Case clerk: AlexandrDmitri ( Talk) Drafting arbitrators: Kirill Lokshin ( Talk) & Elen of the Roads ( Talk)
Wikipedia Arbitration |
---|
|
Track related changes |
Behaviour on this page: Arbitration case pages exist to assist the Arbitration Committee in arriving at a fair, well-informed decision. You are required to act with appropriate decorum during this case. While grievances must often be aired during a case, you are expected to air them without being rude or hostile, and to respond calmly to allegations against you. Accusations of misbehaviour posted in this case must be proven with clear evidence (and otherwise not made at all). Editors who conduct themselves inappropriately during a case may be sanctioned by an arbitrator, clerk, or functionary, without further warning, by being banned from further participation in the case, or being blocked altogether. Personal attacks against other users, including arbitrators or the clerks, will be met with sanctions. Behavior during a case may also be considered by the committee in arriving at a final decision.
Active:
Inactive:
Recused
Excuse me for being blunt, but the proposed decision date has already passed, and there hasn't been any movement at all on the proposed decision page. Now I know that most of your discussion takes place where no one can see it, but I worry that this case might not get closed out by the time that the ArbCom changeover takes place, and that this is percisely the kind of mess that shouldn't be handled over the course of two terms.
Finally, while I'm not accusing anyone of doing this, if action on this is being delayed because it might affect the ongoing elections, I'd like to say that such behavior would be extremely inappropriate. Sven Manguard Wha? 04:07, 3 December 2011 (UTC)
According to "Majority reference" table: 1–2 (Abstentions) result in 7 (Support votes needed for majority) Bulwersator ( talk) 17:34, 13 December 2011 (UTC)
I think there's a bit of cognitive dissonance in holding those positions simultaneously. I think it's fine to say that the sanctions were grounded in real concerns, but on the other hand the whole pattern identification business turned out impractical when participants in the drama couldn't agree that a script-generated sequence of edits was a pattern. The restrictions while well-meaning didn't work as intended. So, it makes little sense for ArbCom to confirm or "take[s] possession" of them as-is. ASCIIn2Bme ( talk) 04:15, 4 January 2012 (UTC)
Picking up on the comment by Courcelles, one possible alternative remedy would be to ban Betacommand from everything except running automated tasks via a flagged bot account approved by the Bot Approvals Group. He would be permitted to edit from his main account only on talk pages or noticeboards and only to discuss his bot operations (with an exception for his or his bots' userspace). Any other edits would have to come through his bots, and must be approved tasks. This would allow, for example, User:Δbot to continue clerking the SPI venue (I've heard no complaints as to his operations there). – xeno talk 13:56, 4 January 2012 (UTC)
When we turn to the proposed decision and consider how to resolve the issue with Betacommand's approach to interaction, I rejected the one-year ban, because the problem is with Betacommand's interactions - which a one-year break from Wikipedia would not solve. An indefinite ban, with a return only upon the presentation of a satisfactory plan for contributing constructively, would be a solution in that it would ensure Betacommand could only return if he became capable of contributing without the generation of a plethora of heated discussion. However, if we could agree that Betacommand was only problematic inasmuch as he runs (and apparently messes up) unapproved cleanup-style runs (not vetted bots), then a sensible solution could be to ban him from making any edit that was not approved in advance by BAG. Of those remedies that would meaningfully resolve the "Betacommand problem", I am inclined to pass the least severe proposal; if we do so and the problem persists, a siteban could be instituted by motion with little difficulty (and a remedy resolving that we will siteban in the event of continued issues with Betacommand may need to be included). Apologies for the verbosity, but I wanted to set my views down clearly before we proceed. AGK [•] 15:16, 4 January 2012 (UTC)
I'd like to explore Xeno's suggestion, if Betacommand believes this would allow him to continue to be useful; I've started two proposals for article-space restrictions on the workshop. See Workshop#Article edit rate restriction. John Vandenberg ( chat) 22:01, 5 January 2012 (UTC)
This seems like a terrible idea. If someone has a long and intractable history of editing disruptively in some area (say articles about penguins), usually they get topic-banned so they have to stop editing about penguins but can still edit any other topic they like. The proposal here is the opposite, something like "so-and-so may henceforth only edit about penguins, subject to the endorsement of the penguin approval group". I don't have any problem with Δ editing like a normal editor (i.e. completely manually and on a normal human scale) but IMHO he should be completely removed for a while from bots and scripts of any sort (thus the 25 page/day limit proposal). To Silktork re "saving useful bots": the bots involved are mostly not that useful (I'd say they're worse than useless, but that's just me), and any useful ones can be reimplemented by others. We should not let "vendor lock-in" dictate our actions in any case. 67.122.210.96 ( talk) 05:30, 9 January 2012 (UTC)
