Talk:Requests for comment/Global arbitration and dispute resolution

Please, continue to discuss at Talk:Requests for comment/Dispute resolution committee.

Interesting, but unnecessary

I disagree with the proposal to make a committee that would serve as Arbcom by default on projects. Each project has a different community and a different pattern of evolution. One body cannot be adequately expected to take the needs and considerations of hundreds of projects into account when handling disputes. We would also be interfering in local autonomy by imposing this from on high. Geoff Plourde 06:37, 19 July 2010 (UTC)

Please read the proposal first. It is not a dictatory body which takes over decisions on a wiki without request. If a small community without local arbcom has a dispute, they can come and can get advise of experienced users and arbitration. They don't have to come, if they don't want. It is not unnecessary, but badly needed, as the past years showed. People from smaller projects just don't know where they can get help otherwise. --თოგო ^(D) 09:56, 19 July 2010 (UTC)

Ah, I must have misread the proposal. It seems from the wording that this would be something that projects would be dealing with by default. So would this body be advisory in such cases where a project seeks advice? Or would it be a binding opinion? Geoff Plourde 18:52, 19 July 2010 (UTC)

Examples needed

To get a better understanding of the problem this is meant to solve, could someone knowledgeable please provide some examples for "there are more and more issues all around Wikimedia projects in which they have to make urgent decisions, which is according to the Steward policy, but the number of needed decisions is above the comfortable limits"? sebmol ^? 06:49, 19 July 2010 (UTC)

Well, I can provide some examples of problems that could have been adequately addressed by Global Arbitration Committee based on what I remember:

• Siberian Wikipedia (creation of attack Wikipedia in an invented language)
• Russian Wikibooks (local bureaucrat who had a sense of ownership of the project)
• Controversy with English Wikiversity
• ace.wikipedia (recent)
• Dozens of "abusive local sysop/Arbcom" complains Meta receive through RFC and other venues. vvv^t 07:41, 19 July 2010 (UTC)

Take a look at Requests for comment. We've just passed the first half of the year and we have almost two times more cases than in 2009 (OK, with a couple of real RfCs, not complains). There are open issues since 2008 (and probably from earlier time; I didn't check those issues which are not categorized). In the most of open cases we don't have ability to solve them, as it would require our (stewards') non-urgent decisions. The last urgent desysoping on ace.wp is not a problem for us, but it is one more issue which makes our position uncomfortable. --Millosh 07:47, 19 July 2010 (UTC)

BTW, I think that large communities which decided not to have ArbComs, like de.wp is, should have right not to pass through DRC, but to go eventually directly to GAC if needed. Also, DRC is much more urgent now than GAC. --Millosh 07:47, 19 July 2010 (UTC)

Well, since DRC's motions will not be binding, I think it can be easily established without community-wide global poll or whatever. I believe we can create it as soon as we agree on what it should be. vvv^t 07:50, 19 July 2010 (UTC)

DRC's motions should be binding inside of the defined scope. --Millosh 08:19, 19 July 2010 (UTC)

BTW, we already have non-binding places, like RfC is. --Millosh 08:20, 19 July 2010 (UTC)

dewp does have an Arbcom, btw. ;) It's only not elected in the same manner as on other wikis, as there are no oppose votes allowed. --თოგო ^(D) 09:47, 19 July 2010 (UTC)

How many stewards?

How many of the stewards were at this meeting? All? Most? Some? Geoff Plourde 06:54, 19 July 2010 (UTC)

I think we where 6 that were present, and 4 more that was on Wikimania but couldn't attend the meeting. It was later discussed among all stewards on the mailinglist, and (as far as I can tell), we all agree that there exists problems that we (the stewards) currently have to solve and enforce (even if it's against the steward policy), but that we don't want to take these descisions (as stewards), since it should be a diffrence between those that make a descion, and those that execute/enforce it. (You could call it separation of powers). Laaknor 13:34, 19 July 2010 (UTC)

Ok, so there is a general consensus of stewards that this is a good idea? Geoff Plourde 18:53, 19 July 2010 (UTC)

There is a general consensus that we (as stewards) don't want this job, and that somebody needs to do it. How we do it is up to the community, but leaving it to a small group of stewards has been suggested by some. Laaknor 22:11, 19 July 2010 (UTC)

Ah, if someone told me that being steward meant that I'd have to take a position on a global arbcom, I wouldn't have run. Dispute resolution is fine, by itself. Technical positions (such as admin, bureaucrat, and steward) are fine by themselves. Combining them into a body that can make decisions and then implement them without oversight, however, is a consolidation of power that I'm not comfortable with: Someone needs to have the ability to say no to orders and enforcement actions. Kylu 23:59, 19 July 2010 (UTC)

Agreed. I see stewardship as valuable in a lasting way because it is a technical service position, and broadly neutral. Resolving disputes demands different skills, and is best kept separate. –SJ · talk | translate 04:31, 20 July 2010 (UTC)

some wikis...

