Discussions on meta are, with some opposition, heading toward support for "global bans" as being something that either could not be locally overridden, or that might require 80% local support, with a minimum of 50 users supporting, to allow a globally banned user to edit locally.

Given that when there is a ban discussion of an user who is unpopular on Wikipedia, on another wiki, Wikipedians pile in, in significant numbers, even though they never edited the smaller wiki, that basically means that Wikipedia Rules. Even if most users of a small wiki want the editor to edit, even if there is no local disruption at all, other than the attempt to prevent the editor from editing.

The only case behind this is Poetlister. As I feared would happen, Poetlister is the poster boy for a campaign to strip local wikis, most notably Wikiversity, of their right to make local decisions about who can edit and who can't. The reality is that meta is dominated by Wikipedians, which alone can muster hordes of users to comment in ban discussions. Wikiversity would be setting records if 50 users commented in a process. That's nothing on Wikipedia.

The principle of local autonomy was well-established. Global blacklists and global blocks have mechanisms built-in for local override, any admin can do it. Global locking was new with SUL and the local tool wasn't created to override, but the renaming trick was discovered, and used on a number of wikis in the case of Thekohser.

In discussing this on meta, it's pretended that Poetlister -- and another obscure user -- are the first global bans. In fact, there were Moulton and Thekohser before, those were declared bans, and they were -- and still are -- what Wikipedia calls "defacto bans," i.e., nobody unwilling to lift the lock.

The Empire Strikes Back.

Poetlister is about as unpopular as I could imagine, but the bottom line is that Poetlister is not really the issue. That's a smokescreen. The issue is the right of one wiki (meta) to decide affairs for another (Wikiversity in this case). It's a crucial case, in my opinion, because my view has been for some time that the future of the WMF lies with Wikiversity, just as the future of encyclopedias lies with academia, which is always freer and far more open than an encyclopedia, which is necessarily very limited.

The proposal, because of the political realities at meta, is like putting the non-academic editors of an encyclopedia in charge of a university, able to eject professors and whole subjects based on their opinions.

And SBJ is leading the charge, carrying water for these forces, blocking Poetlister on Wikiversity per the global ban, in direct and flagrant disregard of Wikiversity blocking policy, which is quite clear on the topic. And which is the consensus he thinks is necessary. It's obviously personal, because with Thekohser, he was clearly on the side of "blocks require consensus, not unblock." And Wikiversity policy absolutely doesn't allow blocking without disruptive behavior. With Thekohser, there were, at least, allegations of disruption. There weren't any for Poetlister, the guy has been nothing but a positive contributor on Wikiversity, and that could really be said about his sojourn as Longfellow on Wikisource. On Wikiversity,though, he's been open as to his identity from the beginning.

And his enemies turn that into an accusation: he's only there to "worm his way into the confidence of the community." Isn't that what half the admins have done on Wikipedia? It's a crime?

It ought to be, at least when done voluntarily. It's more understandable if one is being paid to do it.



Meanwhile, nobody finds the humor in the fact that this entire Meta discussion was brought up by this chicken-petting man?

We need a more effective and lasting way to mock and belittle Walling. He's poison, and nobody cares.

If only I could find some real dirt on his past activities. He's a dull drone and a Facebooker.

Global bans is a blatant power grab. The dominant faction(s) at Wikipedia can easily push meta discussions any way they want, it's a simple matter of numbers. Arguments don't seem to count any more. The process does not work to see that arguments are actually vetted, with evidence. So it's just whatever people think, knee-jerk. As on Wikipedia, that's mob rule, it's not "deliberative democracy."

In a sane process, before the conclusion is even proposed, the evidence is collected and vetted. Then proposals are made and debated, and any conclusion explains what evidence was accepted, and why, and what arguments were accepted. As it is now, discussions often start with the conclusion. It's essentially verdict first, trial later, shades of the Queen of Hearts.

There is a utility to global bans, it's not unreasonable to set a default. What is new is an explicit claim that a discussion at meta can control not only a user, but also whole communities. That is, if a discussion at meta concludes "ban," by the proposed policy, local communities cannot over-ride that.

In a current proposal, override is allowed, but only using unobtanium. I've seen it. When there is some Wikipedian concern about activity on another wiki, based on personal animosity or Wikipedia history, and a user is discussed, Wikipedians pile in, seriously warping what would otherwise be local consensus. The proposed bypass requires a discussion with 80% support and at least 50 editors voting. I have never seen a discussion on Wikiversity with that level of participation, on anything. It's routine on Wikipedia.

There is no way to get 80% for even a popular local user, if Wikipedians are opposed and show up.

Thus the system is being set up so that Wikipedians can control the whole ball of wax. And it's not being made clear what a drastic change this is.

This is, quite possibly, long-term payback for local communities unblocking Thekohser. The excuse is Poetlister, perfect for that because he's so unpopular. What's the harm?, people will think, for, after all, he's dangerous and really a Bad Person. Never mind that his record for well over a year is only positive. Never mind that he's been editing openly without any disruptive actions, and he's easily watched. Begone with him!

And with you, as well, if you oppose this. It was just proposed at Wikiversity that I be banned because I pointed out that policy there prohibits a local custodian from blocking Poetlister unless he's been locally disruptive.

And who is even aware of this happening at meta? Consider all the actions that have been taken before, where meta positions were acted upon on the other wikis. Under the new policy, those actions wouldn't be taken directly. Instead, there would be a ban discussion at meta. These discussions attract only small numbers of editors, most editors never even think of going to meta. They aren't going to see the arguments and understand the issues. All those prior actions, which caused enormous disruption in 2010, will now be justified as "community decisions." But it will be one community making decisions for others. That's quite new.

