Wikipedia talk:Arbitration Committee/Noticeboard

Policy on voting

How far back can arbiters change their vote? Do they need approval from the committee to make such a change? Where in the rules is the right to change ones vote after the close ensconced? Thanks! Hipocrite (talk) 10:44, 14 May 2012 (UTC)

If you're referring to the above incident, there was no change of vote. Courcelles voted on the remedy along with a comment which showed his intent, and his understanding of the issue, but had forgotten to sign off the finding. The guideline for the Clerks is "If there are any ambiguities concerning which proposals have been adopted, the Clerk should identify them so the Arbitrators have an opportunity to clarify them before the decision is finalized and announced". Slips happen, and now that the matter has been identified appropriate action has been taken. SilkTork ^{✔Tea time} 11:13, 14 May 2012 (UTC)

Right, so how far back can arbiters go back and realize they voted wrongly? I have a very specific vote in mind where an arbiter very specifically complained about how they would have voted differently in a circumstance, so I want to know if that complaint was before or after the deadline. Hipocrite (talk) 12:02, 14 May 2012 (UTC)

I don't know which case you have in mind but there's a very big difference between changing your mind and promptly correcting an error or omission. There isn't arbitration-specific policy on this but the ""Not" policy says "A procedural error made in a proposal or request is not grounds for rejecting that proposal or request".  Roger Davies ^talk 12:23, 14 May 2012 (UTC)

How does the committee determine that something is "changing your mind and promptly correcting an error or omission," vs "changing your vote?" What is there to protect users from someone going back 3 years to change their vote on something? Can ex-arbs change their votes in closed cases? How about disgraced ex-arbs? Hipocrite (talk) 17:43, 14 May 2012 (UTC)

It's OK to change one's mind about something - indeed, it can be a sign of reflection. I am quite comfortable with the notion that an arbiter makes a vote for an action and then changes it during a case. Indeed, I did it myself in the just closed Race & Intelligence case when more information was presented. When the case is closed then a vote itself cannot be changed, but the arbiter may later change their mind about making the vote, and is entitled to say so. For reasons of stability of the project, Committee decisions are binding, but that does not mean they are always right, nor that everyone in the Committee agrees with them, nor that a Committee member will not have regrets about voting for a particular remedy. A request for amendment can be made if a remedy appears to be inappropriate, though there is an expectation that - depending on the amendment request - a reasonable time is given to allow the effects of the remedy to be felt. SilkTork ^{✔Tea time} 14:38, 14 May 2012 (UTC)

It appears to me that one can change their mind on something right up until the point a decision is made, and then you can't. Perhaps you disagree - as in this case, where you obviously disagree, in that you believe someone can change their vote after a decision was made, undoing the decision and reforming it in a different way. How long before a decision requires a "request for amendment," as opposed to just canvassing arbs that you think might change their vote after decision are made? Hipocrite (talk) 17:43, 14 May 2012 (UTC)

I'm not an arb, but I would compare this to the practice of judges correcting an opinion after issuing it but prior to publication with issuing an order amending the prior opinion. Either way works, but the correcting prior to publication works only so long as it hasn't been published. Translating that to the wiki-world, I would say that after the point the decision has been archived from WP:AC/N, it should require an amendment to change as it has been "published to the archive," while the notice page itself is a bit like a slip opinion. MBisanz ^talk 17:53, 14 May 2012 (UTC)

It's OK for an arbiter to change their view, opinion, allegiance or religion, and they can express regret for decisions they have made, but they can't unilaterally go back to a closed case and change a vote. Outcomes of a closed case can only be changed by amendment when there is an open and accountable discussion voted on by the whole Committee. If there is an instance of an arbiter going back to a 3 year old closed case and changing their vote that would need to be pointed out as that would be an error. SilkTork ^{✔Tea time} 23:45, 14 May 2012 (UTC)

• It might be a good idea for the Committee to update its Rules & Procedures to clearly explain when an Arb may return to cast additional votes after a case has been closed. MastCell ^Talk 21:56, 17 May 2012 (UTC)

