Talk:Global bans/Archives/2012

Criteria clarification

There are 3 criteria, each starting "A user is ..." or "A user has..."

Are these alternative criteria, ie a user meeting any of the 3 is eligible? Or must the user meet all 3 criteria, in which case they should start "The user" not "A user"? The text needs to be clear whether it's any of these or all of these. FT2 ^{(Talk | email)} 03:14, 3 January 2012 (UTC)

Good feedback. I'll try to clarify that you need to meet all of the criteria, not just one or two. Steven Walling (WMF) • talk 18:05, 3 January 2012 (UTC)

If all of these, then it should be "The user" not "A user" in 2 and 3. They're all discussing the same user, not possibly different ones. Compare:

"A job is in an office. The job pays $20,000. The job is stressful" and "A job is in an office. A job pays $20,000. A job is stressful".

FT2 ^{(Talk | email)} 18:13, 3 January 2012 (UTC)

Sounds fine to me. Be Bold my friend. :) Steven Walling (WMF) • talk 18:27, 3 January 2012 (UTC)

Been bold. Hope you like! If too bold, do revert :) FT2 ^{(Talk | email)} 18:55, 3 January 2012 (UTC)

Previously locked users

I brought up previous locked users to a few Stewards and they said that they are treating them in the same was as global bans. Will this be the default approach? Global bans normally require consensus before putting them forth while many locks seem to be based on the decision of only one or two people with no community discussion. Ottava Rima (talk) 18:14, 3 January 2012 (UTC)

No, obviously going forward with the new policy, a ban will be a ban and otherwise it is not. Steven Walling (WMF) • talk 00:00, 6 January 2012 (UTC)

Right, but the old locks will still be there. Which means that Stewards could potentially lock when there isn't enough to meet the criteria for a ban. I just want to hash out how we are going to treat the old behavioral locks that don't fit in the standard lock requirements (i.e. spam, bad user name, etc.). Here is an example. Ottava Rima (talk) 17:12, 7 January 2012 (UTC)

The easiest way to deal with those is to grandfather them IMO. It is hard enough just getting this proposal on its feet; needing to sort through old locks isn't an issue that we should need to focus on. Ajraddatz (Talk) 18:45, 30 January 2012 (UTC)

What is the problem this will solve?

I do not see that there is a global problem with banned users that this Meta policy will help resolve. On the other hand there is always the possibility that tinkering with the mechanism will actually create new, and totally unforeseen, problems. The rule of thumb advice common among farmers, If it ain't broke, don't fix it, may apply here also. Malcolm Schosha 15:32, 9 February 2012 (UTC)

As we do more outreach to get input/comments going forward, I agree there needs to be clear problem statement. To answer your question here in probably too brief way... The problem is that there is an edge case that a few problematic editors have exploited, which is that, even if there is a consensus to do so, there is no process or policy for banning a user globally. Even in the past cases where people managed hack together a consensus of some kind (1, 2) there is no global agreement on when such a decision is enforceable or not. The broken thing here is that serious trolls, some of whom multiple communities feels are causing offline harm to people, are allowed to slip through the cracks because of a lack of a global policy. Steven Walling (WMF) • talk 00:23, 10 February 2012 (UTC)

The problem this will solve is getting rid of disruptive trolls/stalkers who just won't go away. Kaldari (talk) 05:10, 25 March 2012 (UTC)

But won't they just turn to using sockpuppets? If they're determined enough to troll that they went to a different project to continue trolling, using sockpuppets after the global ban is a piece of cake for them. Deryck C. 10:09, 21 April 2012 (UTC)

Yes, but that happens locally anyway. There is no such thing as a foolproof block on an open wiki. In any case, the point of creating a policy is to have a decision making process that makes it possible for people from different projects to enforce a ban if that is the consensus. It doesn't solve the fact that we still have technical holes to fill. Steven Walling (WMF) • talk 19:11, 22 April 2012 (UTC)

But I think a good point was made for the past two cases of a global ban: if we don't ban them globally, the belligerent users' disruption is likely to be more manageable because that way they're less likely to turn to sockpuppets. Enacting global bans simply shifts the problem from a disruption problem to a sockpuppetry problem, and many on English Wikipedia will know well how this could become a moot and fiasco a few years down the line. Deryck C. 13:22, 23 April 2012 (UTC)

Yes, it's a possibility if there are regular global bans that pile up. I don't personally see the number of successful global bans being anywhere near the size that it could create a sockpuppetry problem that we are not capable of handling. The number of users who actually are dangerous enough that we must enact some kind of Wikimedia-wide ban is tiny. Steven Walling (WMF) • talk 18:22, 23 April 2012 (UTC)

If the number of expected bans is tiny, we don't need a set policy; the ad-hoc process we have so far would be adequate. Deryck C. 18:19, 24 April 2012 (UTC)

There are two reasons not having a policy isn't an option:

1. The lack of a policy has been an excuse for not taking action in the past. (And when I mean action, I mean even just figuring out whether there's a consensus to enact a global ban.)
2. The new terms of use that was just endorsed by the Board specifically mentions a community policy about global bans and links here. If the community doesn't make a policy about global bans, then we go back to the previous state of things, where lack of a clear policy lead to inaction and requests to the WMF to step in to enforce bans (which no one wants, including us).

I would normally agree that creating policies to serve edge cases is a bad idea. But not this time. Steven Walling (WMF) • talk 19:27, 24 April 2012 (UTC)

┌───────────────────┘
Steven, unfortunately point 2 above sounds exactly to me as a case of WMF deciding on a community policy, and forcing the community to accept it. As for point 1, I maintain the position that global bans will create an unfair situation in all circumstances except where there is an overwhelming consensus (lack of policy may have been an excuse, but an invalid excuse - this is a wiki). Deryck C. 20:30, 27 April 2012 (UTC)

Except that there were weeks and weeks of intense discussion about what went into the terms of use. The WMF did not just write it and decide that there must be a global bans policy. It was proposed and then vetted along with the rest of the contents before it went to the Board to review. Proposing an idea and forcing it on the community are two different things entirely. Steven Walling (WMF) • talk 22:53, 27 April 2012 (UTC)

I understand, but I think it was wishful thinking to assume we will have a policy when we don't have one yet, and opinion is somewhat divided about its enactment. Deryck C. 09:32, 28 April 2012 (UTC)

As far as I can see, a big part of the problem this is intended to solve is that some banned users from the English Wikipedia have moved over to small projects such as en-Wikiversity and started trouble from there. On the other hand, I can't think of a situation in the german-language projects where a global ban process would have been useful (obviously excluding ordinary vandalism accounts, but those hardly require any process). I would find it interesting to know whether users from other non-english communities feel the need for a global ban process. --Tinz (talk) 21:03, 27 April 2012 (UTC)

I think someone already posted about the proposal on the DE WP admin's noticeboard, though that was a few months back. Anyway, good point. Steven Walling (WMF) • talk 22:51, 27 April 2012 (UTC)

Cross-wiki outreach

Hey talk page watchers... I just wanted to drop a note here and say that since it looks like the majority of translation work is done so far (or at least as much as we're going to get in the short term), I was thinking the next step is delivering messages to the appropriate local equivalents of the Wikimedia Forum/Village Pump/Café etc. with a note about this policy. If anyone would like to help or can write/translate a message in our top ten languages, that would be fantastic. Thanks, Steven Walling (WMF) • talk 18:22, 23 March 2012 (UTC)

Thanks for your work on this. –SJ talk   15:26, 25 March 2012 (UTC)

not just editing

As we don't just restrict people from editing I've broadened the first sentence, I'll also broaden the ban bit to clarify that the appropriate ban should be related to the access that someone has misused - so if c/u abuse or misuse of deleted contributions resulted in a global ban it would normally be a ban from C/U or access to deleted content as appropriate - so someone can be very welcome as an editor even if globally banned from becoming an admin. WereSpielChequers (talk) 04:25, 24 March 2012 (UTC)

Thank you for adding this. Very good point indeed. Steven Walling (WMF) • talk 17:17, 26 March 2012 (UTC)

An alternative relating to overturning vs opt-in

I skimmed over the discussions above about what level of power local projects would have to overturn global bans or opt-in to participating in them, and was going to reply above but those threads look a little stale.