The software was designed so that the individual wikis were independent. The global blacklist can be overridden by any admin on any wiki, likewise global blocks and the pagename blacklist. Global locks were, I think, the last facility added, with SUL, and it makes sense for vandalism and spam, and the rules used to say that it was only to be used for that, if disruption was massive, cross-wiki. It's crept down to the point where I've seen accounts globally locked with only one or two cross-wiki edits. All it takes is someone to go to meta and make a request that seems okay to a steward, and the local user hasn't been informed, and the action is invisible to the local wiki.

Poetlister was initially locked with no discussion anywhere visible. Clearly to prevent him from editing at Wikiversity, it had no other effect.

There is no local whitelist to bypass locks. It's just a bug, essentially, and it wasn't considered terribly important, because users could just create a new account (which is legitimate on a local wiki, if the identity is disclosed and local admins permit it), or a 'crat could rename.

But now comes the global banhammer. They want to make the whole WMF family of wikis be subject to Wikipedian authority, and I've seen what that authority looks like, as it's operated at meta. It's ugly, it's vendettas and cross-wiki harassment, but tolerated. I confronted some of this, some blatant examples, and if you want to know why I'm currently blocked there, just look at the evidence that was presented!

In a venue where one of my alleged offenses was discussing proposed blocks or unblocks. In other words, an admin noticeboard, and supposedly my offense was being a non-admin and commenting on proposed admin actions. It was literally said, by a meta admin, that this was not to be allowed. What they want is for non-admins to be excluded from participation, yet they will then assert -- it's being asserted -- that the "community" banned. Yeah, "community of admins." Others not welcome!

That's the goal behind this move. Central control. Good luck, folks.

Ya think?!

Actually, if you read between the lines of Steven Walling's comments on meta, he actually defends the local communities' unblocking of Thekohser, which I found rather interesting. Maybe when I finally finish fixing up the chicken house you and he can come over for goat cheese omelets, Greg!

I don't suggest reading between the lines, the print is exceedingly small and you will get eyestrain. Not to mention a bad case of fantasies.

I don't think I've seen reference to Thekohser by Walling. As I read it, Walling is suggesting that local 'crats cannot overturn a global ban. This would also be extended to local crats not being able to permit a user to make a new start locally. He has the idea that if a local "ArbComm" decides to unban, then they should be required to get an meta decision to unban. In other words, if I read between the lines, they *cannot* just act on their decision. The power remains at meta.

It's a radical restructuring, in fact, a boot through the door that protects against domination by Wikipedia.

There is the odd structural weirdness. An "ArbComm" can make a decision to "request an unlock." But that would be a local ArbComm deciding that the rest of the wikis are safe from the banned user. Somehow they think that having a user globally locked, with one or a few local exceptions is "too complicated." It's actually quite easy, and CentralAuth shows the account as detached, if it's been detached.

Anyway, ArbComms typically make decisions by majority vote. Suppose there is a seven-member ArbComm, and suppose it fairly represents the users. I.e., a vote of 4:3 on the ArbComm would be considered adequate to allow an unban request. But if it's not a wiki with an ArbComm, they want a vote with 80% of users supporting the unban, and at least 50 editors commenting. The discussions I saw on Wikiversity and Wikibooks with Thekohers had a handful of editors commenting, it might have been ten or so. You couldn't get 50 editors voting on those wikis without massive canvassing. If then.

Basically, Walling wants to outlaw the process that led to the unblock of Thekohser. Poetlister is just an excuse. Walling also now wants to propose the general principle, abstracted from realities. That's another device for luring communities into making foolish decisions. Real courts don't do that, they *require* real cases, and establish precedents through them. *Law* is created, ideally, through deliberative process, with hearings, evidence, etc. And a process that ensures that alternatives are carefully considered. (In theory. In practice, of course, politics happens, stuff gets railroaded through, etc. But voluntary organizations that use established democratic process mostly operate fairly, Robert's Rules, for example, is designed for that.)

What Walling is setting up is the possibility of a situation where, say, 3/4s of editors on a wiki want to allow an editor to edit (which is enough to elect a sysop, or remove one, typically), but they are denied, by people who are not part of their community. Who don't have to give any reasons, in fact. It's just No.

And the problem with allowing wikis to make their own decisions is? What's the actual damage that has been done? What are they trying to fix?

When there is a discussion of a possible ban/unban on a local wiki, the discussion is supervised by adminstrators on that wiki, who can prevent or sanction harassment. The wikis, in general, have been operating on the assumption that local bans require local misbehavior, and the evidence is available to that community, including material that is deleted. When this is done cross-wiki, it all becomes obscure. People are set up to vote on things where they do not have intimate knowledge of the conditions involved. On meta, it's not uncommon, various trolls slam users, lies are told with impunity, and few have the patience to research claims about what's taken place in other languages. I've investigated a few situations, and it's tedious.

I've seen what looked like admin abuse on, say, es.wikipedia, and I absolutely stayed away from accusations of that, because it's not my community. I saw a problem with what I called the "Dutch cabal," but in no way questioned the right of Dutch users to run their own project. What I didn't like was them pursuing users they'd banned, across the entire WMF, seeking bans elsewhere based on what the poor fellow had allegedly done on nl.wiki. (There was more than one of these, and the cross-wiki behavior of on nl.wiki sysop was utterly outrageous. If she'd showed up at Wikiversity and did what she did on other wikis, I'd have blocked her. It was harassment. Her behavior on meta was disruptive, she filed an RfC on the user that was totally malformed, just a pile of accusations without evidence, incoherent, and then she left on vacation. I closed it. The closure stuck. Most of what I did on meta was supported by consensus, but a few prominent cases where the situation was marginal -- or Poetlister -- is what has stuck in people's minds.)

It's pretty obvious to me. They are trying to fix and eliminate defiance of Their Authority. They are trying to impose a uniformity on WMF wikis, rooted in Wikipedia practice, basically a nightmare.