• • I think when a case has been closed with "ambiguities concerning which proposals have been adopted", or a similar administrative error, then it would be appropriate to notify the Committee as soon as possible so the matter can be dealt with. Ideally a case should be checked thoroughly before closing, but mistakes do occur - such is life. I think that common sense tidying up of clear errors need not be inscribed, though the tidying up should be discussed and noted so everything is kept in the open. SilkTork ^{✔Tea time} 17:42, 19 May 2012 (UTC)

Wikipedia:Arbitration/Requests/Case/Rich Farmbrough closed

Original announcement

This has been a very bad decision. Undue weight was given to Brad being a bit stubborn, uncommunicative, and a bit uncooperative, and short shrift was given to the opinions of the many people who thanked him, gave awards to him, feel themselves and the project indebted to him, and the quite serious disruption to important projects and harm to the entire project that results from this action. Several of the "charges" against Rich seem to be very petty indeed. The words used to support "abusiveness" on his part were the kind of thing more mature Wikipedians would laugh off and take on the chin for the sake of the project. How many of us have suffered much clearer cases of "abuse" but didn't feel the need to push a formal process to have the person banned, effect on the project be damned, just because you have been insulted? And this man was very useful! This was no ordinary editor. His bots do the work of thousands of manhours in very short time and move the whole increasingly unweildy project forward by keeping an eye on thousands of things, catching serious errors that had stood for a very long time, making a mockery of the project. For more details, check out his user and talk pages, look at the praise for and accompishments of this good Wikipedian. For the other "offensives", they these ten administrators point to nothing that can be shown to have harmed the project, it was all depicted as a sort of defense of principles, only, and a look at the long list of accomplishments, praise, and Wikilove was not given due weight! There was even a kerfuffle about him changing tiny bits of whitespace, my lord, who cares? Let it go, for crissake, or better said for the sake of the project. Please run the process again, this time taking into consideration a thourough investigation into his accomplishments and the serious disruption to the project that results from this rash, uninformed, disruptive, and punitive decision. Chrisrus (talk) 20:19, 16 May 2012 (UTC)

I think you meant Rich above where you said Brad. Kumioko (talk) 23:31, 16 May 2012 (UTC)

"Let it go, for crissake, or better said for the sake of the project." That would have been excellent advice to have given to Rich when he used his administrative privileges to do things like adjusting whitespace or capitalization against consensus and/or prior community restrictions. Wikipedia has a high tolerance for individual idiosyncrasies, which tends to evaporate when people try and use administrative and bot operator privileges to enforce theirs on the entire project. It is unfortunate that prior dispute resolution failed to communicate this message to Rich. Jclemens (talk) 01:37, 17 May 2012 (UTC)

Well, maybe the benefits from his bots and abilities are important to consider when banning such important bots as helpful pixie bot and femto bot and others that we all rely on when deciding what to do about such things as whitespace adjusting. Let the committee consider both when coming to their decision this time. This is not just a case about unauthorized whitespace adjusting, it's also about lots of very important and good things that are done as well. It was not right to make such a decision without weighting the pros and cons properly. Chrisrus (talk) 05:07, 17 May 2012 (UTC)

This is a dangerous notion, that the transgressions of an editor can be mitigated or even absolved purely on the basis that they have a certain weight of good contributions. This mindset would allow the doctor that saves a hundred lives to murder his wife, or the soldier who protects his country by serving overseas to steal from people's homes. It sets double standards for editors that are entirely unreasonable - Rich is allowed to break the rules, but someone else isn't, and the decision isn't in any way related to what they actually did wrong, but how many brownie points they've earned? That's an untenable position for one to hold if it is one's goal to maintain the fairest possible environment. There should be no special treatment. It shouldn't matter if Jimbo himself broke the rules. If the rules were broken, an appropriate and consistent response should be given.