The mediawiki software currently has a user right called 'ipblock-exempt' which allows its holder to be exempted from IP address blocks. What about a user right called 'globalblock-exempt'? This would allow a local project to give a specific gban exemption for an editor that applied solely to their own project without needing to fight the global ban itself directly. As a user right it would need to be explicitly assigned, meaning the blocked user would presumably have to apply for it and the local project could set up their own process for assessing and accepting or declining requests. Because action is required on the local project, it would mean the local project also explicitly takes responsibility for the exemption and if the user disrupts that project, they effectively have themselves to blame.

An example could play out as follows: BadUser is active on the wikipedia and wikibooks projects for English and German, and has persistently disrupted enwiki, dewiki and commons, is well-known by admins on those projects, and a decision is made to enact a global ban. BadUser tries to continue his antics at frwiki but can't - frwiki (and every other project) is effectively protected from disruption by the global ban. The enwikibooks project, however, believes BadUser has never disrupted their project and can still contribute constructively there. They're aware of the global ban and the reasons behind it but choose to exempt BadUser from the global ban solely on the enwikibooks project by granting him the globalban-exempt right. They take full responsibility for their decision and if he screws up there, they can simply revoke that right.

This effectively fulfils the goal of the global bans in terms of protecting unwitting projects from highly disruptive editors, but still allows projects in-the-know to make the conscious decision to allow a globally banned user access only to their own project. Good behaviour by a globally banned user on a locally-exempted project could factor in assessing whether the global ban should remain in effect or be lifted.

This would eliminate the opt-in concept by effectively providing local projects with an opt-out instead, applied on a per-user basis. And it may also have the effect of taking some pressure off the overturning process in that it would allow one particularly vocal local project to solve the problem within their own project without exposing all the other projects to potential harm. Overturn discussions would still take place, of course, but would be more likely to involve participants from multiple projects since there would be no need to try to overturn the global ban on behalf of just one project.

Just a thought dump, in any case. NULL (talk) 05:45, 27 March 2012 (UTC)

Giving local projects the ability to exempt any user they like from a global ban before it even happens would totally gut the effectiveness of the policy/practice in the first place. The whole point of the policy is to force a discussion between the projects over whether a global ban is merited – building in the ability to opt-out so easily would not encourage projects to actually participate in the global discussion about a problem user, which is the opposite of what we want. Steven Walling (WMF) • talk 18:03, 27 March 2012 (UTC)

Forcing a process like this on every project is going to be controversial to begin with. The 'sovereignty' and independence of the individual projects has been pushed for a while, and the idea that each project has the right to make its own decisions for its own good is entrenched in places like enwiki. Each time the WMF has stepped in, it has created varying degrees of muttering and opposition. I don't think it's a reasonable expectation that a change like this would be accepted without considering some degree of compromise.

The purpose of the Global Ban isn't to bring the projects to the discussion table, that's a desirable side effect. The purpose of the Global Ban is to protect the projects from the persistent disruptive behaviour of particular users. Forcing a project to the negotiation table isn't something that will happen as idealistically as you imagine. More likely, they're going to do what they always do - mutter about WMF overextending its authority and ignore the process altogether. There's some underlying resentment that meta has some of the power over other projects and outside of their self determination as it is already, I just think you're going to butt heads with projects like enwiki if yet another part of their self-determination is offloaded to meta like this without giving them a local 'override' button. NULL (talk) 20:06, 27 March 2012 (UTC)

And just to clarify, my suggestion above wasn't that local projects exempt users from the global ban before it happens, it's something that would happen after the global ban is put in effect. I'm not familiar with MediaWiki's codebase but as a developer myself I can't imagine it would be difficult to only allow the flag to be added to accounts that already have a global ban in effect, and remove that flag on any local projects either at the same time the global ban is lifted, or the next time a 'can I log in' check is done locally. The notion is global ban→local exemption→global discussion, not local exemption→global ban→?→profit. NULL (talk) 20:23, 27 March 2012 (UTC)

I think it is a good idea. There are not many users that deserve a global block, but recently I have seen such a case where it would make sense to me that a project that is willed to keep it's contributer can make an exception. That someone behaves badly on one or more projects but not on another can have various reasons and it's often not the editor alone that makes two sides (opinions, etc.) collide. If a project really wants to keep such an editor because he is productive then it would be to the advantage of the project. We will also have to note that an editor blocked globally would have to ask for the permission on a project itself. That means that a global block would be in full effect until an exemption is made. --Niabot (talk) 21:55, 27 March 2012 (UTC)

There is a role for both global blocks and global bans. The software should definitely allow people to make global blocks, and local "globalbockexempt" decisions. Like the ban/block distinction on individual projects, a global ban is simply a global block that by social convention can't be overturned by a single project (conrast w: 'by a single admin' on an individual project). Let's hope we never have cause for a global ban, but let's also define it to keep the terms straight. People do propose a global ban every few years. (And if I understand WSC's comment above, where one could have 'activity bans' for a contributor, that apparently would also be a sort of ban, not suitable for individual overturn by a single project.) (someone who has abused access to private data might not be allowed such access again on the good faith of a single project). –SJ talk   03:49, 28 March 2012 (UTC)

I think distinguishing global blocks (which could potentially be overridden per wiki) from global bans (which are universal) makes sense. If a user has been disruptive enough to warrant a global ban, giving them a potential loophole to exploit will just prolong their disruption. Kaldari (talk) 05:34, 28 March 2012 (UTC)

I can get behind that too. Sounds like we need to take a stab at solidifying Global block too, and including a mention here about the more specific differences. Niabot, NULL... does that split make sense to you too? Steven Walling (WMF) • talk 07:16, 28 March 2012 (UTC)

If i understood correctly then you propose two kind of "blocks". The global ban without anything to add and the global block which can be lifted by single projects at own will/risk. But isn't it a bit complicated? Wouldn't be a global block alone enough to satisfy the needs if local projects can make single exceptions? What would be a good reason to disallow local projects to remove a block/ban? --Niabot (talk) 07:28, 28 March 2012 (UTC)

Are there two or three kinds of blocks? You talk about global "blocks" and global "bans" here, but a few weeks ago there was an office action which resulted in a global "lock" of a user account. So do we have three kinds of blocks ("block", "ban" and "lock")? --Stefan2 (talk) 15:24, 28 March 2012 (UTC)

The distinction between a block and a ban are not as stark on some projects (like Commons), but it should be made clear that they are very much different things. A ban is a consensus decision that someone is not welcome on the projects or is banned from certain activities. A block is a technical mechanism for preventing editing, and it can and is often applied without a ban. Slightly complicated, but important in cases like these, because not every block is a ban. Steven Walling (WMF) • talk 18:11, 28 March 2012 (UTC)

If a global block becomes possible, I'm not sure there would be a reason to continue locking accounts. That has always seemed like a hack to me. But it's probably worth pinging the stewards' noticeboard to see if any of them have a reason to continue locking accounts in that case. –SJ talk  18:17, 28 March 2012 (UTC)

I can see benefit in the block/ban distinction here, yes. As you stated, Steven, blocks are issued for any number of things, made by individual admins, revoked by individual admins and so on. Bans are decisions by the community that an editor is not welcome. A block is a way of enforcing a ban if it becomes necessary, but blocks are used more broadly than just to enforce bans. I'm imagining a system where an admin from any project could request a global block of an account provided it's blocked on at least 2-3 projects minimum (including their own), but blocks could be overturned either locally with globalblock-exempt or a petition to remove the global block itself. But as with enwiki, blocks enacted because of bans are almost never lifted until the ban itself is lifted. The enwiki admins are mostly pretty good at respecting this, so as long as a global block came with a note saying it was by a global ban decision, I think admins will respect that enough not to give local exemptions. And failing that, it can be codified in policy on the globalblock-exempt user right or something. I agree with others above that with a global block system in place (which is more diverse and perhaps more appropriate than the global lock system) then there should be no need for locking that I can see. NULL (talk) 20:59, 28 March 2012 (UTC)