I've become quite clear on the goal. It's not protection of the small wikis. It's empowering and encouraging some of the worst elements in the community, those who want to ban others, instead of developing cooperation between diverse people and interests.

It's sick.

Ottava babbled forth the following over there:



Jeffrey, could you point to one iota, just one smidgeon of "Wiki based insanity" that has come as a result of my lock being broken by local Bureaucrats on Commons, or Wikiversity, or Wikisource?

This will hurt those banned editors who take the "Redemption Island" challenge in which they work on different wikis to prove they are repentant and worthy enough to return to their mother wikis. Now if you piss enough people off on one wiki, you have no hope of ever returning or even positively contributing to other WMF projects. This also undermines each community's dispute resolution processes. No chance at redemption unless the Meta people decide one is repentant and worthy enough to return.

http://meta.wikimedia.org/w/index.php?titl...7&oldid=3051830

Let's see if they'll consider lowering the threshold to 65%.

Let's not. Please. Don't encourage them in any way.

Let's force the little toads to give up and go away. Hand the whole thing over to professionals,
or at least something more closely resembling professionals, than flakes like Walling or Darklama.
They are man-children, not "professionals".

It won't solve the problem. We know from the past that the cabal can muster lots of editors to pile in at any discussion anywhere. On local wikis, decisions are made locally, especially by local admins, who can deprecate votes from non-participants at the wiki. Stewards are not going to go to that work, they want to look at a discussion and avoid complexities.

Further, setting the threshold above 50%+ creates minority rule, it's well known that, long-term, consensus organizations run into the problem of persistence of past errors if it takes a supermajority (not to mention full consensus) to reverse a past decision. There are serious problems with supermajority requirements even for administrator promotions, the Wikipedia situation can be ascribed to this.

Basically, supermajority for *demotion* produces a situation where an admin is actually disliked by a majority, but continues. Supermajority for promotion sets a requirement that results in under-representation of even majority positions, someone supported by a majority is still unable to obtain the tools. Difficulty in removal makes this seem reasonable! I argued for easy suspension of privileges, almost automatically by ArbComm, or easy issuance of injunctions prohibiting certain actions or classes of actions. Voice crying in the wilderness. Again, supermajority election (by plurality-at-large) of Arbitrators guarantees overrepresentation of a popular faction, thus ArbComm does *not* fairly represent the community.)

All this is well-known among election methods experts, who tend not to fare well on Wikipedia if they point it out. There is an easy proportional representation method invented by Lewis Carroll, in about 1883, which could easily put together a representational system that would truly represent *every user*, directly or indirectly. This could be done ad hoc, using the representational system suggested in [[WP:PRX]], or, as with the original invention, using secret ballot election of what might be called "delegates," who would then openly assign the votes they "own" to create seats on the Committee, according to a quota (i.e, if the Hare quota is used, if there are N seats to be elected, and there are V votes, the quota is V/N. And it's even possible to allow direct voting by delegates on issues -- but not their right to participate as arbitrators, in deliberations. If that's done, *every decision becomes a representation of true majority.*)

Conditions to enable this are being set up at Wikiversity, and you can bet there are forces which do not want to see that happen. Not a majority on Wikiversity, but if Wikipedians can control, as the new proposals are seeking, it's over, I predict. This reform, which could transform decision-making (long term, this won't be quick), not only for Wikiversity, but, then, by imitation of an efficient, functional example, the whole WMF family of wikis. And more. That's not possible without a working model to study.

[[WP:PRX]] was explicitly an experiment, to see what would happen if the file structure were set up on Wikipedia. It was all voluntary, no policy changes were involved. Crushed, as quickly as possible, the proposer banned in short order, for offenses that would ordinarily have done nothing more than raise eyebrows and maybe a warning and, in the extreme, a short block. A long-time user, by the way, with a clean block record (as to recent accounts, I think he had really old accounts, years back, not disclosed, with problems. Those were not the basis for any blocks. This was an early Wikipedian. A bit crazy, like many of us.)

Supermajority requirements for changes or decisions lead to extreme conservatism, the creation of a dominant faction which can be in the minority, as to actual support, but which retains control. That is why no sane deliberative body uses supermajority. ArbComm doesn't! They use majority rule like every other body that must make decisions. That requires, of course, a defined membership. Supermajority rules were based on the idea that a subset of the "membership" can be presumed to be representative of, at least, the majority. It's not completely crazy!

Traditional process, in fact, has something like that. Under Robert's Rules, an absolute majority of members can change *anything* about the organization, including the Bylaws. Otherwise, to change Bylaws requires a supermajority (2/3) of all members voting after notice. Which is often a small fraction of the total membership.

What has fascinated me about Asset Voting (that 1883 thing, Lewis Carroll -- Charles Dodgson) is that a body elected by this means, representing *all* the participating members, with individual members not needing to know the complex politics of the organization, they only need to know whom they most trust, could, in theory, exercise the full power of the members, a *represented absolute majority* does become possible.

It's never been tried on anything other than an extremely small scale, where it worked spectacularly, achieving what many experts commonly think impossible (when they don't consider the Asset process, which is not what experts call a pure "election method," but only, say, Single Transferable Vote, which is next-best, and which really requires a party system to work, Dodgson knew that and was trying to fix it).

abd, I just want to say to that I agree with your views about supermajorities and minority representation completely. During the 2011 ArbCom Election RfC, there was a push to establish a threshold when such a threshold where one didn't exist previously:

http://en.wikipedia.org/wiki/Wikipedia:Req...nt_to_ArbCom.3F

In previous ArbCom elections, all open seats were filled regardless of whether the candidates received over 50% votes of support from voters (although they always managed to received over 50% anyway). In this RfC, they wanted to create an official threshold. A lot of users even wanted to threshold to be 60%.

