Jump to content

Topic on Talk:Code of Conduct/Archive 5

Proposed amendment: Ban on soliciting tech support for hate groups

49
Ladsgroup (talkcontribs)

In accordance with CoC's amendment policy. The committee proposes an amendment on behalf of one of the community members to be added in unacceptable behavior section:

Soliciting help, support or technical assistance for websites whose predominant activity or content is behaviour that explicitly and significantly violates provisions of the code of conduct is forbidden in Wikimedia technical spaces. For example, websites advocating violence or hate speech are not welcome to ask for support on Wikimedia support forums.

This is similar to w:en:Wikipedia:No Nazis

Pbsouthwood (talkcontribs)

Why is this necessary?

Bawolff (talkcontribs)

There was a discussion about whether or not it was within the rules for the operators of fascipedia.org to ask for assistance on Project:Support_desk. The general conclusion was there is currently nothing in the rules forbidding that. It then lead to a discussion on whether or not that should be the case.


Relatedly, not quite the same thing, but a little bit similar, phab:T323956 happened a little while ago as well about whether WikiMANNia's extensions should be indexed.

This post was hidden by Ladsgroup (history)
Pppery (talkcontribs)

This comment is really not helping your case.

WikiForMen (talkcontribs)

Your comment does not help anyone, but supports new emerging fascist movements.

Ladsgroup (talkcontribs)

Hi, member of CoCC here, comparing other editors with Nazis is an explicitly defined type of personal attack and won't be tolerated. As such, we deemed the above comment inappropriate and in violation of CoC and removed them. This also serves as a warning to @WikiForMen to avoid making such remarks in the future.

This post was hidden by Ladsgroup (history)
Ckoerner (talkcontribs)

>Hey, you, are you literated?

I think you mean illiterate, which is not an inappropriate accusation to make to another community member. We don't do that here.

WikiForMen (talkcontribs)

No, I asked: "Are you literated?"

I remember, Ladsgroup wrote this

"This is similar to w:en:Wikipedia:No Nazis"

putting our editors on a par with N*. Later he deleted replies.

This is the fascist way. Fascists can be brown, red, green and purple coloured.--WikiForMen (talk) 11:42, 17 November 2023 (UTC)

Ladsgroup (talkcontribs)

Notifying that I have blocked WikiForMen from editing this page given the constant violation of COC

Yaron Koren (talkcontribs)

Whatever WikiForMen's transgressions, he's right that linking to the "No Nazis" essay at the top was a mistake. That essay is itself a massive code of conduct violation - it explicitly says that any editor with "somewhat-less-than-complimentary views on other races and ethnicities" is a Nazi. (Which I dare say would put the majority of the human race in the Nazi category.) Linking to it - and then blocking anyone who throws the "Nazi" term in the other direction - does seem to strengthen the case that any enforcement of this proposed rule would be arbitrary and political.

Bawolff (talkcontribs)

For the avoidance of doubt, I certainly did not mean to imply with my comment that WikiForMen or their editors are nazis. Only that the discussions had some similarities in that the impetus for them was that some contributors view the content of that website objectionable. Thus the discussions have some similarities, but of course are not identical.

I do think fascipedia.org are nazis, which i don't think is a controversial take as they themselves identify as such afaict.

Pbsouthwood (talkcontribs)

I can see that people don't want to have anything to do with such websites, but is there any contractual obligation to deal with such a request? If not, just ignore them, or is it so that the buck can be passed?

JCrespo (WMF) (talkcontribs)

This is my personal opinion, but one of the goals of any CoC is making sure people feel safe to participate in our community- that all voices can be heard and are not "lost". Having clear rules about what kind of behaviors are not welcome I believe are a key aspect of it- and saying "just ignore them" (not a criticism of your comment, I mean having that as a policy) I believe can make people feel left aside or ignored/not heard, while having a policy with concrete actionable for something brought up as an issue that actually happened can help getting the right people in and the wrong people out. So I applaud the committee for proposing the change.

Pbsouthwood (talkcontribs)

I don't have any objection to the principle, but I think you should leave out the example. Examples tend to distract from the general philosophy of the argument and attract straw man arguments by focusing on the specific.

Noloader (talkcontribs)

Small nit: Don't personify a website, as in "... for websites whose predominant activity..." People are "who's." Use a word like "that" or "which" for a non-individual, like a corporation or website or other object. In this case, maybe "with" would do: "... for websites with predominant activity or material content ..."