• I have an impression that now we are talking about something different from what we were talking before. First, community bans do not exist on all projects, and there have been projects where community bans were explicitly rejected by a vote. Second, I technically do not understand, even if community ban existed, how it could have been implemented technically. By a vote on meta? This is unacceptable, since a typical vote on meta is a couple of dozen users, and on top of that, the votes are very vulnerable to the external influence such as en.wp, WR and God knows what, like we have recently seen on the example of Mbz1. Performing community ban votes on several projects and then ban a user if say three or more projects banned them? Has the same problems as the previous version (for instance, Commons votes are vulnerable as well), in addition, it is unclear which projects would be eligible: I can easily solicit users on a dormat project so that nobody would even know until the ban has been enforced. Honestly, I do not see any way this idea could be implemented. I think we should drop it.--Ymblanter (talk) 23:27, 29 March 2012 (UTC)
  • Oh god no. A vote is a terrible idea... The process here is described as a Request for Comment, which is explicitly not a vote. As for the rest: yes, all processes are sometimes vulnerable to gaming or influence. But on the other hand, the projects are surprisingly resistent to being forced to ban a user just because another project did. In any case, I would rather see a community decision on bans, rather than one where people try to force the Stewards or the Foundation to intervene without any kind of attempt to gain an honest cross-wiki consensus. Even if we're only talking about less than 10 global bans ever. Steven Walling (WMF) • talk 04:07, 30 March 2012 (UTC)
    • I think that in theory, it makes sense for a project to be able to override a global ban. If the WMF wants to force discussion or override the project for other reasons, it still has the power of office action - the only point of the exemption is to create a class of users who are banned by default rather than allowed by default, without choosing to override the projects in every single case. Nonetheless, admittedly, the creation of a new software feature to benefit less than 10 of the most problematic users ever to participate in the projects is, well, not the first priority for any sane developer. If a project wanted to invite one of these users back, they'd need merely invite him to start a new account under a slightly different name, with the agreement that he use it only on that project and that they will oppose any global ban request against it. (note that this proposed policy would prohibit this non-software workaround, but whether this discussion here should override an individual project's consensus to do otherwise seems debatable to me) Wnt (talk) 14:07, 11 April 2012 (UTC)
      • That is basically the situation we already have. For example: with Thekohser, a Steward decided to enact a global ban that ended up being watered down to just a series of individual local blocks across the wikis, and naturally they were all undone just on principle. I think a global ban needs to be all or nothing: if someone really is not enough of a danger to the projects that we can pick and choose which wikis they can be active on, then there's no need for a global ban in the first place and it should be left up to local autonomy entirely. Steven Walling (WMF) • talk 20:03, 11 April 2012 (UTC)
        • Whichever way we weigh up arguments on a global ban discussion, it doesn't work. Absolute consensus? If there's absolute consensus I'm sure we don't even need a new policy; the existing global block rules are adequate. Majority rule or "rough consensus"? That would intuitively lead to the scenario in which one or two projects who value the contribution of editor A is forced to accept a ban on A imposed by the consensus between 5 other projects on which A was disruptive. A global ban policy without local opt-out will simply do harm and no good. Deryck C. 10:18, 21 April 2012 (UTC)

• Question: Earlier in the discussion Niabot made reference to a "recent" case where there a user may have been allowed to edit on one project although globally blocked. I am not sure if we are thinking of the same case, but there was a recent case on Commons where a user had been blocked and unblocked, after much controversy and discussion, only for the account to be globally locked. This same case lead to renewed interest in Pedophilia and a similar policy proposal being created on Commons. I note that Meta:Pedophila is only a proposed policy, but if there were to be a WMF policy that covers all projects, how would this be handled by the proposal? Delicious carbuncle (talk) 10:29, 9 April 2012 (UTC)
  • Hey. Thanks for the question. The case on Commons was an office action, and the way this policy is written would mean that user would not be eligible for a global ban (they were not blocked on multiple independent projects). As for the proposal at Pedophilia, the text says that any such user "...will be indefinitely blocked." Which I take to mean is separate from a formal ban, and because it's a specific reason for indef blocking codified in a policy, would not take a consensus-building discussion like other cases might. If that policy is accepted, then I would definitely amend this one to mention it as an exception. Steven Walling (WMF) • talk 17:35, 10 April 2012 (UTC)
    • Steven, you say "if that policy is accepted". I am not very familiar with the workings of Meta, but the policy has been in a "proposed" state for nearly two years now, despite the clear statements from both Jimbo and Sue that this is de facto policy. What will it take to have the policy "accepted"? I fear that the next such incident will be a publicity disaster if this is not policy when that happens. If I were the WMF, I would not like to see some of the comments on Talk:Pedophilia reprinted in the media. Delicious carbuncle (talk) 19:18, 10 April 2012 (UTC)
      • I personally have no answer about how/when/if that page will become a community policy, though the "policies in limbo" problem is something that clearly impacts both pages. The right person to ask is probably Phillippe or legal in general. Steven Walling (WMF) • talk 20:00, 10 April 2012 (UTC)
        • I worded my question poorly. Let me try again. How does a proposed or draft policy on Meta become just "policy"? Delicious carbuncle (talk) 02:11, 11 April 2012 (UTC)
          • There is no set process. We're going to have to pick one. Steven Walling (WMF) • talk 19:58, 11 April 2012 (UTC)

Some comments

Hello all, I'll leave a few comments here. For those that do not know me: I was a steard and did not get re-elected. I do however write this as a normal user. The comment is also not about me, but simply tries to point out possible problems.

• Long term / indefinite bans suffer from the fact that the banned user has no way of "evolving", and that the community on whose decision this ban is based, will change. In such a case, there is a ban, but without community support. In addition, the actions for which the ban is enacted tend to be forgotten rather soon. For the long term users of a project: do you remember the banning discussions/reasons of people banned from your project, say 9-12 months ago? - Can you make sure the community would still support thiese bans, in a vote?
• Wikipedia comprises many projects; in order for the ban to have any legitimation, a majority of users and of projects must support the ban. Of course, I am only talking about short term bans.
• There is no need to introduce this instrument; stewards already have the option of globally locking an account, which will result in the same. As pointed out above, no community is able to meaningfully reach an agreement on banning a user from all projects.
• Just out of curiosity, how much work could have been avoided, say in the last 3 months, if the instrument were available?

Whe you as a community elect representatives, please give them the power to actually make decisions. --Eptalon (talk) 13:10, 30 March 2012 (UTC)

In regards to: "There is no need to introduce this instrument; stewards already have the option of globally locking an account, which will result in the same. As pointed out above, no community is able to meaningfully reach an agreement on banning a user from all projects."...

Both of these things are demonstrably untrue. First, as the back-and-forth among Stewards over the request for a global ban of User:Thekohser demonstrated, Stewards do not currently have a clear mandate for enforcing any request for a global ban. Yes, they have the technical ability for global locks, but no, there is not a set process or policy that covers this. Current Stewards have expressed deep reservations about any request to them for a global ban that isn't backed by a community consensus. Second, we've already proven the cross-wiki consensus model can work with a minimum of drama. Just see the two RFC examples I linked to the section at the top. Steven Walling (WMF) • talk 20:44, 30 March 2012 (UTC)

In order for a ban to be legitimized, it must be possible to "repeat" the discussion / vote that precedes this ban at any time during the duration of the ban; if at some point in time, the result is different, the ban lost its legitimacy, and must be lifted. As to stewards not agreeing on some issue: this is normal. They have been elected to serve a term - how they reach their goals is not the concern of the people who elected them. In addition, we are talking about a multi-lingual project. In the case of a global ban dicussion I would expect all commuication to be localized, so that any user of any wikipedia project can follow it, without the need to understand another language, other than the language of the projects where this user is active. In addition to creatng a problem of legitimacy of a ban, this requires a huge effort of translation, which in my opinion does not outweight the benefits.--Eptalon (talk) 20:27, 1 April 2012 (UTC)

I think Abd's comment on the Poetlister case summarized precisely my worries about this proposal. It doesn't work because it simply forces dedicated trolls into sockpuppetry; and it causes tension between projects by allowing larger projects to gang up and rule over smaller projects. Deryck C. 10:31, 21 April 2012 (UTC)

Some comments (2)

Currently with not so much time so I don't expect replies and if there are any I won't be able to get back to you soon for now. That said, I completly oppose this proposal as written for the following reasons: the first one is that it seems that there is an option allowing a project to lift a global ban locally. In my humble opinion that absolutelly wrong. Global means global, all projects. If we allow projects to overturn global bans locally then this policy looses its spirit and efficacy, and has the potential secondary harmful effects. The second one is that the proposal is unclear in regards of which ammount of consensus the request must obtain in order to the ban be implemented as it forces stewards, again, to decice. A formal threshold of support must be set, so all proposals can be treated with the same standards. These are my concerns as of now which I consider serious enough not to support this policy at this current stage. Thanks. —Marco Aurelio (Nihil Prius Fide) 21:39, 1 April 2012 (UTC)

There is not an option to lift a ban locally. The message is supposed to be that if a local community disagrees with a ban that they should take the disucssion back here where the ban originated. If that's unclear let's rewrite it. :) Thanks for your comments, Steven Walling (WMF) • talk 18:53, 2 April 2012 (UTC)

Principle of least restrictive ban

I didn't think this would be controversial so I boldly added it as principle 4, but as someone objected to it I thought I'd start this thread.