I made the following statements:

http://en.wikipedia.org/w/index.php?title=...oldid=451861770

http://en.wikipedia.org/w/index.php?title=...oldid=451989706

I was against establishing a threshold. I knew that such a threshold would disenfranchise minorities on Wikipedia.

Towards the end, it appeared as if those supported a 60% threshold were going to win out, so I made one final, bold comment:

http://en.wikipedia.org/w/index.php?title=...oldid=458952065

I subtly suggested that "no minimal percentage" !votes should be added to the !votes for 50% in a manner reminiscent of preferential voting.

Unfortunately, the decision closed with 60% stated as the new threshold:

http://en.wikipedia.org/w/index.php?title=...oldid=458995226

But then, Timotheus Canens convinced TParis to change his mind:

http://en.wikipedia.org/w/index.php?title=...799#ACE2011_RFC

http://en.wikipedia.org/w/index.php?title=...oldid=459309740

Although I didn't want any threshold, I had to settle for the 50% threshold. Just as !voters for "no minimal percentage" couldn't outvote !voters for thresholds, I can't "outvote" those in favor of the global block policy. I don't have the means to prevent a threshold from being established, but I can negotiate for a lower one.

50% makes sense when one is seeking a single seat or position, single-winner elections.

The method being used is, generally, Plurality-at-large, except that one may vote for more than one candidate, which makes it kind of Approval-at-large, which is known to be worse than plurality-at-large (in which you get one vote only). For plurality at large to produce some approximation to proportional representation, it's important to limit the voting to one vote per election. Otherwise the following happens, easily:

Suppose the membership is divided into two factions: Call them the inclusionists and the deletionists. Suppose the deletionists are 51% of the voters, and the inclusionists are 49%, and suppose that the voting for each seat is separate, so you can vote yes or no for each seat. If all voters vote according to their faction, all seats will be awarded to deletionists. There will be no representation of the inclusionist faction on the body to be elected.

This really happens in legislatures which are single-district elections, if the parties are spread evenly through the population. If a party has a majority in each district, it gets *all* the seats. Because many districts are being separately elected, real proportions are merely distorted, often drastically, toward the majority faction. It is easily possible that a majority in the legislature represents a minority of the voters, it depends on how they are distributed.

Single transferable vote was invented in the 19th century to address the problem. I won't detail how that works in detail, but for multiwinner elections, each voter, in a district election, gets only one vote, but multiple seats are being elected from each district. Voters vote a preference list, and candidates are eliminated, and the votes for those candidates are transferred to the next preference, so only one vote is being counted at any given time.

It's extremely complex, I think it's an NP-hard problem. But I'm not an expert on that, I just know that canvassing STV elections can be enormously complex, and an error at one stage, not discovered until later, can require redoing all subsequent counting. Australia uses the method, and, while supposedly it helps minor parties, in reality, as shown by a lot of analysis, it doesn't. What it actually does is make the presence of small parties safe for the large parties.... so they can ignore them. The method, when used single-winner, as in Instant runoff voting in the U.S., can result in very strange results, where candidates preferred over every other candidate, if the ballots are analyzed for that, can easily lose, and have (as in Burlington VT, recently).

There is a quite simple solution that would be easy to apply on-wiki. Asset voting. I'll state it how I think it might be implemented. The election is held with a secure server as at present. Voters can vote, anonymously, for one candidate, the only restriction is that the candidate must be eligible, and I won't describe eligibility. Suffice it to say that it might be a long list of candidates, it doesn't matter, because no votes need be wasted (as happens if you don't happen to pick a "winner" in an STV or plurality method).

The ideal vote in this system is the candidate whom you most trust to represent you if you can't participate in a decision yourself.

All candidates receiving votes are now called "delegates." In a lot of writing on this, I've called them "electors." These candidates, in a negotiation in which only they may participate, may recast their votes to create winners, as if the votes were their "property." That's why it's called Asset voting. It takes a quota of votes to create a winner. I suggest the Hare quota, for reasons I won't explain; that means that it's possible that not all seats will be filled, if a delegate refuses to compromise as needed. But it also means that all seats are elected from exactly the same number of votes. And it encourages delegates to make the necessary compromises.

If the delegates refuse to compromise, those who voted for them will know, because all the subsequent process is completely open. They may decide to vote for someone else next time!

Generally, voters will know exactly whom their votes elected. That's spectacular, as a result.

There is a reason why systems like this have not been implemented. It would create true, fully-representative democracy, which is quite threatening to some. Even inferior methods (but improvements) have been killed by central authorities, such as Bucklin voting in the United States, based on spurious arguments generally rejected by experts. Bucklin was very popular with voters! (It's a single-winner method, and it worked. It was also used for multiwinner elections, where it might be better than plurality-at-large, but where it is still vastly inferior to PR methods (such as STV-PR).

In the U.S., STV-PR was rejected based on public campaigns where it was pointed out that it would elect Socialists and -- horrors! -- Negroes. New York. Basically, that was a rejection because it worked.

So Wikipedia machinations surprise me not at all. Okay, I was a little surprised. I did think that there was a little more sentiment in the community for fairness than actually turned out to be the case. I was fooled, for a time, by the presence of old-timers who did understand some of the issues. They mostly left, burned out by the mob.

At the end of my tenure as an unbanned editor, I'd found that there were very few willing to investigate issues, i.e., to actually vet evidence in a ban discussion, not to mention to look for evidence independently. Too much work, probably. Easier to just join whichever side you favor, and that's what most have done for a long time. I'll confess, as well, when I realized that ArbComm was quite the same, with only rare exceptions (as happened in RfAr/Abd and JzG. The same arb did the same in RfAr/Abd-William M. Connolley, and he was simply shoved aside.)

And there goes neutrality. Unobtainable.

By the way, Michael Suarez did review the evidence presented in my block discussion on meta. Shoved aside. That experience, repeated over and over, leads to loss of people who will do that.