Regarding to First level of dispute-resolution for the communities without arbitration committees. Some wikis don't have an ArbCom because they don't want to, they only use a noticeboard to admins, so global ArbCom will override this local decision? --by Màñü飆¹5 ^talk 08:09, 19 July 2010 (UTC)

As I mentioned above, DRC shouldn't be used for that purpose in the cases of [very] large communities which decided not to have ArbCom, like de.wp is. --Millosh 08:18, 19 July 2010 (UTC)

Thanks :-) --by Màñü飆¹5 ^talk 08:23, 19 July 2010 (UTC)

There is a problem here with the "no translation/lost in translation" issue--no practical way could this body cover all wikis adequately. As to wikis with arbcoms, it's not needed. I do think there's some use here for things like pedophiles, cross wiki issues such as those who disrupt across several wikis, but it'd have to be clearly defined. and that hasn't been done yet. — Rlevse • Talk • 11:50, 19 July 2010 (UTC)

Stewards already have the problem with translation (how do we know when we should oversight something that none of us can read?), but we do manage, and I think that the global arbcom can manage. Some times we only use Google Translate, other times we need to discuss the case on Instant Messaging/IRC to fully understand the case, or get someone uninvolved to translate for us, but I do not think this is a large problem. Personally, I'm opposed to having an global arbcom that works as a supreme court, where every decision made by a project ArbCom can be tried once more, and would oppose that particular part of this RFC, but there is a problem with some wikis where the ArbCom isn't working, so there might be a need for a body that can disband ArbComs, but I see no reason why that comittee needs to handle "ordinary" cases. Laaknor 13:18, 19 July 2010 (UTC)

I agree that GAC is not urgently needed yet. However, ArbComs are judicial bodies and if you want to have fair treatment, the second instance is needed. And, of course, GAC would deny the most of cases (let's say, some obvious case of sockpuppetry), but it is always possible that some ArbCom wouldn't make the right decision because of many reasons, including benign mistakes. However, I agree that it is not yet needed urgently and that that DRC is much more needed. Which, again, doesn't mean that we shouldn't create GAC before its existence is urgently needed. --Millosh 13:54, 19 July 2010 (UTC)

It isn't clear to me what the GAC would be needed for. A group to resolve cross-project disputes makes sense - that seems to be what you mean by a DRC, and there simply isn't something like that today. Can you describe a case that would require a GAC? Are you saying that you see a GAC as an "appeals court" for local AC decisions? This sounds like another layer of red tape, worth avoiding if possible. –SJ · talk | translate 04:31, 20 July 2010 (UTC)

I am not thinking about GAC as a kind of supreme court, but as a kind of body which deals with rights of Wikimedians. Thus, I think that it is natural to expand Ombudsmen commission in that way. They would deal not just with abuses of permissions by checkusers, oversighters and stewards, but with public abusing rights of Wikimedians, including local ArbComs decisions and DRC. This shouldn't be an appeal court, but court for human rights. Thus, the most of local ArbComs decisions shouldn't be considered by GAC. Anyone who would be willing to make a case before GAC should give proves that his or her human/Wikimedian rights are abused. --Millosh 06:03, 20 July 2010 (UTC)

DRC should deal with all kinds of disputes, not GAC. --Millosh 06:03, 20 July 2010 (UTC)

I think that this has to be removed. Not only big wikis, but also small wikis may have chosen not to have an ArbCom at all. Just think how much opposition there was to global sysops, a global ArbCom for ordinary dispute resolution will never be accepted. --Nemo 22:16, 19 July 2010 (UTC)

Decision of every community not to have ArbCom should be considered. In the case of large communities, GAC would just continue to do the job of Ombudsmen commission. For smaller communities decision should be made on case by case basis. Some communities are not very large but very transparent (for example, en.wn), some other have problems and avoiding local or global instances could be just a reflection of inner power struggles. DRC should serve in such cases. --Millosh 06:03, 20 July 2010 (UTC)

Good idea

I think this is a good idea. Many projects do not have a large enough community for an ArbCom to be formed. However, once in a while decision need to be made. That's where global ArbComs will be useful. Kayau (talk) 12:56, 19 July 2010 (UTC)

Such as Wikiversity, which is filled with conflict right now at Wikiversity:Community Review and its transcluded subpages. It's hard to expect that the community will be able to review the current situation there when pretty much the entire community is involved in some way. There are no unbiased participants. Adrignola 13:43, 19 July 2010 (UTC)

I'm pretty uninvolved :p, but we could always revert to dueling to resolve our issues. On a more serious note, external intervention would probably only lead to a bigger mess, judging from previous attempts Geoff Plourde 18:59, 19 July 2010 (UTC)

Brickbats at three quarters of a mile, I assume? Kylu 19:02, 19 July 2010 (UTC)

Expand Ombudsman Committee?