4 The Ban is the least restrictive needed to protect the project from similar incidents by the same editor. So, if the ban were due to the misuse of extra userrights, then the user would be banned from having those userrights on any Wikimedia wiki, but might still be welcomed as an editor.

This proposed principle is different from, but complementary to, the idea of global bans being a last resort. Some people simply can't be trusted with deleted edits or the block button, others are too disruptive even to be allowed to edit. Under this principle it would be entirely appropriate to ban someone from a position of trust whilst still allowing them to edit, providing that was the least restrictive ban needed to prevent similar incidents recurring. WereSpielChequers (talk) 10:21, 6 April 2012 (UTC)

There is already language in the lede which calls for the least restrictive measure ("...should only be sought when the problem cannot be addressed by less restrictive means..." and "...may be selective...") as well as the statement which follows the criteria, "Please remember that global bans are not a common occurrence, and that merely meeting the above minimum criteria does not mean that a global ban is required." I think it's unnecessary that we add such a criterion, as the language is quite vague and it's already covered. Steven Walling (WMF) • talk 23:42, 6 April 2012 (UTC)

I disagree. "May be selective" is a long way from the principle of least restriction. As for the "should only be sought when the problem cannot be addressed by less restrictive means" I read that as complementary to the proposal but as more about global bans being a last resort. I think this would benefit from bans being both a last resort and also minimally restrictive, and that would require the proposed 4th principle. WereSpielChequers (talk) 22:03, 8 April 2012 (UTC)

I don't think that the clause works as a prerequisite to considering a ban, which is what it would be ("Global bans are only considered if all of the following criteria are met:"). Making it a prerequisite seems to open up a world of wikilawyering (Example: "We can't discuss this because the contributor never misused talk pages and so the least restrictive ban needed would allow him to edit talk pages; since that wasn't proposed, we have to close this consideration.") Surely discussing the appropriate parameters of a ban would be part of the consideration? --Maggie Dennis (WMF) (talk) 11:03, 9 April 2012 (UTC)

WSC, I think a discussion about the principle of least restrictive ban is fundamentally moot here. Global bans are fundamentally about banning a user from using things they've never abused before because of fears that they will abuse them later. (Otherwise what would be issued is a series of local bans.) If you believe in the principle of least restrictive ban, you oppose this proposal outright. Deryck C. 10:34, 21 April 2012 (UTC)

Discussion close by Meta admins

I'd be much more comfortable with the policy if only stewards could close these discussions, as I trust stewards much more than I do meta admins. Stewards have a more public and through vetting process, and the most recent two incidents that I've come over to Meta to see have left me very, very unimpressed with Meta admins' judgements. Sven Manguard (talk) 03:23, 10 April 2012 (UTC)

Issue is that determining consensus on just about anything except global rights requests is outside of the steward scope. Stewards are supposed to implement consensus only, not define it. That being said, with something of this proposal's scope closure by a group of stewards might be best. Ajraddatz (Talk) 04:09, 10 April 2012 (UTC)

Yeah, we really have no set way to approve policy here on Meta. However, if we hold an RFC or something about it to garner wider opinion than the talk page, I would also request that Stewards close the discussion. The other good reason for this is that the policy is a global one which directly involves Steward work, and it has little to do with the usual kind of sysop work here on Meta. Steven Walling (WMF) • talk 17:26, 10 April 2012 (UTC)

Since you're the one that made the image, you might want to change File:Global-bans-process-workflow.svg then; that's the source of the concern that triggered this thread. Sven Manguard (talk) 18:40, 10 April 2012 (UTC)

Where is the Global arbcom ?

Where is the Global arbcom that can review and possibly overturn a Global ban ? On the French Wikipedia, any block decision by any admin can be reviewed by the Arbcom. Only the Arbcom has the final authority on disciplinary matters. The admins or stewards, or whoever on meta does not have the power to overturn the supreme authority of the French Arbcom on disciplinary matters on the French wikipedia. If the French Arbcom decides that the accused is innocent and is free to edit the French Wikipedia, no admin or steward or whoever on meta is allowed to overturn that. Teofilo (talk) 00:20, 22 April 2012 (UTC)

I am against bans as opposed to temporary sanctions

This idea of a ban is associated with death penalty, because it is considers that some human beings cannot improve themselves, even years after the event. I am against death penalty. I am opposed to any indefinite blocking on any Wikimedia project. Sanctions against a specific user should never be longer than 10 or 15 years. Blocking a user for more than 10 or 15 years is similar with cruelty. Banning someone forever is always an abuse.

I would not necessarily be so much opposed to a Global block for X years, provided X is less than 15. Teofilo (talk) 23:20, 21 April 2012 (UTC)

I agree; however, I believe that "X" should be less than five years. --Michaeldsuarez (talk) 18:01, 1 May 2012 (UTC)

Unfortunately (if you believe that) then your problem is not just with this proposal, but blocking/banning policy on most of the largest projects. Indefinite bans for egregiously bad faith editors is an extremely common practice on the local level, so arguing against it here is kind of fruitless considering the number of global bans has been and is likely to remain very low (e.g. less than 10-20 ever). Steven Walling (WMF) • talk 19:21, 1 May 2012 (UTC)

That system should be opt-in only

This system should not be imposed on projects that do not have a policy of banning people. For example, as written on en:French Wikipedia, the French Wikipedia had no sysop between May 2001 and August 2002. That means that during that period, nobody was banned, and nobody could be banned. Projects that do not have a policy of banning people should not be compelled to taking part into that system. For that reason, it should be opt-in only. Teofilo (talk) 23:25, 21 April 2012 (UTC)

Meta has no power to impose a decision on local community

No vote whatsoever on meta can have the power to overturn policies decided on local projects by their local communities. This is by the same token that a vote at the United Nations does not have the power to amend or abrogate a US law or a French laws or whatever. The United Nations is a place to regulate relations between nations, but it is not a place from where the nations are ruled. Meta is in the same position as regards to Wikimedia projects. A French citizen is allowed to be judged according to French laws. The united nations have no power to overturn a decision taken by a French judge. So by the same token, meta does not have the power to overturn a decision taken by the French wikipedia's arbcom. If the French Wikipedia's arbcom decides that the accused is innocent or should be blocked for 1 year, meta does not have the power to change that into an indefinite ban. Teofilo (talk) 23:33, 21 April 2012 (UTC)

Actually, Meta does have that power. It's exercised all the time on the Steward's noticeboard and via global locks of vandals and spammers. Steven Walling (WMF) • talk 19:07, 22 April 2012 (UTC)

The Falklands islands dispute example again

When discussing the Global blocking in April 2008, I took the example of the Falklands islands dispute, and the possibility to block a user on a different wiki for POV pushing, because that user edits that other wiki in a way that is seen as POV on the blocking party's wiki. I had been given assurances that the tool would not be used that way : Right, well it's possible to do so, but that's not what the tool is for; misuse will be handled in the same way as misuse of any other tool. Issues affecting one wiki stay on that wiki, and global blocking is not to be used for disputes of that nature anyway. I think you misunderstand how this tool will be used. – Mike.lifeguard | @en.wb 17:20, 25 April 2008 (UTC). See Talk:Global blocking/Archive 1#only for IP addresses, not usernames. See also Mark Graham's article in The Atlantic, 6 April 2012 in which he defends the view that Wikipedia should not unify its communities (in projects such as Wikidata) but provide the possibility for different world views to coexist.