And now meta heads want to impose this on all the other wikis, forcefully. If that succeeds, well, fixing WMF process will become far more difficult. It's still possible now, though not easy.

It strikes me that if all of those who have been banned on en.wikipedia (but not elsewhere) were to go to Meta and !vote on this ridiculous proposal, what would be the outcome? I predict that, despite no negative history there or on other Wikis, we would all be summarily global-banned. Any other guesses?

It might make the proposal fail. It depends.

It would expose the hipocrisy of the proposal.

By the way, from one point of view, the proposal is sound. Before, "global bans" were determined, essentially ad-hoc, by stewards, and with Thekohser, it was clearly done in contradiction of a previous apparent consensus. Was there some private agreement that a majority of stewards signed up for, or was it a unilateral action that other stewards didn't care to challenge? I don't know. I just know that Mike.lifeguard stated that his reinstatement of the global lock for Thekohser was claimed to be a result of "discussions." When he was asked where these discussions happened, he declined to answer.

The problem is this new idea that local wikis *cannot* reverse the effect locally. That's the radical change.

I've pointed out before that if banned editors were to cooperate, the position of Wikipedia about banning would become untenable. I.e.., every time they banned someone, they would add power to this dissident faction. If the dissident faction were organized off-wiki, with efficient organization that actually finds consensus rapidly -- it's possible! -- it would be unstoppable, it could outmaneuver the highly inefficient and divided Wikipedia on-wiki structures, easily.

I will not detail what such an organization could do. But it definitely could shift the balance of power. It would also be far safer than people would imagine. The organizational structures I'd suggest are designed to be difficult to corrupt, and a consensus that was actually harmful to Wikipedia would be very difficult to maintain. Rather, this "cabal" would work for fair and inclusive process, I'm sure of it. That's what most banned editors want, though they often have had their own individual factional agendas.

They actually want the original wiki vision, what we signed up for originally, most of us. They want what Wikipedia pretends to be, but isn't.

I seriously doubt it... do you really believe that?

I wrote that "it depends." If WP banned editors make civil comments, and don't argue tendentiously, the chance of them being banned on meta for it are very low. Notice how much Ottava has had to do to get blocked there, in the past, and how much he's doing now. The chance of a global ban is even lower, unless they have made separate enemies at meta. Simply being banned on Wikipedia is not normally enough.

However, conditions aren't really "normal" lately. I was astounded to be blocked for as little as I did, so maybe.

My point is that if the door is opened, it will eventually be used. And will be used more and more. But right now, that door could probably be closed, if enough warm bodies could make cogent comments. No socks, please (if you are blocked on meta, please don't create a new account to comment, it will have a poor effect). Real editors, and arguments are what's important. The Old Way might still work.

The issue is not really global bans, per se, there is no way to stop them, because of the lock tool. However, the dangerous idea is that local wikis cannot overturn this, if the proposal goes that way. Looks right now like the 80% idea is dead, or was a misunderstanding from the start. Or a trial balloon. The key is that the decision to overturn or not should be local. Not judged at meta. This is the status quo, it's been established (with a lot of work, in fact, it would be a shame for that to have been wasted.)

In the global ban discussion, there is reference to the draft Terms of use. It's claimed that for a "banned user" to access the service would violate the TOS. Steve Walling has written on the Global bans talk page, that
This is a reference to section 10 of the TOS, which is about site management. However, notice: this is legally dangerous for the Foundation, because the issue would be repeated violation of section 4 of the TOS, and it would not be enough for "the community" to determine that, if the Foundation were bound by the Community, it would be bound by a fool, and it knows better. If the corporation makes a decision infringing on the rights of a user, and especially defaming the user, it would become liable. This is the last thing that the WMF wants to do, unless it's really necessary. The TOS goes on
The legal issue I see here is that if the TOS is abrogated by the WMF, whether or not the user remains bound by it. "Banning" is a term of art on WMF wikis, it has no legal meaning. Blocking has a legal meaning, because it is a physical situation. An account is blocked or locked. As I see it, this can only be done, under the TOS, upon a finding of violation of section 4, and the corporation cannot delegate this as a matter of TOS enforcement to the community. The community may block, but it may also unblock. What is being implied in the TOS discussion is that a "global ban," which is being defined by meta, becomes legally binding on all the wikis. Legally, I think this thing is unenforceable. The WMF cannot evade its responsibility. It's not clear under the existing TOS if the WMF can take legal action against a block evader, Newyorkbrad tried to threaten me with that, and I pointed out that it would be laughed out of court. I'd love to be sued by the WMF, because it would be like shooting fish in a barrel. I know already that I'd get support with legal expenses, and it would be a winnable case.

I don't think they have any interest in going there. They lose too much if they lose, and they gain almost nothing if they win. And I am under no injunction from the WMF not to edit Wikipedia, say, nor is anyone, to my knowledge. I believe that Poetlister is not. "The encyclopedia that anyone can edit." It will make for splendid news if they try. Bad Idea, I'm sure that they aren't stupid enough to try this. But, hey, I've been wrong before with predictions like that.

The WMF avoids liability by allowing the communities to run the wikis, including blocking under whatever cockamie policies that the communities choose. What is being attempted with the new ban policy proposal, combined with the new TOS, is to have their cake and eat it too, to maintain the idea that the communities are responsible, but to enshrine in the TOS, by reference to ban policy, the superior authority of meta, which contradicts the rest of the TOS. Is this a meta ban or a WMF ban? The WMF has kept its hands off the meta ban of Poetlister, the only ban so far to attempt to follow the proposed ban policy. Since what was considered there was not violations of the TOS (I know people argue this, but those arguments would not stand up in court, where mob rule and mob thinking fails to impress), it's really all a charade.