Noloader (talk) 19:30, 17 April 2023 (UTC)

Bawolff (talkcontribs)

Support

24.185.185.231 (talkcontribs)

Support easy agree

TheresNoTime (talkcontribs)

Support

Quiddity (talkcontribs)

Support

Skizzerz (talkcontribs)

Support

BDavis (WMF) (talkcontribs)

I very much personally agree with w:WP:NOFASH, but the CoC already forbids personal attacks, derogatory comments, and discrimination. Are there particular circumstances where the committee envisions that the current protections against disruption of our community by people exposing fascist ideology would not be sufficient?

Samwilson (talkcontribs)

Support

@BDavis (WMF): I'm not sure, but it sounds to me like this is about excluding people who run sites that exhibit those attitudes, but who aren't actually saying anything here on mediawiki.org that goes against the CoC (but rather asking for help with wiki configuration etc.).

Bawolff (talkcontribs)

This is more meant to address if someone posted something along the lines of, "please help, my website kill-all-the-minorities.com is showing an exception error, how do i fix it?"

Legoktm (talkcontribs)

Why is this limited to tech support? Shouldn't it include any activity within Wikimedia technical spaces? I have no intention of reviewing or merging patches written by people who operate hateful wikis, but that's also not "tech support". As written this would not be applicable to the situation at T323956, which IMO is case in which better CoC guidance would've been useful.

On the text itself, I would avoid the term "hate speech" because it does have legal meaning in some jurisdictions. The newer Wikipedia:Hate is disruptive essay (which IMO is better argued than the No Nazis one) uses "expressing hateful views", which is functionally equivalent.

Ladsgroup (talkcontribs)

@Legoktm Any better wording would be appreciated. It's hard to word it properly. It can be also two parts, one being that "we as a community won't provide any help including but not limited to tech support, code review, etc. to running hate groups" and keep this part verbatim as part of unacceptable behavior. It is important not to ban "helping" because the person might not know they are helping a hate group but to ban "asking for help".

Bawolff (talkcontribs)

To split hairs, i guess there's a bunch of different things here all related to the same idea - to wit, conduct that is primarily off-wiki but at least is a little bit on-wiki and having a chilling affect on wiki. i.e.

  • Doing on wiki activity that furthers hateful content (e.g. "Help fix my KKK wiki or whatever")
  • Documenting technical things that themselves are ideologically neutral, and potentially usable generically, but are primarily used by hateful content (e.g. documenting extensions primarily used by/made for such types of sites, even if the documentation does not reference anything problematic).
  • To paraphrase w:Wikipedia:Hate_is_disruptive#Hateful_conduct - "Referring to oneself on-wiki as being a member of a hate movement, or outing oneself as supporting such a cause off-wiki."


I generally think there is a real chilling affect if such conduct is allowed. Its difficult to work productively in an environment that includes people who essentially are like, I won't go into it on wiki, but by the way I think you should be murdered.

But i can also understand fears that such prohibitions, if taken too far can turn into morality police or witch hunts. Quis custodiet ipsos custodes and all that. To some extent, that risk is just the price of having rules.

My personal view is that the further off-wiki the content is, the more extreme it should have to be to get them blocked. For example, I don't think anyone should be sanctioned for making a joke in poor taste on twitter, if it is not referenced on wiki or connected to the wiki in any way, even if the same joke made on wiki might get you sanctioned. But someone actively advocating for genocide, should be unwelcome in our community because fundamentally if they're not unwelcome someone else is being made unwelcome.

Abbe98 (talkcontribs)

Support


"Soliciting" could maybe be replaced with something more accessible like "Requesting", "Asking for", etc? The later part could probably address persons, not "websites"?

Ckoerner (talkcontribs)

Support

GregRundlett (talkcontribs)

Support

Tgr (talkcontribs)

The Code of Conduct is a list of behaviors that are not wanted in Wikimedia- and MediaWiki-related technical spaces. It is fairly extensive in disallowing things, which makes sense; these spaces are for technical collaboration, so most of the things that one could do online are off-topic and irrelevant in these spaces. One can limit which cultural conventions are appropriate, which worldviews are appropriate to represent, what kind of interactions are considered productive etc, without much restraint - it's not hard to follow those rules during one's technical interactions, and they aren't particularly limiting. It's like workplace etiquette or rules for being able to use a nice restaurant.