But what is being prepared here includes the possibility of blocking a user on a third wiki (and on all wikis) if he is blocked on two other wikis. If two wikis having the same view on the Falklands Island dispute decide that a user who edits the third wiki in a way which they frown upon, is a POV pusher, then they are allowed to ban that user on the third wiki, by concluding an alliance of their two wikis against the third one. Teofilo (talk) 00:15, 22 April 2012 (UTC)

Please read the examples of global bans given in the policy. POV pushing or editorial disputes are not what this policy is for at all: it is to prevent harm to the projects and other users, not solve simple disputes. Steven Walling (WMF) • talk 19:08, 22 April 2012 (UTC)

Global bans violate Ne bis in idem

Global bans are exclusively applied where multiple independent communities have previously elected to ban a user for a pattern of abuse constitutes to a violation of Ne bis in idem. See en:Ne bis in idem. Teofilo (talk) 03:37, 22 April 2012 (UTC)

Even if they do, it hardly matters. Since when have we had a prohibition of double jeopardy on Wikimedia wikis? A user cannot be re-blocked for edit warring a second time?--Jasper Deng (talk) 05:21, 22 April 2012 (UTC)

No it doesn't. It's like moving a case to a higher court which can apply a stronger sanction for the same decision than the lower court can (projects can't ban a user on other projects). Rd232 (talk) 10:26, 22 April 2012 (UTC)

Ratification

I've proposed a ratification process at Global bans/ratification, where the global policy would go into effect when communities representing 75% of active users of Wikimedia projects endorse it. This process gives us clear steps to get the policy endorsed, namely a series of local discussions which approve it. Failing that, we seem to be stuck in a never-ending limbo. Since this talk page is already confused enough, I suggest the details of the ratification process (eg different thresholds or criteria) be discussed at Talk:Global bans/ratification. Rd232 (talk) 20:12, 24 April 2012 (UTC)

I think we should put more thought into mapping out our options before there's any movement on any one threshold for ratification. Personally I think individual/local ratification a bunch of separate times is way too inefficient. Steven Walling (WMF) • talk 21:04, 24 April 2012 (UTC)

It's not "inefficient" at all. Tapping into existing processes for approving a local policy, in a way which can proceed along parallel tracks in different projects is much more efficient than ... whatever the hell the current non-process for never getting anything decided is. And one or two "super-major" projects aside, any one major project getting stuck in its local process doesn't matter for endorsement of the global policy. Rd232 (talk) 07:36, 25 April 2012 (UTC)

Incidentally, the threshold and ratification criteria should be part of the policy draft, so it gets approved as part of the ratification process as well. Rd232 (talk) 07:36, 25 April 2012 (UTC)

I think this is way to inneficient and complicated too. This is not the General Assembly of UN or a Treaty being subject to the Vienna Convention; this is a proposal for a global policy that the community needs to agree or the Board needs to approve and make it official. The new Terms of Use recognizes Global bans as policy; so once those Terms of Use come into effect this should enter into effect too at the earliest. I think there's no need to complicate things further. They're already too complicated. Thanks. —Marco Aurelio (audiencia) 13:30, 25 April 2012 (UTC)

I think it's a perfectly sensible approach to ensuring a large part of the active Wikimedia usersphere actively endorses a new global policy - and really not that difficult to implement, or complicated to understand (legwork is required to get the necessary stats, but that's not complicated either). If the board is willing to make it policy by fiat, fine, but there's been no indication of that. All I've seen on the issue of how to make this policy is handwaving and helplessness. Rd232 (talk) 15:31, 25 April 2012 (UTC)

And what about simply using a vote process like Global sysops/Vote? --MF-W 16:31, 25 April 2012 (UTC)

(ec) Quoting: While the global ban is authorized by the terms of use, it will be implemented by community policy. I myself am not a fan of complicated procedures, but we need to run some procedure. I would be more in favor of a vote similar to the vote for the Board.--Ymblanter (talk) 16:32, 25 April 2012 (UTC)

Or, indeed, like for global sysops.--Ymblanter (talk) 16:32, 25 April 2012 (UTC)

┌─────────────────────────────────┘
So we have a few ideas so far. I'm going to update the ratification subpage with our options so we can see them all and decide. Steven Walling (WMF) • talk 17:34, 25 April 2012 (UTC)

The global sysops vote example seems reasonable to me. If a bunch of people dislike the result, they can come and try to change it through a similar process. Don't make change hard; make it easy. –SJ talk  03:07, 27 April 2012 (UTC)

I personally favor something more akin to an RFC, just because we don't really have the option to just say "no" now that the Board approved a ToU that mentions the bans policy. I'll post a draft of RFC questions soon, but I was thinking something like, "Yes" and "Not yet, the policy needs work". Steven Walling (WMF) • talk 16:53, 27 April 2012 (UTC)

This is a sensitive question, and I would recommend the system similar to how it was done recently with the Pending Changes RFC on English Wikipedia: to choose three uninvolved stewards who will in the end sum up the discussion. Their names will be known from the very beginning.--Ymblanter (talk) 20:12, 27 April 2012 (UTC)

Agreed. I'll make a request on the Steward's noticeboard today. Steven Walling (WMF) • talk 20:14, 27 April 2012 (UTC)

I have started a draft Request for Comment slightly like the global sysops vote here. Please feel free to edit. Steven Walling (WMF) • talk 22:26, 30 April 2012 (UTC)

Too many processes?

The way I understand it, we're now up to five global processes against a user: global blocks, global bans, global locks, office actions, and legal actions under the terms of service. I take it w:WP:BURO is not Meta policy... Wnt (talk) 00:09, 25 April 2012 (UTC)

AFAIK, locks are a "temporary" (maybe not) way of blocking accounts globally. Global blocks currently only apply to IPs and ranges. Is there a difference between an "office action" and a "legal action"? PiRSquared17 (talk) 00:15, 25 April 2012 (UTC)

Global blocks don't actually exist. There literally is no technical mechanism for doing so. Steven Walling (WMF) • talk 03:20, 25 April 2012 (UTC)

Overturning

I decided that folks are probably right: the text about overturning was too confusing. I've changed it now to say you simply need a broad community consensus to end a ban. Leaving it open ended can give us more wiggle room for cases when a particular project disagrees with a ban any time in the future. Steven Walling (WMF) • talk 00:12, 28 April 2012 (UTC)

I have changed it again because where overturn discussions take place was not stated, overturn discussions may be about a decision not to global ban a user as well, and hopefully avoid endless requests. --darklama 14:40, 4 May 2012 (UTC)

Hmmm, you also introduced the requirement of an RfC, which isn't really what the original points towards as a process... Deryck C. 23:07, 4 May 2012 (UTC)

I am okay having it or not. In the latter version which only required a "broad community consensus" and requiring notification to all the communities that helped decide a ban... I could easily see a situation where one or two wikis don't want a ban, and their local consensus is convincing enough to show that it merits overturning. On the other hand, explicitly saying you need a Meta RFC will prevent someone from jumping the gun and trying to get a ban overturned through the backdoor without notifying all the communities that contributed to the original decision. Steven Walling (WMF) • talk 23:16, 4 May 2012 (UTC)

I've changed the sentence after that because I think "informed and commented" is ambiguous when it comes to deciding who needs to be informed of the overturning discussion. Deryck C. 09:42, 5 May 2012 (UTC)

I had hoped to keep the text simple, but since it was ambiguous, I have expanded it to be explicit that similar requirements exist for overturning a decision as there are for a global ban. I think conceivably project "A" is informed of a global ban discussion because the user is recently active there, project "B" is informed of a global ban discussion because the user was banned there, members from project "C" participated in the decision even though no bans or activity from the user, a decision to globally ban a user requires project "D" be informed because the user becomes active there under another account, a member of project "E" makes a overturn request because project "E" is informed of a global ban decision when a user becomes active there under another account, and somehow project "F" is dragged into it as well maybe because that community also saw activity from the user under another account after being globally banned. I think a reasonable expectation is that projects "A", "B", "C", "D", "E" and "F" are all informed of the overturn discussion. --darklama 14:13, 5 May 2012 (UTC)

I think that's too much of a formal process - I reckon overturning should be a simpler procedure than blocking, because the "default" situation on wikis is the absence of blocks. Deryck C. 23:34, 5 May 2012 (UTC)

What do you suggest should be done to make the procedure simpler? An overturn discussion may be about overturning a decision to ban or not to ban. If the procedure is too simple, an endless cycle of discussions may be initiated to ban, overturn the ban, and reinstate a ban which may do more harm than the harm claimed to have been done by the user the discussions are about by beating a dead horse. --darklama 04:06, 6 May 2012 (UTC)

┌─────────────────────────────────┘
What we're describing is essentially the same as the process for requesting a ban in the first place, so let's not duplicate content. The shorter the better. Steven Walling (WMF) • talk 17:09, 7 May 2012 (UTC)

Archive and freeze for translation?