So what if the WMF formally "bans" a user? How would this be done? Legally, it would be necessary to notify the user, and normal on-wiki notice wouldn't legally suffice. Do they really want to open this can of worms? Because if they violate due-process rights of users, in order to jump in, they become liable for errors -- not to mention malice. They would have to actually investigate claims of violations of the TOS. In the past, they have not wanted to touch this with a 10-foot pole. That's why I think that the Board members who have been involved in this truly are as they say: operating on their own, not approved by the Board and not representing Board policy.

The proposals are naive and not crafted by a decent attorney, my opinion. The only thing that I can expect this would do is to terrify certain naive users, who imagine that the WMF is going to shut down the wikis if anyone defies a meta ban, or Jimbo ban, in the past. Not likely, not likely at all. If a local wiki truly does act to cause damage to other wikis, sure, then the WMF would need to choose. Short of that, no. But they can make it look like they will, so that's why it takes users to stand up to this nonsense.

Post on meta that shows the problem:
Right. The last thing Wikipedians want is outside interference. A global ArbComm, could it overrule a local ArbComm? Could it overrule a local consensus? The WMF can, that's clear, because it owns the sites. In other words, the WMF is the "Global ArbComm." It only takes cases where there is legal hazard to the WMF.

Yes, weird things can happen on really small projects, but really large projects can be dominated by small groups, if the groups coordinate (deliberately or naturally), and that includes Wikipedia. About two dozen editors, for years, were able to paralyze response to blatant recusal failure on the part of the group's administrators. They'd pile in and oppose a finding of recusal failure, even when it was blatant, they were able to run about 50% of the voting, easily, and then they would claim vindication because no consensus was found. Usually these discussions were not closed, they just faded into archives. And other sysops were in despair over it, and I found some examples where admins left the project over this. ArbComm didn't want to touch it, because they had no solution to the problem of participation bias, the whole system, on which they rely, is vulnerable to it.

A global ArbComm may indeed have been a Bad Idea; the basic problem is setting up structures where one group of editors can control the activity of another, different group. When I participate in Wikiversity, I don't want to have to look over my shoulder at, say, meta, and the vast majority of users don't. The Poetlister ban discussion was indeed announced on Wikiversity, but it did not invite discussion on Wikiversity, even though that Wikiversity is the only site that was actually affected by the proposed ban, the only place where Poetlister was openly editing. So editors who assumed that local wikis control local business may not even have looked at it, and the issues were quite complex (I spent days researching this), so what did they know?

But the issue of whether or not Poetlister was causing problems at Wikiversity was -- and remains -- simple. Even those supporting the global ban acknowledge that he hasn't been violating policies. They then turn this into an argument against him: he's trying to "worm his way into the confidence of the community" by making positive contributions. I think this is the most beautiful ban argument ever! "User makes positive contributions with Bad Intent." Look! He made positive contributions in the past, such as at Wikisource, and Wikisource banned him! (Uh, why? That remains entirely unclear. I think it was because 'crats and checkusers at Wikisource ended up with egg on their faces because they allowed him to edit as a fresh start, and they allowed him to seek admin status without disclosing his identity. In other words, they were embarrassed that they'd been foolish, and they blamed it on him. So they took it out on him. This could not happen at Wikiversity, because he is openly Poetlister there, openly connected with his past.)

So any discussion of whether or not Poetlister should be allowed to edit Wikiversity should be on Wikiversity, so that the affected community decides, not some other community. While it is true that, in theory, there could be exceptions, they have been rare and limited to emergency situations, and the local community can reverse emergency actions taken by stewards, and they have. And it's not a problem, it is not "unresolved," it's been clear all along.

In the discussions we see now at meta, meta is attempting to decide, at meta, that it can intervene and control the individual wikis. I'd say that wikis could rationally opt-in or opt-out of that. If it's imposed from the top, then it's simply an attempted power grab, self-justified. And who is going to enforce this concept? The local communities? Not if they are not voluntarily a part of it! They have been a part of something very, very different, for years. It would be possible to make this enforceable. Simply ban the approximately three-fourths of the users who support local independence, if they dare to protest, as being "disruptive" and "against global consensus." Easy!

http://meta.wikimedia.org/w/index.php?titl...1&oldid=3063243



Abd, is this addition satisfactory?

Apparently on Meta, the EYE-talian mafia doesn't want you to know all of the reasons why a global lock may be implemented.

You know Abd, in the end they'll be obliged to find a way of globally banning you, however much you attempt to try and remove their ability to do it. In my opinion you will completely deserve it. There are people in life who are so excessively both congenitally and wantonly illogical, so stiflingly overbearing, and so utterly dysfunctional, that they eventually become nothing other than counter-productive. It's not like they would be banning you from your local shops, or there is nothing else you can do with your life.

It can hardly be a 'power grab' by them unless you have any powers yourself: but you don't represent anything other than the ridiculous idea that people signed up to Planet Wikiversity are 'innocent until proven guilty', despite what they have unremittingly done elsewhere in an environment that is tailor made for surreptitious opinion-mongers, insidious perverts and sleazy bullshitters. WikiMedia simply deals in immorality - liars, bullies, controllers and thieves.

It's just a shame for you that your cybersex with Poetfister will have to end when you become no more use to him. Just think of all the money you'll save on paddles and brillo.

There's this whole self-fulfilling-prophecy take on that that I'm not going to bother going into, because I can't be bothered to try to talk sense into Abd.

Satisfactory for what?

Properly, a "global ban" is a decision to apply a global lock.

A subtle shift has occurred. In theory, if true and present consensus is to rule, for a ban to be maintained, there must be consensus *for it*. For efficiency, though, communities don't re-discuss everything, and "consensus communities" have a tendency to become rigid. The rigidity arises when a past consensus is considered to bind the present, requiring consensus (i.e., supermajority) to overturn it.