Give how discussions seem to work at BAG [3] I don't know if this is actually going to solve anything with respect to drama. ASCIIn2Bme ( talk) 17:46, 11 January 2012 (UTC)
I am glad this comity is considering clean start as a possible remedy. I beg members to be objective in weighing this option. There are assertions that this form of remedy implies an external blame is at core. Please read the policy again and if I may be so bold as to ask, please also show the prose that supports the assertion. It is problematic that Betacommand is under current sanctions, because that would preclude availability of this remedy. The proposal to vacate sanctions could be adopted and the clean start immediately ensue. With well crafted word, and the prohibitive sanction against evading scrutiny, this remedy does no harm to the encyclopedia, nor place jeopardy upon it. With respectful intent, thank you for considering this case, and this request as well. - My76Strat ( talk) 03:16, 5 January 2012 (UTC)
@Jtrainor - What do you find difficult? Can I ease the difficulty? I do stand available to answer any direct question you may have. In general, I do oppose the comparison of Betacommand's conduct with criminal conduct. This because generally I see no comparison. And I am just naive enough to believe it is not necessary. We have a southern saying: "you don't kick a man when he is down" which I have come to understand, and agree. Perhaps we could see a bit less "kicking", or not. My76Strat ( talk) 16:56, 11 January 2012 (UTC)
I know I found them when reviewing evidence a month ago, but I am not entirely certain that the version of community sanctions I have found is the most current one, particularly given all the amendments and modifications done over time. Could someone please link me to the *current* iteration of Betacommand/Δ's community sanction? Risker ( talk) 04:50, 5 January 2012 (UTC)
There absolutely are ways forward instead of back. And Xeno's proposal above is one such. The attempt to follow the "community imposed sanctions" was another, which resulted in more problems than it resolved (perhaps suggesting that the "community sanctions" were a complete failure, in terms of drafting, implementation and observance).
Regardless, if the result of this case is a ban, or something that destroys the value (either net, or potential) that Δ can bring to the project, it will be a sad day.
Rich
Farmbrough, 16:34, 5 January 2012 (UTC).
Just adopt my proposal about mentoring proposed some time ago on the Workshop page. So, some editor in good standing who has the relevant experience with operating bots, will be in charge of setting restrictions on what Delta is allowed to do. This solves two problems. Not only are the problems that Delta has caused contained, but also the drama generated when Delta is alleged to have violated some restriction will end, because Delta's restrictions are now off topic to the community. This is now the sole task of the mentor. The community may raise some perceive problem about Delta to the mentor, but it's always the mentor who takes some decision w.r.t. restrictions. If Delta violates a restriction set by the mentor, then that's also an issue between Delta and the mentor only. The mentor does have the duty to report Delta back to ArbCom if Delta were to misbehave and/or not stick to the rules set by the mentor. ArbCom can then revoke the mentorship agreement, which automatically bans Delta from undertaking any automated tasks. Count Iblis ( talk) 23:14, 7 January 2012 (UTC)
Assuming the community restrictions are continued/imported as AE-remedies, who is going to be considered uninvolved for the purpose of AE discussion and decision-making? This is a serious question because these are not typical content disputes focused on a topic area, but are about systematic changes to many unrelated articles. Reading the proceeding here and also the previous AN[I] threads, one cannon escape the feeling that over time Δ has engendered a dedicated fan club who shows up with regularity at most discussions about him. And the positions of the participants in previous disputes involving Δ are a very good predictor of their position in subsequent ones. A rather obvious and prolonged factionalism has developed around the "Δ-problem". ASCIIn2Bme ( talk) 06:57, 6 January 2012 (UTC)
Also, the recently but inconclusively closed discussion Wikipedia:AE#Suggested_change_to_practice:_comments_by_non-neutral_editors should be of relevance here. ASCIIn2Bme ( talk) 07:02, 6 January 2012 (UTC)
I think there were some CU-confirmed socks besides the "official" list given in this Arbitration. See Category:Wikipedia sockpuppets of Betacommand. Amusingly, one of these was doing in 2009 the kind of "clean up" that was again in focus in the autumn of 2011: Special:Contributions/Shimane Kanagi. ASCIIn2Bme ( talk) 10:18, 8 January 2012 (UTC)
The antecedent is stated
here. I.E. the case was a request for clarification as to whether a proposed task fell under the aegis of a previous ARBCOM imposed restriction.