Hey folks: I was thinking in prep for the RFC, we should archive the old threads and put a temporary freeze on English edits so the translators can catch up. That sound ok? Steven Walling (WMF) • talk 17:57, 9 May 2012 (UTC)

I would support archiving old discussions in order to have fresh discussions take their place, and I would support freezing the page in order to allow for the translators to catch up. --Michaeldsuarez (talk) 03:10, 13 May 2012 (UTC)

Done -- I fully protected the page for 30 days (will take it down if translations catch up sooner) and archived the threads older than this one. Steven Walling (WMF) • talk 17:24, 18 May 2012 (UTC)

Hi Steven, why is WMF doing project's policy here? Or did you forgot to change your accounts? ;-) --Saibo (Δ) 14:37, 1 July 2012 (UTC)

Respected users who are also banned

What about the case of users who have been banned on two or more projects, but are considered to be respected members of another project. Particularly, what if they hold a rollback, admin, etc. flag or would be eligible for such a flag but haven't pursued it? --Philosopher ^{Let us reason together.} 19:28, 17 May 2012 (UTC)

I think the answer lies in this part of the policy: "Merely meeting the above minimum criteria does not mean that a global ban is required." Someone can be banned in 100 projects but still be allowed to contribute to another. The point of the two project minimum is to prevent requests for global bans that are completely frivolous, not say that someone must be banned globally for any particular reason. Steven Walling (WMF) • talk 21:41, 17 May 2012 (UTC)

Wouldn't it be a very bad idea to ban a user globally if the user contributes constructively to some project? I know of one user who is banned on one big project while at the same time being a bureaucrat on another big project, and I think that it would be a very bad idea to completely disallow such users to contribute to Wikimedia projects. --Stefan2 (talk) 22:27, 17 May 2012 (UTC)

Yes, that is the point of the above statement in the policy. Just because a user is banned on two or three projects does not mean they must be globally banned. If someone brings up a proposal anyway, every community where the person is active has to be notified and given the opportunity to object if they want. Steven Walling (WMF) • talk 17:12, 18 May 2012 (UTC)

What happens if there is a clear consensus to ban someone when the views of all projects are considered, but there is also clear consensus from a particular project that the one project doesn't want them banned? Can they be banned from all but the one project, or will the consensus of the larger community override the consensus of the one objecting wiki? Monty845 (talk) 17:42, 6 July 2012 (UTC)

I think it should be made clear: it's not a "one wiki, one vote" situation. Like all Requests for Comment, the consensus is between the individual editors participating. What constitutes consensus is up to the Steward who closes the discussion, and since it's not a majority vote, it's not really a question of what happens when 10 people vote one way and 3 vote another. Stewards are chosen for their experience doing work across many of the projects, and extremely trustworthy when it comes to adjudicating cross-wiki disputes and decisions. Steven Walling (WMF) • talk 20:13, 6 July 2012 (UTC)

I disagree. We should behave as if we were a federation of peoples instead of behaving as if we were a nation-state of one people. Wikimedia isn't just multilingual; it's multinational as well. I like the concept of "one wiki; one vote", and the current numbers at Requests_for_comment/Global_bans give weight to the "one wiki; one vote" point of view. --Michaeldsuarez (talk) 21:00, 6 July 2012 (UTC)

I didnt read everything in this case. But let me ask u a question. WP is considered an "open system". As far as I can see, those global bans would be only to support switching to a closed one because each community would like to have their sticky users banned globally. What if an admin of a local community calls for a global ban of a user? If it goes like this: "Bann that user, he is obsolete", then Admins are not only acting locally but globally. Who will be able to certify if there are not any personal reasons behind that wish? It has a snowball effect, u know?--Angel54 5 (talk) 15:24, 25 July 2012 (UTC)

The absolute minimum requirement for proposing a global ban is the agreement of two communities to ban the user, so a single person from one project is not able to call for a global ban. WhatamIdoing (talk) 18:16, 26 July 2012 (UTC)

RFC ready for translation

See the draft posted at Request for comment/Global bans. Steven Walling (WMF) • talk 22:15, 19 May 2012 (UTC)

Language nitpick

"A global ban discussion may be invalided when the nominator is clearly not impartial..."

"Invalid" cannot used as a verb in this context. This should read:

"A global ban discussion may be invalid when the nominator is clearly not impartial..."

or

"A global ban discussion may be invalidated if the nominator is clearly not impartial..."

     Jim . . . . Jameslwoodward (talk to me) 20:13, 25 May 2012 (UTC)

Good eye. I've simplified it substantially. Steven Walling (WMF) • talk 23:15, 25 May 2012 (UTC)

RfC ready (translations still welcome)

I've dropped a note at the Wikimedia Forum, but most of the key translations are finished for Request for comment/Global bans. We should get started, and if folks can help advertise it locally on the projects, please do so. Steven Walling (WMF) • talk 21:45, 27 June 2012 (UTC)

I suggest we conduct its final approval via SecurePoll. It's much more easy, have integrated translations, better security options, will give a clear result and will save Meta folks a lot of job. After an RfC regarding whether this version of the policy is OK, securepoll would do it IMHO. Regards. — MA (audiencia) 22:32, 27 June 2012 (UTC)

I've never used SecurePoll, but if it has integrated translations that is a big, big plus. The number one thing slowing this whole process down has been translation. Steven Walling (WMF) • talk 22:40, 27 June 2012 (UTC)

I mean the interface. Tho when I was an election scrutineer for enwiki's ArbCom we had links for translating messages and the like. Regards. — MA (audiencia) 23:40, 27 June 2012 (UTC)

Should we run smth like a global sitenotice, so that it is visible on all projects?--Ymblanter (talk) 07:49, 28 June 2012 (UTC)

Is there any way to translate the headers as well?--Ymblanter (talk) 07:49, 28 June 2012 (UTC)

I was thinking that a banner would be overkill, personally. As for headers: if you want to wrap them inside the ls template, that's the only way I can think of. Steven Walling (WMF) • talk 16:55, 28 June 2012 (UTC)

This has been advertised on wikimedia-l, etc. I think this should be promptly put on CentralNotice, as it affects all projects. John Vandenberg (talk) 07:04, 6 July 2012 (UTC)

Second. My request for a watchlist notice at enwiki have been ignored so far, unfortunately. --Michaeldsuarez (talk) 12:36, 6 July 2012 (UTC)

I'll plus one that now Michael. Wikimania is coming up, so I will likely be mostly unavailable to help with setting up CentralNotice, but otherwise I'm open to helping advertise this more. Thanks, Steven Walling (WMF) • talk 17:32, 6 July 2012 (UTC)

Unregistered Users

Is there a discussion anywhere, on Wiki's policy to allow non-registered users the privilege of Editing? It seems to me in my short year of editing, that 99% of our vandalism is from "non-registered users". Why not require registering and eliminate a giant chunk of this ongoing problem? I was invited here by Meta-Wiki. Is my name coming up red because I never posted here before? Pocketthis (talk) 15:24, 13 July 2012 (UTC)

Because it doesn't work. See en:Wikipedia:Perennial proposals#Prohibit_anonymous_users_from_editing. WhatamIdoing (talk) 16:04, 13 July 2012 (UTC)

Thanks for your reply, I'm sure you won't get angry if I respectfully disagree. This sentence:, "about 76% or 82% of anonymous edits are intended to improve the encyclopedia" may be true, however, those 76-82% would still contribute. Like the article says... "It only takes a few seconds to register". Many Grammar Schools with public computers for the kids are a big part of this problem. Registering may be well beyond their capabilities. It's just too darn simple for a drunk or vandal roaming Wiki and spotting the "Editing Tab" to dive in and wreck our work. That tab should only be visible to a registered user in my humble opinion. This issue was dismissed way to easy by Wiki, and should be reviewed again. I would like to propose a "1 week trial", with no visible editing tag for non registered users, and let's see just how "It doesn't work". Thanks