As for Rich's bots, nothing his bots do can't be done by bots written by someone else. He isn't indispensable - nobody is indispensable - and that's a good thing. Others will step in to fill the void. I looked at the BAG page last week and already saw a proposal or two for bot tasks to replace HPB. – NULL ‹talk›
‹edits› 06:20, 17 May 2012 (UTC)

We're not here to have the fairest possible environment. We're here to build an encyclopedia. I don't have enough understanding of the issue one way or the other to weigh up the damage he was causing over the good he was doing, but my instinct is that the former was outweighed by the latter. I am not confident in the decision anyway. Egg Centric 22:10, 17 May 2012 (UTC)

We're here for everyone to build an encyclopedia. That doesn't happen when the environment in which you're doing so isn't conducive to teamwork. Fairness is an essential component of meaningful collaboration, because its absence breeds contempt and an 'every man for himself' mentality. This is basic leadership knowledge, really. – NULL ‹talk›
‹edits› 23:22, 17 May 2012 (UTC)

(e/c_ I thought the idea was leave the damn whitespace and similar non-rendering cosmetic issues alone. And those bots did a mix of useful and useless stuff, and of the useful stuff, nothing jumped out at me as being especially important that couldn't be done by someone else without much fuss (and anything important and hard to replicate shouldn't have been approved without published source code). I commented for a while in the workshop (then disengaged because of RL stuff etc.) and supported somewhat different remedies, but I think we will get along ok without those bots. More importantly this decision makes me feel like arbcom's eyes are starting to open about automation misuse, compared to some earlier decisions that didn't do enough. Wikipedia is not a programmers' playground.

A quibble: I don't think the reminder to Elen should have equated Rich with his bots. Blocking a bot that's causing problems shouldn't be considered a big deal in general, unless there's wheel warring or similar involved. I can accept that in this particular situation it would have been better for someone else to do the block. So the remedy's rationale was mostly ok but I think it should have been framed as "Elen advised" rather than "reminded" (since reminding implies there is a pre-existing notion that someone didn't remember). 66.127.55.46 (talk) 06:30, 17 May 2012 (UTC)

I take your point - Elen didn't forget, so much as think of the bot purely as a chunk of malfunctioning code, rather than an extension of Rich (who I wouldn't have actually blocked, regarding myself as WP:INVOLVED). It was a mistake in that it's obvious when you think of it that the bot could be seen as Rich-by-extension, so it was always going to be contentious. I certainly shan't do it again, should I find myself in that situation. Elen of the Roads (talk) 22:02, 17 May 2012 (UTC)

Understandable, but that comment cuts both ways. The main reason (from my reading) that Rich was desysopped was because he unblocked his own bots – who's to say he wasn't working under the same misapprehension (i.e. that bots are not users-by-extension, just bits of code)? Jenks24 (talk) 22:29, 17 May 2012 (UTC)

Now again, I thought the problem was more editing through protection, and various things to do with deletion/recreation, other than one?? cited instance where he unblocked the bot without taking any notice of the problem...? Again, don't want to impinge on what was in the decision makers heads. --Elen of the Roads (talk) 22:36, 17 May 2012 (UTC)

I can't see editing through protection mentioned at all in the final decision (and rightly so, see workshop discussion). I also see no mention of deletion/recreation. It's interesting that you say there was only one instance where unblocking his bot was improper, as that was also my understanding – the FoF states that he did it "on many occasions", though fails to cite specific instances. Makes the case for desysopping him look rather flimsy. Jenks24 (talk) 23:15, 17 May 2012 (UTC)

Prohibited

I don't see "prohibited" on Wikipedia very often, so I'd like a clarification. Is this just another way of saying "banned", or is there some significance to this usage? Nyttend (talk) 05:59, 17 May 2012 (UTC)

Prohibited is the same as banned. SilkTork ^{✔Tea time} 21:52, 17 May 2012 (UTC)

Have a bad feeling about this...