Rich
Farmbrough, 18:16, 11 January 2012 (UTC).
Actually I realise there is an antecedent, in that the complexity and finickiness of the restrictions meant that there was alot of squabbling and lots of people unhappy. Hopefully the new remedy sorts this out. Casliber ( talk · contribs) 11:16, 12 January 2012 (UTC)
Some months ago, I began developing a list of Δ's accomplishments on Wikipedia. This was in response to a rather vocal group of editors who despise Δ, and seem to think everything he has done is sort of a inverse Midas; 'anything he touches turns to crap'. Given the committee seems to be hung up on trying to decide whether Δ is a benefit to the project, it would be in their interest to consider this non-exhaustive list of his accomplishments here. See User:Hammersoft/Δ vitae. -- Hammersoft ( talk) 23:08, 11 January 2012 (UTC)
Why is it the business of Arbcom to hand out back-pattings and attaboys? Jtrainor ( talk) 23:34, 12 January 2012 (UTC)
Thanks Hammersoft; can you finish adding the dates, order it by date, and put it in /Evidence. Thanks. John Vandenberg ( chat) 10:00, 24 January 2012 (UTC)
I have followed this case as a completely uninvolved observer. I expected a choice between a ban and something like this (Bot Group approved tasks only) - I think they are choices that would address the issue rather than skirting it or nibbling at it, maybe the only choices that would do so. However, if Beta's to stay active, it is not enough to allow interaction with users - the quality of that interaction should be addressed as well. I'm not workshopping a proposal for you - I think it needs to be handled more skillfully and gently than I could - but there needs to be a remedy directed at the quality of Beta's interaction with other users in conjunction with the BAG-approved only remedy. Jd2718 ( talk) 11:50, 12 January 2012 (UTC)
I'm basically saying SPI bot is irrelevant to his ability to operate an article-editing bot. As to the last question, we can look at the history of arbitration cases, community sanctions, bans, and the evidence in this case, and everyone can come to their own conclusions.
My opinion, which I have previously expressed, is that if we don't want a full ban, the best option is to impose much firmer sanctions that do not rely on the ability of Δ to modulate his own editing. The BAG-only option seems like it has a possibility of achieving that, I think. It's parallel to the previous requirement that he has to post on VPR, but it no longer assumes that he will follow the rules of his own accord. And, by moving enforcement to AE, it makes the penalties more likely to stick if Δ violates the restrictions. — Carl ( CBM · talk) 16:50, 13 January 2012 (UTC)
Remedies:
I feel if we are to allow bots, that there should be a limit for at least a year to see how it works. Perhaps limit to three? The community sanctions would not apply to the bots, though the topic ban should still be in place. All forms of editing sanctions should come under ArbCom (which is what PR 1.5 is saying). SilkTork ✔Tea time 15:11, 12 January 2012 (UTC)
I've looked in the evidence page and don't see evidence for recent poor interactions. Would someone please supply such evidence, or we'll have to assume that Betacommand is responding within reasonable parameters. SilkTork ✔Tea time 15:30, 12 January 2012 (UTC)
Anything could be interesting when taken out of context. In reference to the "blinking text"; what actually happened is this:
There's the context. I find it interesting that Δ is taken to task for making some text red and blinking, yet Intoronto1125 is never remonstrated for calling Δ an idiot or for edit warring in direct violation of WP:NFCC #10c. -- Hammersoft ( talk) 15:32, 13 January 2012 (UTC)
The above example dif, and the explanation, and the discussion surrounding it have been useful. I have got a good feel for the situation, and it confirms other impressions I have picked up. Thank you. SilkTork ✔Tea time 17:02, 13 January 2012 (UTC)
Is this a joke? The case languished in the Workshop phase with little movement for just short of two months. TotientDragooned ( talk) 20:10, 12 January 2012 (UTC)
For whatever it is worth, I am very willing to mentor Betacommand through whatever sanction, short of a site ban the committee may willingly adopt. I am generally available, and have an established rapport with him, such that I am confident "intense mentorship" could succeed.