The real problem is the lack of a reliable decision-making process.

Practically speaking, discussions at meta will always, at least for a long time, be dominated by Wikipedians (from the various languages, but especially en.wiki).

This is the real problem, and the idea of overturning global bans assumes that it's necessary to do so, in order for a user to edit a local wiki. Global locks protect the small wikis from encountering users considered disruptive elsewhere. However, the wikis offer differing conditions. Wikipedia is a floating opportunity for dispute, with this rather vague concept of "neutrality" as if neutrality were an absolute, when, in fact, the only reasonably reliable way of assessing neutrality is through a measure of true consensus. Which is largely prevented at Wikipedia, "POV pushers" being readily banned. Get rid of the people with strong opinions, does the expertise, the ability to understand issues, in the remaining population increase or decrease?

Wikiversity, on the other hand, by the nature of the project, does not need to make decisions about neutrality as commonly, because forking is allowed. (Spooning, too, off in the corner there. Knifing seems to be de rigeur lately, but none of this has to do with content. Rather, fulfilling a major function of the wikis, and, yes, it's true: it's about Ottava and Poetlister.)

So it's entirely possible that a user may be unable to cooperate on many WMF wikis, but might be able to work collaboratively on one. Why should the one have to overturn a global ban in order to allow the user to edit there? The idea that this is necessary would be entirely new. Up to now, local wikis had autonomy. The edges of that were frayed with Thekohser's ban, but, in fact, the wikis ended up being quite clear. Nolo me tangere. Mind Your Own Business.

The global ban for Thekohser stands, it's still in effect. It's highly unlikely that it will be overturned. Yet, Thekohser *is* allowed to edit on certain wikis. Where is he disruptive? Not there!

Does anyone notice that bans create ongoing disruption?

This much is true these days... on the wikis where my global lock has been overturned, I've presented zero disruption. On the wikis where the global lock is enforced, I do seem to cause disruption, at least how they define it.

Right, but you were "banned" because Jimbo said so, and he said so because he doesn't like you. I don't think it's necessarily a bad sign that teh communiteh decided to ban the poetlister.

Summary: SB Johnny is a far more serious danger to the independence of Wikiversity than anyone else.

It's a distinction without a difference. "The community" is the community aware of and active on meta. This is not the Wikiversity community, nor is it the global community.

Meta is utterly unsafe, except for Ottava, who has a Special Immunity Badge. Same at Wikiversity. SBJ has protected him because he's a convenient attack dog.

Of course SBJ approves of the Poetlister ban. But Poetlister wasn't disruptive at Wikiversity. SBJ is, on Wikiversity, pretending to be unclear about the community consensus. That's because with Ottava and others around, it's possible to gather quick comment that rejects the long-established policy that non-disruptive users are not to be blocked without a community consensus, and those who are obviously not present on Wikiversity to build the place, but to either tear it down (which could even include SBJ) or to make sure that personal enemies are not allowed to function there, can present an appearance that SBJ and others then use to claim that my unblock of Poetlister wasn't "consensus." It wasn't consensus. It was an action within custodial discretion, and clearly so.

By standard practice, an unblock doesn't require consensus. SBJ even noted this himself when unblocking Thekohser. Basically, SBJ uses whatever arguments he can find to rationalize his actions. And the claque falls for it.

There was no emergency with my sysop bit, as claimed by SBJ. I'd made two button pushes considered inappropriate. Both were quickly reversed. The first, and only really important one, was the unblock of Poetlister. I did not unblock immediately, I waited almost five days, making sure that there was some agreement with the Wikiversity blocking policy. The relevant sections of that "proposed policy" have stood for years without challenge. That's [also] an expression of consensus.

Block log.SBJ pretends that there was no community consensus on meta. What's so is that there was no RfC. There was an operating consensus. What SBJ is doing is ratifying meta consensus when he agrees with it, and not when he doesn't. The prior position was clear: Wikiversity manages its own wiki, including deciding who may edit. Meta may make a global default, by locking, but local wikis may decide about local accounts. Attempts to shut down Wikiversity because of "defiance" there went nowhere. The real danger to Wikiversity comes from within, as we will see. More accurately, it comes from a certain faction in the community, largely technocratic, mostly non-academic.
I mentioned discussion. Specifically, besides prior email discussion with my mentor, Jtneill (the other active bureaucrat on Wikiversity), and with Thenub314, there was
On Poetlister/Poetlister1
Request custodian action
User talk:Bilby
User talk:Poetlister1
User talk:SB Johnny
Colloquium discussion of Poetlister1 Because of controversy, I appended a notice that I'd unblocked to this discussion.
On blocking policy/bans
Permanent custodial discussion for Thenub314 (also mentions Poetlister) Thenub, in this candidacy statement, clearly said that he'd block to enforce a global ban, if the ban followed global ban policy. That policy doesn't yet exist. And who decides "global ban policy"? The WMF has the right, but it doesn't want to touch this. However, it might be convenient for some if meta becomes a global government. The practical effect, though, will be to make the rest of the WMF wikis helpless appendages to Wikipedia, for Wikipedia users can easily dominate at meta. A large non-English 'pedia may also do this; it's an issue of mobilizable user base.
Blocking policy has not changed significantly.
Community Review on global bans as of my unblock action.

Wiki decisions, in theory, are not made by voting. Rather, arguments are to be considered. Comments may be deprecated by a closer based on the nature of participants; a participant who shows up just to comment may not carry the same weight as one who is a regular contributor to the wiki. So, given established policy and practice (confirmed by SBJ in the past), given a discussion showing significant weight toward confirming the policy (most notably see the opinion of Jtneill in the Global Ban Community Review), I acted within custodial discretion, and because of the extensive discussion, and because of policy on unblocking, I was well within WV policies and practices.