Once you start applying the CoC to behavior that's not in any way meant to target or affect our spaces, to determine who are the good people who should be allowed to report MediaWiki bugs and who are the bad people who shouldn't, that extensiveness of the CoC becomes very problematic. Are you allowed to ask help for Boobopedia ("Gratuitous or off-topic use of sexual language or imagery")? For a citizen journalism wiki ("Unwanted photography or recording" of public figures)? Would you ban Wikileaks (yes, yes, it's not actually a wiki these days) for "Inappropriate or unwanted publication of private communication"? Who will decide what counts as "Offensive, derogatory, or discriminatory comments" when it's not about one Wikimedia technical space participant commenting on another, but some random wiki hosting political commentary or literature or forums or whatever?

Yaron Koren (talkcontribs)

Oppose Oppose

@Tgr above said it best about how arbitrary and capricious things can get when you can start policing things outside of Wikimedia technical spaces. I'll just add that even a restriction as simple as "no hate speech on your wiki" is ultimately impossible to enforce fairly, I would think. What is hate speech? Nowadays we are told that seemingly innocuous questions like "where are you from?", or terms like "grandfathered" and "master bedroom", are at least problematic, if not hateful - not to mention now-verboten statements of opinion like "men should not use women's bathrooms". It's an endless - and unnecessary - rabbit hole to go down.

Ckoerner (talkcontribs)

So you’re saying it’s contextual?

Pbsouthwood (talkcontribs)

I think they are saying people will game the system, and/or some people are unreasonable.

Bawolff (talkcontribs)

Would it help to make the language more strict to "advocating violence"? If i understand correctly, the concern is the slippery slope.

Fwiw, i disagree that this is about activity happening elsewhere. Once something is brought here it is happening here.

Tgr (talkcontribs)

It would definitely be an improvement. It doesn't entirely solve the problem. Consider a wiki organizing donations for Ukraine - it will very likely advocate for violence against some people, is that a good reason to ban it from technical support? Or a wiki about organizing not entirely peaceful protests in an authoritarian country?

Saying people should not advocate for violence while they are using Wikimedia spaces is a no-brainer; there is no legitimate reason anyone would need to do that as a part of discussing MediaWiki support and development. Saying people should not advocate for violence, ever, is enforcing a value choice. The world is a complex place, coming up with values which are sufficiently sophisticated that they can handle that complexity is hard. It's not something this CoC can reasonably undertake.

To put it in another way, this amendment would expand the CoC from a judgement on what values work for building collections of free software and free knowledge together in a specific project that we have decades of experience in, to a judgement on what values work anywhere for anything. That won't work; at the very least, you'd need a different set of values.

If you really feel the need to regulate, I think the amendment should say something like "admins can ban you from soliciting help if they judge your usage of Wikimedia software to be so unethical that having discussions with you in Wikimedia technical spaces would be disruptive to the normal functioning of the community" instead of pretending that that judgement somehow objectively follows from the text of the CoC, which will never be the case.

Anomie (talkcontribs)

I like Tgr's suggestion much better than the original proposal here. First, it's honest about what it's doing rather than hiding behind a facade of CoC respectability. Second, by giving the power to admins rather than the CoC committee it's more open to community oversight and, if necessary, challenge. The one thing I might change is to not make it a CoC amendment at all, just make it part of a regular "no disruption" policy.

It's still highly subjective, but at least its honest and open about its subjectivity.

Platonides (talkcontribs)

Oppose Oppose

Soliciting support is a technical matter, I don't think we should ban some people saying they are not allowed to ask for assistance, for the same reason freedom 0 allows using free software for any purpose. Even if we would prefer that our code is used for "Good, not Evil". Policing the Good websites vs the Bad websites brings similar issues.

Plus, it could be trivially workarounded by not stating which website they are having trouble with (which would be a good idea if mentioning the site might cause controversy, anyway).

Of course, it is perfectly acceptable for anyone not to provide any support to others based on whatever reasons e.g. Sorry, I prefer not to continue helping you with this, since I profoundly disagree with the view stated on your website that dogs are better than cats

But this is quite different than forbidding people from even asking. Anyone reading the CoC in good faith would need to (somehow) determine if they are welcome or they would be caught in that definition (which may not be so straightforward, a common issue with these kind of rules). Consider for instance a broken Wiki Farm where one in a hundred website might fall into such definition.