I hope that I'm proven wrong, but I don't think the automation restriction will work well. I think the suggestions on the talk page of the proposed decision (by CBM, IIRC) would have been better for all involved. Anyway, I have the first of what will probably be many questions about this automation restriction: I notice Rich has (without requesting it) been given the rollbacker userright; would using that be considered a breach? What about Twinkle? Huggle? Semi-automated scripts? Jenks24 (talk) 20:40, 17 May 2012 (UTC)

Agreed it's not a perfect solution, but it is the one we cobbled put together in an attempt to deal with the situation. The essence is that the community had expressed concern about Rich's use of automation and trusted tools, and those concerns had not been adequately addressed by Rich; as such the Committee have taken away Rich's access to the admin tools and prohibited him from using any other form of automation. This, by implication, would include the Rollbacker, Twinkle, Huggle and AWB tools. Rich is advised to edit manually and in a manner that could not be mistaken for automated. If in doubt if he is permitted to use an aid he would be advised to avoid using it, and if he feels he needs or would like to use it, to seek clarification if he can use it. SilkTork ^{✔Tea time} 22:03, 17 May 2012 (UTC)

If you cobbled it together does this suggest this is temporary, and you are going to come up with a better solution given time? Egg Centric 22:12, 17 May 2012 (UTC)

My bad. I didn't mean to suggest hasty or temporary - I had misunderstood the phrase "cobbled together" to mean put together out of the available material rather than the "ideal" material. I think the solution is understandable and workable and not too complicated. That doesn't mean that it is faultless, or there won't be an attempt to subvert it, but it is what we have, and will work with the right attitude. SilkTork ^{✔Tea time} 10:02, 18 May 2012 (UTC)

Gotcha, cheers Egg Centric 12:23, 18 May 2012 (UTC)

(edit conflict) Well, I'm disappointed with your answer, but at least it's clear. Thanks.

On a side note, if the decision was just "cobbled together", what was the rush? I know each year incoming arbs pledge to get through cases more quickly, but surely if the solution was imperfect, a few more days could have been spent discussing and trying to improve it? Jenks24 (talk) 22:15, 17 May 2012 (UTC)

rollback a form of automation? suddenly I've got a bad feeling about this... Rd232 ^talk 22:19, 17 May 2012 (UTC)

I believe SilkTork meant cobbled as in crudely constructed instead of quickly constructed. Arbcom has never found a good definition of automation to deal with editors who do quick/mass edits that are controversial. See Betacommand and Lightbot for two multi-year examples of trying different restrictions that ultimately failed. MBisanz ^talk 22:25, 17 May 2012 (UTC)

Understood, but surely the point still remains that if it is crudely constructed, wouldn't it be in everyone's best interests (including ArbCom's) if some time had been spent refining it? Jenks24 (talk) 22:31, 17 May 2012 (UTC)

You may find Wikipedia:Administrators' noticeboard/Δ instructive. That set of pages documents a small part of the efforts of dozens of editors over 3 years to refine the definition of automated editing as applied to that user. They failed, as did Arbcom in at least three cases on that user. Time won't refine this matter, only rather crude and blunt restrictions. MBisanz ^talk 22:36, 17 May 2012 (UTC)

I don't wish to contradict my colleague here, and of course I wasn't in on the decision, but I thought the discussion had concluded that because Twinkle and Huggle can't be 'hacked' but have to be used out of the box, and Twinkle certainly forces the user to confirm the edit, they would be OK. Elen of the Roads (talk) 22:22, 17 May 2012 (UTC)

The discussion on the talkpage was that Twinkle, Huggle and Rollback would be included in the restriction. The alternative proposal (that was being drawn up but wasn't implemented) did specify their inclusion: "No automation on the main account. This includes semi-auto editing, AWB, Twinkle, custom javascript, everything." . I'm seeing "any automation whatsoever" along with "any edits that reasonably appear to be automated shall be assumed to be so" as a bright line that is easy to monitor. SilkTork ^{✔Tea time} 10:28, 18 May 2012 (UTC)

Since WP lets anyone nobody Me! use Twinkle it should be fine per common sense. Pushed to its illogical extreme "no automation" would preclude using brower spelling auto-correct. Nobody Ent 22:46, 17 May 2012 (UTC)