The one who wheel-warred was Thenub314. He did not discuss the unblock with me, nor with the community, before reversing it. He unblocked, rather, on a procedural argument that is widely understood as bankrupt. (I.e., improper custodian action. It's mostly done in the other direction, i.e., admin blocks, admin was involved, should not have blocked, therefore unblock. It's an error because the block reason, hazard to the wiki being asserted, should always be considered.) SBJ did this in the matter of Ottava Rima's block.

"No discussion"? Is he crazy? (No, he's obtuse and argues from conclusions.) If anything there was too much discussion, not "no discussion."

With unblocks, an unblock template is routinely reviewed by another custodian, who may decide, especially after discussion, but may also decide immediately if policy is clear. The policy suggests an uninvolved custodian if possible. That's why I first went to Request custodian action and RCA. However, policy was clear, no policy-based arguments had appeared in almost five days, hence I saw my responsibility as being to unblock.

And what this revealed was some serious corruption, specifically ownership of the wiki by SBJ, and his willingness to lie to keep control. There are some who considered the action foolish, but that assumes that I'd care about the consequences of confronting SBJ.

That I undid an action of his, no matter what the arguments, was, for him, an emergency, and so he went to meta. When a relatively straightforward request for emergency desysop did not work, he lied. And Thenub314 supported this.

There was no violation of my custodial agreement. There was a dispute over the agreement page, and I'd deleted it, for technical reasons (page in user space, sole author, not wanted in user space, revert warring on move back to Wikiversity space) -- and revised the agreement, but that was merely a proposal. I undid this rapidly. I never violated the agreement. There was no risk; had I violated the agreement, SBJ -- or anyone, really -- could have requested a steward desysop immediately.

The meta decision, given the BS they were eventually fed, was proper. Jtneill, very conservative, is not willing to re-sysop, preferring to present the matter for vote, which is a reasonable decision in itself, though it will probably result in "failure." I prefer not to call in the reserves on a matter where I'm involved. Guess what? I won't be involved next time. (If you want to understand this, read the Talk page for the voting page, together with the original voting. Analysis reveals much about what is going on.)

I've always said I didn't need the tools to do my work on Wikiversity. Indeed, they were a distraction and a restriction. I spent months handling spammers and vandals, welcoming users, and generally supporting the community, highly active and highly useful, recognized by SBJ in the first voting on my permanent custodianship. I'm now free from that obligation. Feels good, frankly.

The meta request. Notice that SBJ does not mention the "rash actions" at first. He doesn't want to call attention to his personal involvement. Notice that he leads with the BS issue of "altering the terms of his agreement." What I proposed would have made me like any other probationary custodian. However, when there was objection, I reviewed the original discussion and concluded that some, at least, had depended on the agreement for their support, so I undid those changes while I still had the tools. That was transient, and involved no emergency. Any custodian could simply have gone ahead and enforced the agreement, as it existed when made (the original agreement referred to a permanent version. In moving this to my user space, SBJ broke the links, that was part of my objection). SBJ took advantage of that opportunity, fully. It allowed him to actually ignore the agreement, while blaming me. But the substance, clearly, was the Poetlister unblock, and the principle, being asserted by Darklama -- with Ottava Rima -- of unblocking allegedly requiring consensus, which is a principle which removes the protections of policy and allows a transient majority to dominate.

SBJ is clearly on the side of the conditions that have largely paralyzed the wikis, not allowing them to adapt to changing circumstances. These ways of operating favor the continuation of power of people like SBJ. It's ironic that one of the claims made about me is that I supposedly was rigid and didn't respond to criticism, but, in fact, the record shows the reverse, and those making the claims of rigidity are the most rigid. Not a surprise, eh?

There is discussion of this, to some extent, in my current Permanent Custodian voting, currently in the site message on Wikiversity. I just love one of the arguments: candidate is not qualified because candidate has been rejected, three strikes and you're out. Besides being circular, this prejudges the outcome. This is the third probationary custodianship, which was made "permanent" by a compromise in the second voting. (There was no PC vote in the second period.) It is the third voting. SBJ now claims (in the meta discussion, misrepresenting the history) that the compromise was a mistake. Nice that he admits his errors, eh? Mistake because?

In that vote, SBJ would not have been able to legitimately close. Jtneill would have closed. Jtneill is conservative, for sure, but the pattern shown on the Talk page is unmistakeable. I don't know what he would have done. The compromise avoided that mess. A desysop in midstream misrepresents the situation to the community, some of which will assume that if this was done, must have been a good reason, which is a misunderstanding of how stewards work.

It's just occurred to me that nobody represents Wikimedia better than Abd. In a peverse way, he is the ultimate Wikimedia mascot - representing the very warped stupidity that the 'projects' ultimately bring about. Wikimedia is the death of intelligence, and Abd is the angel on the tree. He is 'POV' and Verify, and 'bold' incarnate - and absolutely encapsulates the worst elements of each. He is the reams of eye-crossing archived text. He is attrition itself. You cannot argue with him just as you cannot ultimately argue with the mindless chasm of WP. There is no actual ground with either, only the abyss. Human life is absolutely valueless to Wikimedia, and it so it will ultimately spurn everyone involved in it. The unremitting and imbecilic force of Abd, while he lasts, represents it like nothing else.

Wait, you don't think he's an eternal fixture?

It's not for nothing that my name is in the Qur'an.

It may be that Wikipedia brings out the absolute worst in me. In real life, I can speak to people and they listen and are moved. Wikipedia depends on text, which filters out maybe 99.9% of human communication.

"Eye-crossing archived text." I don't know if anyone has noticed, but Wikipedia typography, on talk pages, is not designed for readability. The medium encourages brief. Very brief. That rules out any deep exploration of a topic. In real life conversations, there is constant interchange between speaker and listener, even if only one is speaking. A "boring speaker" is one who is unaware of the listener, the equivalent of a "wall of text."