The CoC caters for actual behavior in the Wikimedia technical spaces. Shunning people for outside contents deemed[by whom?] to be "evil" would be problematic by itself, and might also cause other unexpected chilling effects. Would I be in violation of the CoC if I helped someone asking about what turns out to be (after it works again) a website advocating violence? What if, when trying to help the OP, I had asked another developer about the error they were facing?

I don't think we should be adding indirect prohibitions like this.

01tonythomas (talkcontribs)

Support

More like should make it wider, thinking in line of what @Legoktm said.

MarkAHershberger (talkcontribs)

Oppose Oppose

Someone who runs one of these problematic wikis offered a patch for an extension I wrote. I would like to accept their solicitation of help even though I do not want to endorse their beliefs. I could redact anything that ties the contribution to their odious wiki. But this policy seems like it would prevent me from accepting their patch.

For this and the reasons that @Tgr and @Platonides mention, I support this from @tgr:

I think the amendment should say something like "admins can ban you from soliciting help if they judge your usage of Wikimedia software to be so unethical that having discussions with you in Wikimedia technical spaces would be disruptive to the normal functioning of the community" instead of pretending that that judgement somehow objectively follows from the text of the CoC, which will never be the case.

Risker (talkcontribs)

Oppose Oppose

Is a pro- or an anti-Putin website a hate site? What about a pro- or anti-Zelensky one? What about sites that are pro- or anti- either of the two warring factions in Sudan? One that actively opposes the reign of King Charles and permits posting of proposals for civil disobedience? Depending on the definition of "hate speech", it could easily be all of the above.


Generally speaking, software is agnostic to the political positions of the people who operate it. And it has always been a conscious decision of the MediaWiki community to be a completely open source software that anyone can use or reuse; thus, it's even more agnostic than closed source or commercial software. At the same time, there has never been any obligation on the part of any member of the MediaWiki community to provide support to any and all non-Wikimedia usages of MediaWiki. This should not change - but neither should there be even the slightest suggestion that a MediaWiki community member will suffer consequences because they do provide assistance to another MediaWiki site without thoroughly researching their contents. As currently written, the proposal doesn't say that, but it certainly implies that it could happen.


I think Tgr's suggestion makes more sense, and does not require an alteration of the CoC. It's really important to provide hard definitions of any term that is used such as "admins" and "unethical" - I recommend avoiding the use of the phrase "hate speech" because it has specific legal meaning in several jurisdictions, and those legal definitions do not align with each other, let alone what is being proposed here.


I'll also point out that the various English Wikipedia policies and essays do not advocate blocking or removing editors who, external to the project, may make hateful statements. As long as they keep it off the project, there's no reason for anyone to even investigate whether an editor is a hate-monger outside of the project. There are a few exceptions to that rule of thumb but they aren't formalized or made into policy for good reasons. Those people are just removed from the project, normally with the intervention or assistance of T&S so they can be globally banned. I think some sort of parallel system should be considered here. I can certainly think of one MediaWiki usage outside of Wikipedia that totally freaked me out, and the operator of that site was an active member of the MediaWiki community until someone from the editing side red-flagged the account.

Bawolff (talkcontribs)

i disagree with the framing of "unethical" here. That implies we are primarily concerned with the behaviour of the wiki, which is an external matter. The matter at hand should be the affect of the behaviour on our community.


E.g. if someone asks for help on their wiki named www.LetsKillBawolff.com - it is probably an unethical site, but the real concern is that they are creating a hostile environment on mw.org through asking for help with their attack site. That's the connection to mw.org, and that is the thing i think would be in scope of our code of conduct.

Thiemo Kreuz (WMDE) (talkcontribs)

After a month of silence and being told the committee has no interest in an actual discussion I would like to repost my Wikitech-l response here. While I'm not opposing the idea that drives the proposed amendment, I have questions. TL;DR: I think my concerns can all be resolved with careful rewording.

Please bear with me and analyze this idea: Does the proposed amendment imply that I'm violating the CoC the moment I help someone who runs or supports such a website?

Or to ask the question differently: What are we banning here? The act of helping? The act of asking for help? The fact that a person belongs to a specific group? What a person believes? Am I supposed to check every time I help someone to make sure they don't work for such a website? How can I tell if a website matches the criteria?

The proposal is rather different from how the CoC and especially the enforcement process currently works. Usually the idea is to ban people because of their behavior. But the proposed amendment isn't about behavior: As hinted above it doesn't make sense to ban someone just because they ask for help. Still the proposed wording suggests that "soliciting help" would be the same as "advocating violence or hate speech". This doesn't seem right, especially when the person's intention clearly isn't to advertise their website.