Pocketthis (talk) 17:23, 13 July 2012 (UTC)http://en.wikipedia.org/wiki/User_talk:Pocketthis

• One more point I'd like to make. This issue really hit home with me for the first time this week, when an article that I had contributed most of the info on the page to, was replaced by a vandal with 10 sentences of profanity. This article gets over 100,000 views per month. It took over an hour for some good watchdog Samaritan to undo the edit. I did the math. In that hour, hundreds of folks hungry for information on that topic (and counted on Wiki for the answers), were greeted by a page of profanity. This should NEVER happen. Wiki is earning a better reputation each year as an aid in education for the masses; and we owe it to all who would search Wiki, to be "Intact" when they get here. Also, many of those who contribute articles but don't register, are those writing about themselves, (like Hollywood) and don't want to expose that fact. For them, the more anonymous the better. I can't think of one good reason a person would refuse to register to contribute; but I can think of hundreds of reasons why a Vandal would prefer not to. Thanks for allowing me to vent on this pet peeve. Pocketthis (talk) 18:31, 13 July 2012 (UTC)

http://en.wikipedia.org/wiki/User_talk:Pocketthis

This page is for discussing Global bans. If your proposal is that an individual wiki should change its policy, the place for that is on that particular wiki. If your proposal is that WMF wikis as a whole shouldn't allow anonymous editing, Meta:Babel is probably a better place to discuss it. --Philosopher ^{Let us reason together.} 19:34, 13 July 2012 (UTC)

• Thank You, and I apologize for the rant. I was extremely surprised that the only place that 'WAID' could point me to was a one paragraph explanation of why someone on Wiki doesn't think registering will help Vandalism; so I vented here. I realize however, that trying to change a policy that Wiki was founded on is nearly impossible.

As far as the issue at hand on this page... I'm for any rational act that will help keep vandals off of Wiki. Pocketthis (talk) 22:36, 13 July 2012 (UTC)

Yes, you're in the wrong place. But since you're here, let me add that it's strange that ClueBot didn't detect the vandalism. What article are you talking about? WhatamIdoing (talk) 19:30, 14 July 2012 (UTC)

other thoughts

I don't fully understand proposal, but I suspect it begins with a false premise. After contributing intensive and informed research on various topics, I have been permanently banned from Wikipedia via one solitary person; not a "community consensus."

The term is pure fantasy used to cover the inevitably and invariably peculiar and generally abusive Wikipedia administrators.

Who ever is suggesting this is probably interested in another petty power play.

Eventually they add up.... Then somewhat later, those who have won power will lose interest or die off..... and the whole project may collapse. Which could be an okay thing. ~~

*Oppose* There is a worry trend where a group of editors spend an inordinate amount of time devising pseudo legal actions when then should just be producing content. Each time is attempts to restrict the free speech of someone they don't like- have they heard of conciliation as a means of co-existance? Not content to limit their power play to one Wp- this one attempts global domination. I oppose in principle and suggest they drop it and do some serious content production. --ClemRutter (talk) 15:14, 22 July 2012 (UTC)

I suspect that you meant to comment at Requests for comment/Global bans. WhatamIdoing (talk) 05:11, 23 July 2012 (UTC)

What if the user does not easily write the language in question?

What will happen if the user does not easily write the language in question, but writes well enough to be understood.--Jax0677 (talk) 16:53, 20 July 2012 (UTC)

If you write well enough to be understood, then there should be no problem. It's not school work; there is no teacher to mark down errors in your writing. WhatamIdoing (talk) 15:16, 21 July 2012 (UTC)

Anyway, if someone wants to ban a user, I think that the discussion about this should be held in a language spoken by the user who might end up being banned. In that case, there is no problem. --Stefan2 (talk) 15:47, 21 July 2012 (UTC)

"Past reasons for requesting a global ban have included"

The language "Past reasons for requesting a global ban have included" does not limit to that list; it seems designed to leave it open for any number of possible global banning justifications. My concern is that people might enact this policy intending that it be used for those enumerated situations, and then find that it ends up being applied much more broadly. Can we figure out what situations this is likely to be used for, and write it so that its potential application is limited in scope tho those situations? Leucosticte (talk) 21:26, 5 October 2012 (UTC)

I believe that you have accurately identified the goal. That list does not include every acceptable reason. Requests may be made for any reason at all. The community can and will reject request it believes unjustified or inappropriate.

I do not believe that it is good to list every possible reason. We do not want to encourage people to believe that bad behavior is acceptable, so long as it is not one of four particular types of bad behavior. For example, I would not want someone to say, "It's okay to harass and threaten my ex-girlfriend on 25 different WMF projects, because she's not 'a contributor' to any of them, and the global bans policy only says that harassing and threatening contributors is a reason to implement a global ban." I would also not want to make an attempt to list the hundreds of specific behaviors that (at least if carried to an extreme) might result in someone requesting a global ban. WhatamIdoing (talk) 18:43, 6 October 2012 (UTC)

Isn't part of the point of having all these different projects that it allows each one to have its own community with different standards, and does not allow consensus on one project to override consensus on another? So, for instance, two communities might have different standards as to what constitutes harassment and how harassment allegations are to be adjudicated. Enwiki has decided to put in place an ArbCom that is empowered to decide that a matter is so sensitive that the community shouldn't be informed of the precise nature of the accusation, of the evidence against the accused, of the content of the deliberations concerning the adjudication, or of how the individual Arbitrators voted; they can just say, "The user is banned by the ArbCom."

There are also some wikis where it's pretty easy to get accused of abuse because the sysops have little oversight, or there's a bona fide cabal. E.g., software developers are showing up all the time at MediaWiki.org, posting a userpage link to their personal software development website as a way of introducing themselves, and getting indef-blocked for spamming. MediaWiki.org doesn't really have much of a Requests for Comment process in place (I think people would rather be coding than getting involved in such stuff), and the policies on such matters are vague or nonexistent, so the sysops pretty much do what they want, unless another sysop happens to notice, disagree, and revert.

But perhaps there are a few WMF wikis, out of the hundreds, that have figured out a way to avoid banning people who don't need to be banned. Unfortunately, the bans on the other wikis would probably be admissible at meta as sufficient evidence that the user has engaged in cross-wiki abuse to clear the procedural hurdles needed to start debate on his global banning, and he could be banned by consensus at meta against those dissident wikis' wishes. So, why not, instead of having global bans that are binding on every WMF wiki, just make it advisory? A consensus at meta to globally ban a user would put his name on a blacklist, but each wiki would be free to ignore that, if it wished, depending on its policies. Or if a wiki wanted, it could have a policy of automatically regarding globally banned users as locally banned too.

If we are going to have global bans that are binding on local wikis, then we should limit the scope of allowed reasons as strictly as possible. I would favor limiting it what is necessary to protect WMF from legal liability; and because most of the community are not lawyers, those are probably matters for the WMF office to decide, rather than the community. So basically, what we need is already in place, since the WMF office already does globally ban people when a legal matter causes them to deem it necessary. Leucosticte (talk) 08:27, 7 October 2012 (UTC)

It seems that the WMF office decided that they wanted to have a community-based process for global bans. This has already been decided. The only decision we need to make is to decide how the community will make decisions about whether a person will be globally banned. Will it be fairly minimal (as proposed here), or will it be practically impossible (e.g., requiring a signature from the Pope and the President)? Will the community take on as much power for itself as possible (as proposed here) or will it limit itself to considering very few situations, so that as much as possible is handled secretly by the office (as proposed by you)? Those are the decisions we have to make here.