Automated spell correctors can make mistakes so they need to be checked. I have a spell checker which underlines what it thinks are spelling mistakes, but allows me to make the decision. We have different spellings for British and American articles, and sometimes we use quotes that may have archaic or incorrect spelling. So, yes, it would make sense for Rich to turn off any automated spelling changes and allow his browser to instead indicate what it thinks are spelling errors. SilkTork ^{✔Tea time} 10:28, 18 May 2012 (UTC)

The issue of rollback, specifically, may be something Arbcom wants to nail down, because among the non-admin rights restored to him after the close of the case is - you guessed it! - rollback. If it is the case that arbcom feels that rollback is subsumed under "automation Rich is banned from using", it may be better for an arb or clerk to remove that right, rather than leave him with it and see drama happen when someday he accidentally or absentmindedly clicks the button. A fluffernutter is a sandwich! (talk) 14:16, 18 May 2012 (UTC)

I agree with SilkTork, for the purposes of restrictions like this one, rollback is automation, and, as such is prohibited from those restricted from using all automation. If this becomes contentious, though, it would be better to discuss this as a formal clarification request than here... Courcelles 14:28, 18 May 2012 (UTC)

 Done --Guerillero | My Talk 16:41, 18 May 2012 (UTC)

That's a very liberal interpretation of the meaning of the word "automation". Malleus Fatuorum 11:53, 19 May 2012 (UTC)

And that was a very polite way of putting it, Malleus, thanks. I happen to agree: automation is about affecting multiple articles in rapid succession, and it's not really possible to do that with rollback absent MEATBOT like behaviour, or some other automation scripted to invoke rollback in such a manner. While I appreciate Guerillero's promptness, I fear that the discussion had not reached a point where his action was yet needed. Jclemens (talk) 16:14, 19 May 2012 (UTC)

I'm afraid it also looks like indecent haste to further humiliate a fellow editor who's become unpopular in certain quarters. As I said further down on this page, one of the most disturbing aspects of ArbCom decisions is that they're entirely predictable, based as the evident preconceptions of those voting. Malleus Fatuorum 16:48, 19 May 2012 (UTC)

If anything, I think that what this illustrates (and I'm speaking to the below conversation as well) is that requests for clarification are best handled on Arbcom's "Requests for clarification" page. We have a discussion here and a premature request for amendment, when in fact all that is needed is a formal requests for clarification. It was not my impression that "votes" were conducted or binding interpretations were to be made on this page, when we have others designated for that purpose, but if my colleagues disagree, we may need to alter the structure of this page to facilitate that, which would end its usefulness as free-form discussion. Jclemens (talk) 16:59, 19 May 2012 (UTC)

Ah, apologies to all involved. I'm still pretty unfamiliar with the whole ArbCom process (this is the first case that I have followed at all), so I didn't know there was an official process for requesting clarification. I just saw that Rich had been given a userright which he may might unconsciously use and thought it best to get some quick clarification before someone claimed he was violating the restriction. Should I (or someone who knows what they are doing) open a formal request? Jenks24 (talk) 17:08, 19 May 2012 (UTC)

Obviously I have been an opponent of these restrictions from the start so what I am about to say is goign to come as no great surprise but here goes. Contrary to what is said above and in the Arbcom ruling it wasn't the whole community that had a problem with Rich's edits. It was certain members and somem powerful admins and arbs. There were a lot of us that think that the Arbcom decision is extreme, was poorly constructed, poorly written and will be even more poorly implemented. The purpose is and always was to stop Rich from editing Wikipedia. They tried to ban him but the vote didn't go their way so they did the next best thing and worded the ruling in such a vague way that nearly anyone with half a brain could argue that almost anything he does is a violation of it. This ruling is a recipe for disaster not only for Rich but for other editors in the future and for Wikipedia in general. They blocked his bots and revoked his ability to use automation. Fine. This ruling should NOT also restrict his ability of using tools that are available in the GAdgets menu of preferences or in at least some of the available rolls such as File mover, Rollbacker or the like. I agree I may be alone in this but thats how I feel. At this point I completely question even having a need for an Arbitration committee in Wikipedia and if someone where to ask me I would say we should shut the group down and write it off as a failure. Kumioko (talk) 15:24, 18 May 2012 (UTC)