I think the issue here is a conflict with how a CoC works: A violation is always a violation and must be handled as such, no matter if people know better or not. It must be like that. Otherwise anyone could simply claim "I didn't know" and get away. But this idea is not compatible with the proposed wording alone. It becomes actionable only with additional information: Either the person asking for help openly mentions their website, or they spam our support channels. Both actions should absolutely be covered by the CoC, sure. The question is: Aren't they already? Is there really a gap we need to cover with an amendment?

Unfortunately the publicly available information about the processes of the CoC committee and how it evaluates cases is extremely sparse. How does it decide, for example, what the "predominant activity" of a website is? What publicly available sources does the committee use to be sure they have a common, transparent definition of hard to define things like what qualifies as "hate speech"? Do they consult experts? Do they have a catalogue of words and phrases? Nobody knows.

The last significant edit to the committee's FAQ was in 2017.

Personally I would love to have a different enforcement process that – after an initial intervention, if one is needed – then focuses on giving people empathy, time and material to learn, to grow, to become better, and to say sorry. Unfortunately this is – from all I know and was able to observe – not at all how the committee operates. It's about "protecting the space". That's all. It's the reason why – for example – the first time a person might learn they violated the CoC can actually be a public announcement. It's not about the person, it's about the space.

Can we please rephrase the proposal and make it very clear that it's not about the act of helping but about the act of consuming our support resources? "Soliciting" is a word I find extremely hard to understand. According to my dictionary it could seriously mean both "asking for" as well as "offering help". Furthermore, I suggest avoiding the word "help". "Help" shouldn't be used as if it's something negative. For example, no doctor would deny saving someone's life, no matter how many hate tattoos the person has.

Bawolff (talkcontribs)

I have no objection to using a simpler word than solicit. However my understanding is that solicit (wikt:solicit) always means to ask for something. It would not mean to help someone in this context. It does not mean to respond but to ask or offer. I think the confusion maybe comes from that it can also be used to mean to offer something, particularly for pay (e.g. it could mean to try and convince someone to buy a service you are selling. So if someone went around to people who posted their mediawiki problems, offering to solve their mediawiki problems for $10, that would also fit the definition, albeit not in the way that is meant here). The person soliciting is the person making the request, not the person responding.

Re coc being overly punative and secreative with little room for regabilitation. I agree and have said similar things in the past. However i think that is a separate topic.

I guess the thing trying to be banned is the act of indirectly bringing such material here. Nobody can read what is in people's hearts and minds, and i don't think anyone is concerned with the consumption of support resources (they are there to be consumed after all)

Thiemo Kreuz (WMDE) (talkcontribs)

Thanks. Considering this I find the proposed wording even more alarming. If this is about promoting "websites that advocate violence or hate speech", why do we discuss "support and technical assistance" then? Why do we put pressure and risk on the people who give that support? Why don't we just ban "promoting or openly supporting websites whose predominant activity or content is behavior that explicitly and significantly violates provisions of the code of conduct"?

Jdforrester (WMF) (talkcontribs)

If this is about promoting "websites that advocate violence or hate speech", why do we discuss "support and technical assistance" then? Why do we put pressure and risk on the people who give that support?

That's not remotely what it says. The wording is:

Soliciting help, support or technical assistance for …

To re-word it to be perhaps clearer by making all the implicit words explicit, you could say:

Soliciting of help, of support, or of technical assistance for …

The prohibition is extremely clearly on the soliciting, and not on community members who might unknowingly provide such help, support, or assistance without being aware.

Thiemo Kreuz (WMDE) (talkcontribs)

The ban is clearly about "help, support or technical assistance". That's literally how the sentence starts. I'm afraid adding some "of" doesn't change anything. What even is "soliciting"? Does "actively offering help" count as "soliciting"? What when I am aware and still help?

Platonides (talkcontribs)

I think it means "Asking for help or asking for support or asking for technical assistance for websites...". So not as bad as you thought. Still, talking about the issue with other people while helping someone (that happens to be related to such a website) might be construed as Asking for help for that website.

And we would have a funny situation if a bug was found by such a banned website, since submitting a patch would require asking for a review (i.e. technical assistance) to get it merged...

Reply to "Proposed amendment: Ban on soliciting tech support for hate groups"