As for your proposal, do you really want the office to take on such a role? They have no obligation to limit themselves to legal necessity. They have no obligation to listen to your defense. They have no obligation to be transparent with the community (and they traditionally have not explained their blocks publicly). The office is legally allowed to ban people whose usernames have the wrong number of letters in them, and we have no agreement from them to be reluctant to ban people. They are even legally allowed to block you and make it look like they've done nothing at all, and the reason you can't login is because you forgot your password. On twenty-five accounts in a row. I think that some of the staff members would be perfectly happy to have "globally ban whoever irritates you, whenever you want, with no oversight or input from the community" added to their job descriptions. Is that what you want? Because I believe that's what we'll end up with, if we don't have a process that gives the community a formal role in the decision making. WhatamIdoing (talk) 16:03, 7 October 2012 (UTC)

It is true that Sue did say that she wished we had "a community body with decision-making authority to enact global blocks and bans." I think it's a false dichotomy, though, to say that it comes down to a choice of either having the community handle, say, cases of cross-wiki harassment of non-users or having the office handle those cases. The third option would be to leave those types of cases up to individual wikis. If the harassment doesn't have legal implications, then why does the office need to get involved?

If the office is concerned about individual wikis letting, say, harassers harass freely, then how does it solve the problem to empower the meta community to enact global blocks and bans? If individual wikis can't be trusted to make the right decision, then why would meta be any better at making such decisions? In other words, what makes the larger community a better judge of whether a user needs to be blocked than each of these smaller communities acting independently?

If anything, the smaller communities are better equipped to make decisions for themselves because they can take into account their own unique needs. E.g. suppose a user keeps getting banned from various wikipedias for posting biased political content. That might not be an issue at, say, MediaWiki.org, which is devoted to documenting software; that user might be capable of being a good contributor there, but a global ban would prevent that.

What evidence is there to suggest that the office is worse at making these decisions than the community? Shall we look at individual cases? I've never seen anyone have a problem with the office that wasn't corrected, but I've seen a lot of people have problems with certain communities. One of those communities is enwiki, whose users would probably tend to dominate discussion at meta since enwiki is the largest and most active project.

So yeah, I'd be inclined to trust the office more than the community. At least at the office, people can get fired for making bad decisions. What happens if a user makes a bad decision as to whether to support or oppose a global banning? Nothing; he is allowed to continue participating in global banning debates, as long as he doesn't break any rules (e.g. by being uncivil). At the office, everyone is accountable ultimately to the CEO; in the community, there is no such accountability. Leucosticte (talk) 16:32, 7 October 2012 (UTC)

That sounds remarkably like, "I trust everyone except the people who banned me". WhatamIdoing (talk) 16:25, 8 October 2012 (UTC)

That would tend to make one not trust a certain process or group, would it not? The self-serving bias tends to make people believe they are in the right; therefore, those who took the contrary view will, from that perspective, be deemed to be in the wrong, and therefore the action they took in accordance with that belief will be deemed to be wrong, which in turn will tend to lead to a perception of a problem that should be corrected; in this case, that problem being the banning system. Of course, those on the other side have their own biases.

It's not just my case that I thought was poorly handled, though; I disagreed with the banning of Suarez too, and I think Abd's banning was also probably unnecessary. And of course there is the ArbCom-supported Wikipedia censorship that has been documented at NewgonWiki. People who have been through the system and seen how the banning and ban appeal processes operate are, in some respects, in a better position to understand the failings of those systems. Those who haven't been through them may not be in as good a position to judge, because those systems are not always transparent. Without transparency, it's hard for outsiders to know whether good procedures are being followed or not.

But "sour grapes" is a bit of an ad hominem argument anyway. Of course it's going to be the people who have bad experiences with something who tend to oppose it. If you buy a product and find that it was a ripoff, is it "sour grapes" if you go around telling everyone how bad it was? Or if you get fired from a job and then try to trash that company's reputation? The bias of such people has to be taken into account, but that doesn't mean there's no truth to what they say. They were there; others were not, so in some ways their information could be better than those who are in a position to be totally objective due to having no involvement. Leucosticte (talk) 17:35, 8 October 2012 (UTC)

@WhatamIdoing: Experience tells me that the WMF is hesitant to take any action. The WMF isn't bold. They do their best to avoid controversy and backlash. They won't made the "hard decisions." The only reason that they finally decided to do something is because they were under pressure and, in my opinion, afraid of how the media might react if they didn't take action. The "community", on the other hand, will make bold, hard decisions. The "community" isn't afraid of backlash; they shrug backlash off. the "community" doesn't care about the media; they believe that the media (e.g. Fox News, copyright advocates) is evil. The "community" isn't afraid of becoming fired or being forced to resign. The "community" will ban more people than the WMF. --Michaeldsuarez (talk) 18:20, 8 October 2012 (UTC)

http://meta.wikimedia.org/w/index.php?title=Requests_for_comment/Global_bans&diff=4223637&oldid=4218099 – Leucosticte isn't the only one who's concerned. --Michaeldsuarez (talk) 18:39, 8 October 2012 (UTC)

Binding or advisory

Is there any reason the global bans have to be binding on local wikis rather than advisory? Might we not allow each wiki to decide for itself whether the decision made at meta, and the evidence presented there, warrants banning the user on the local wiki? Some wikis might choose to have a policy of automatically banning globally banned users, while others might choose to have a policy of ignoring global bans altogether, and there could be many positions in between. It seems like this would be more in accordance with wikifederalist principles.

We could simply have a global page for informing the Wikimedia community as a whole of what users have been banned from what projects. That would be an informational resource that local wikis could use for whatever purposes they wished. There would be no need for a global Request for Comment; anyone could simply add the information as soon as a user were banned, and link to the appropriate page on the local wiki providing evidence of the user's being banned.

This could allow local wikis to adopt more nuanced solutions than what a global ban would accomplish. E.g., one of the wikipedias might have a policy that anyone banned from three wikipedias with article counts of more than 1 million would be automatically banned from that wikipedia as well. There are any number of creative possibilities that wikis might adopt if given the freedom to tailor solutions to their unique needs.

This type of advisory or informational system would have a similar function in Wikimedia as a credit rating or reporting system has in a market economy. If you are a businessperson who knows that three prestigious companies have reported being ripped off by a particular customer, you are still free to either engage in a transaction with that customer or not. Perhaps you have reason to believe that the nature of your proposed transaction with him is such that you are safer from being ripped off than those companies were, or you simply have a different idea than those companies do about what it means to be ripped off. E.g., if the customer can give you an explanation for his failure to pay that you find acceptable, then you might do business with him, despite the fact that those companies have vowed never to do business with him again. Leucosticte (talk) 09:09, 7 October 2012 (UTC)

Yes, there is a reason why global bans have to be binding on all projects: the Terms of use say that they are.

There could additionally be a "recommended but not required ban" or a "global-except-for-this-project ban" system, but there must be a process for deciding whether to issue a ban on all projects.

There is no requirement that any such ban ever be issued. The community is free to reject every proposal proposed ban on an individual. And the fact is that very, very few current users have been banned on two or more projects, so almost no one could even be nominated. I can only think of one, and his current activity is so limited that it would be pointless. WhatamIdoing (talk) 16:13, 7 October 2012 (UTC)

What are the consequences if the community simply refuses to adopt a global banning policy? What would the office do then? Maybe it would lead them to reconsider their stance and change the terms of use.

What is your point in saying that the community is free to reject every proposal? It's debatable whether it's good or bad to have a system that doesn't apply rules the same way consistently; see wikibureaucracy. Wikibureaucracy has been mostly rejected, so people are at the mercy of arbitrary community decisions that ignore all rules. However, one of the few safeguards is that if you get banned from certain wikis, you're not banned from them all. This would weaken that safeguard.

Even if it's only a few cases, that doesn't mean we can necessarily expect that it will remain only a few cases; and these sorts of policies, once established, are notoriously hard to get rid of. I hope it will remain only a few cases, and that limiting the policy's application to those who are banned on more than one wiki will make it not apply to very many people, but who knows. And in any event, it seems likely to do more harm than good in those few cases, and could also contribute toward the development of an anti-wikifederalist trend in policymaking, so I vote against it. Leucosticte (talk) 16:55, 7 October 2012 (UTC)

I have clarified my comment in re rejecting "proposals". The community is free to reject every single proposed ban of any user at all.

I don't know what the office will do if this (or any other) proposed policy is rejected. I wouldn't exactly be surprised if they simply imposed an interim policy by fiat, and then left revisions up to us. As I read the TOU, they actually need to have something in place, whether we want it or not. WhatamIdoing (talk) 16:20, 8 October 2012 (UTC)