Could some of the problems mentioned here be addressed by expanding the Ombudsman Committee or setting up a Ombuds program to provide advice to projects? Geoff Plourde 18:57, 19 July 2010 (UTC)

That's an interesting idea. Ombudsmen review and field specific concerns, I don't know if that extends to resolving disputes between a number of people. You might ask the current omubdsmen if that is something they feel they do already. I could certainly see that program expanding to cover various types of concerns and conflicts, to generally help keep the peace, ensure people are heard who might otherwise feel silenced, and suggest good resolutions. We should look at the actual cases from the past year to see if that would suffice to handle cross-project problems that have arisen (and not been resolved terribly quickly). –SJ · talk | translate 04:31, 20 July 2010 (UTC)

What I'm thinking is that the current committee gets renamed to something more snazzy, and then having an actual Ombuds program to work with the projects (i.e. people are selected based on clue-level and language skills, then trained and work with specific projects). This would allow projects to maintain autonomy, while providing a new resource for them to utilize. Geoff Plourde 04:50, 20 July 2010 (UTC)

On reflection, I think this is a very good idea. I also think we need a narrowly-framed meta arbcom, that simply deals with the non-trivial global block/lock requests. we've had 2 in the past 6 months; let's say there are 4 a year - it would be better if a small focused body reviews evidence and makes recommendations on those requests. (at present an arbitrary one or two stewards will tackle and try to make sense of these requests and decide whether or not to act... it's not very efficient, and a bit out of scope for the steward role.) –SJ · talk | translate 05:20, 11 September 2010 (UTC)

Separate the two proposals

There are two very different proposals here.

• A dispute resolution committee that can resolve or advise on cross-project disputes or issues.

  Yes, we have a vague process for Meta RFCs, but that often fails to produce coherent advice or any specific outcome. And noone feels ownership of the process; whereas a committee tasked with providing advice on such matters would handle them clearly.

• A global arbcomm, which is more controversial and less clearly defined.

I recommend separating these two proposals, and clearly listing a few sample issues from the past year that each one might deal with. –SJ · talk | translate 04:31, 20 July 2010 (UTC)

If I count well, the only case with which GAC would deal would be Violation of local policy by Ukrainian arbitration committee and just in case that person who complains shows that their human and Wikimedian rights were abused. Otherwise, GAC would continue to do just Ombudsmen commission job. --Millosh 07:35, 20 July 2010 (UTC)

All other issues would be dealt by DRC. --Millosh 07:35, 20 July 2010 (UTC)

The very nearest situation with rules, policies and violations by administrators, checkusers (on my mind) is in ru.wikipedia.org.--Prima klasy4na 22:09, 26 July 2010 (UTC)

(Which was mentioned in wm2010:Submissions/From Russia with love and squalor: an overview of Russian Wikipedia.) --Nemo 10:33, 27 July 2010 (UTC)

As for the two proposals, I think we should proceed with the DRC proposal only, and do the GAC later. I think the much more urgent thing is DRC, and it might not frighten so many people like a GAC proposal would maybe. So, for tactical reasons, it might be better to leave the GAC stuff out completely for now. ;) --თოგო ^(D) 10:53, 27 July 2010 (UTC)

Revise with narrower scope

Update: I've created a new RFC at Requests for comment/Global ArbCom with this narrower scope. Please comment there. –SJ · talk | translate

I've been trying to clear up in my own mind how global locking is used today; what is needed to make global blocking possible for named accounts, and how it would be used; and what other cross-project decisions might need to be made about behavior that can't be described purely as 'counter-vandalism and -harassment'. A few examples:

1. Recent global lock requests / actions affecting Tyciol and Thekohser
2. The recent discussion about a global lock for user:השואה, who is making useful contributions on at least one wiki
3. Cross-project disputes about how to react to off-wiki insults and harassment on ru:wp and ru:wv (where as I understand via Milos, it is possible to be desysoped for an off-wiki insult until you apologize)

I don't know how to handle the third type of issue mentioned above, but the first two issues -- when to implement global (often long-term or indefinite) account blocks (using whatever tools are available) -- is something that normally large projects have ArbComms for.

I would like to see a body separate from the stewards that handles that very narrow class of requests. This body would only handle requests to globally block IPs or accounts for issues more complex than vandalism and basic abuse. These requests would have to come to Steward requests/Global, following a standard process (offering significantly more information than the vandalism requests do). And requests would need to list the local responses to the problematic actions on any wikis where the account had been active. I am guessing that they would (for now) handle some 4-5 requests a year. A single clerk might suffice to help organize information and flesh out related policies.

Thoughts? –SJ · talk | translate 04:42, 11 September 2010 (UTC)