last visit" workflow. Anomie (talk) 14:33, 27 November 2017 (UTC)

@Tgr: That was my hope, but I didn't want to disrupt on-going conversations. (It's called Structured discussions now, BTW.) Jdforrester (WMF) (talk) 21:46, 27 November 2017 (UTC)

Structureddiscussionify just doesn't have the same ring to it :) I think SD is not so great for very active discussions, for the reason Anomie mentioned (and converting a very active discussion is confusing anyway), but since the creation of the committee, not much discussion is actually happening, and that doesn't seem very likely to change. There are two active threads right now (well, three if we count this one), and since SD diffs works fine for a single thread, it doesn't seem like too much of an effort to keep up. --Tgr (talk) 06:19, 28 November 2017 (UTC)