I'm an ordinary editor. I could not disagree more with the above summary. Arbcom are elected and did what they were elected to do. If you don't like it do something at the next Arbcom elections. Stop soap boxing. Leaky Caldron 15:40, 18 May 2012 (UTC)

I intend too but I also should be clear that I think that we need to do a better job of screening next time. It is becoming clearer that selecting the most popular Wikipedian to be an Arb is not the best path to follow for the future. Kumioko (talk) 16:02, 18 May 2012 (UTC)

I agree. It's become very obvious with this case and others that there ought to be no place for ArcCom as it's currently constituted. This case is by no means the first time we've seen these post hoc amendments to the terms of some sanction or other, but it needs to be the last. Malleus Fatuorum 11:53, 19 May 2012 (UTC)

• hm, maybe using browser or computer also counts as automation Bulwersator (talk) 19:48, 18 May 2012 (UTC)

@SilkTork: "So, yes, it would make sense for Rich to turn off any automated spelling changes and allow his browser to instead indicate what it thinks are spelling errors." - So whether or not he does that, all that is needed is waiting for Rich to make a spelling mistake that could have been an automatic update from another common spelling mistake, and that shows enough reasonable appearance of automation. --Dirk Beetstra ^{T C} 17:15, 19 May 2012 (UTC)

Follow the bouncing ArbCom decision

So, not only do we have the follow the bouncing arb count on Rich's case, now we have a follow the bouncing ArbCom decision. Above, we have two arbitrators...out of ten (or was that 9? or 11?) deciding that the rollbacker right is "automation", and Rich's rollbacker rights are stripped. Neither arb member indicated they were speaking on behalf of the entire arbcom that sat the case, just unilaterally posting their opinions. Nevertheless, the case is retroactively modified to strip his rollbacker rights. I wonder, how much of this arbcom decision is going to be retroactively changed? What traps are awaiting Rich to stumble into unawares that he is doing something wrong because there's no actual decision saying it's wrong? What if he'd used rollbacker already? Since the case was retroactively changed (though, of course this "decision" is not posted on the case...very professional), would Rich now be banned from the site because he had used rollback? Thankfully, he didn't. --Hammersoft (talk) 17:36, 18 May 2012 (UTC)

Decisions such as this one are clearly designed to ease the path for future blocks and bans, so it's likely that Rich's days are numbered. His only real hope is that that doesn't happen before the next ArbCom elections. Malleus Fatuorum 11:59, 19 May 2012 (UTC)

ArbCom elections are not going to change the system. --Dirk Beetstra ^{T C} 12:58, 19 May 2012 (UTC)

The ones last year and the year before certainly didn't, so I remain sceptical that this year will somehow be strikingly different. --Demiurge1000 (talk) 13:15, 19 May 2012 (UTC)

You may be right. But what's obvious is that arbitrators largely vote in accordance with their own preconceptions, often with very little regard to the actual evidence presented, or its quality. So different people could at least change that. Malleus Fatuorum 13:47, 19 May 2012 (UTC)

The damage is done, and it will perpetuate at least until the next ArbCom elections (unless the community starts disallowing more damage). --Dirk Beetstra ^{T C} 14:21, 19 May 2012 (UTC)

It seems that this case, as with several others of the last few months have shown us several things. These include but are not limited too: That not everyone in the community believe that Arbcoms decisions are above reproach, that the Arbcom process needs some significant redesign and rethinking and that we need a process by which even an Arbcom decision can be overturned. Every society needs a system of checks and balances, Wikisociety is no exception. Kumioko (talk) 14:45, 19 May 2012 (UTC)

Hammersoft, the revocation of Rich's rollback rights on the basis of only two arbitrators' say-so was a mistake which I am trying to have corrected. AGK [•] 18:57, 19 May 2012 (UTC)