In offering this service, I am willing to accept a tandem block, if and whenever Betacommand may warrant one. I would agree to be bound by a clause equivocating his failure, as a failure of my own. I would like to establish advance benchmarks that Betacommand and I can work in good faith to achieve which would clearly enunciate the path to normal editing. It would be a great advantage to success if a streamlined process was in place to request and perhaps effect temporary modification along the way if and whenever circumstances may warrant. And I assure with the weight of promise and the bond of word that I will monitor Betacommand and maintain a record of every potentially relevant action, discussion, or third opinion sought to show, demonstrably, the proactive steps taken to achieve "productive editing rehabilitation" and analysis of progress made. Respectfully submitted with prayerful hope and regard. My76Strat ( talk) 16:33, 11 January 2012 (UTC)
The above offer as well as some recent edits to the workshop page may be impossible to consider for procedural errors, missed deadlines, or any other reason, that being very new, I would reasonable not know. I hope where I may have been out of process I haven't been out of line. If the burden of my being tardy is too great, please advise me that I am too late. If there is any reason otherwise that I should compile what I believe is compelling, please advise so that I may begin in earnest, the proper construction of a document I do wish to respectfully present. My76Strat ( talk) 06:25, 13 January 2012 (UTC)
Please don't do this. It's hard enough recruiting and keeping BAG members. Instructing us to wade into this field of figuring out what Betacommand can do with a bot, both in light of the community consensus and this case's restrictions, is simply sliding the headache to us. MBisanz talk 13:00, 13 January 2012 (UTC)
Like, say, running unauthorized tasks with the bot? Given the years of drama surrounding this, ArbCom needs to spell out the enforcement procedure, if any. And most importantly, who is authorized to make this determination: ArbCom, BAG, or any "uninvolved" admin? And what are the appeal venues? ASCIIn2Bme ( talk) 13:52, 13 January 2012 (UTC)
The main remedy currently passing almost passing is 6 (
3.3.6 Betacommand limited to BAG approved tasks). There is also a proposal to vacate current sanctions. (
3.1.6 Community sanctions vacated). Most of the tangle of previous sanctions would be superseded by 3.3.6.
However, the interaction sanctions stand apart; they would be voided, and not replaced or superseded. There is no finding currently proposed that identifies an improvement in the quality of Beta's interactions with other editors, nor am I aware of evidence that would support such a finding. Arbitrator AGK, above, offered that the interaction problems occurred around scripts, not bots. Arbitrator Silk Tork, two comments further down, offered above that he would prefer look at only current evidence from a bot account. Neither reason adequately supports an affirmative action to void interaction restrictions.
A remedy addressing the quality of Beta's interactions with other editors needs to be crafted (or continued). If, on the other hand, the arbitrators choose to eliminate sanctions related to Beta's interactions with other editors, they should do so explicitly by remedy, and such remedy should be supported by findings of fact. Jd2718 ( talk) 15:18, 14 January 2012 (UTC)
Since this was another editor restricted to use automation only in BAG-approved cases (and was mentioned by Kirill), the discussion surrounding the final block should be informative here: [41] ASCIIn2Bme ( talk) 15:27, 15 January 2012 (UTC)
I am afraid Remedy 7 ("Betacommmand restricted") will leave us in essentially the same position.
There is a small advantage that #7 would be enforced at AE instead of ANI, but overall #7 seems like a very small tweak to the existing sanctions. — Carl ( CBM · talk) 12:25, 16 January 2012 (UTC)
Actually, creating such lists could also be used to deal with Delta's communication problems. If we implement something along the lines of Remedy 6, then Delta could be put on a 0RR or 1RR restriction. Then he could run a bot to check on reversions of his bot edits and then post them on some file which others can deal with. You can also imagine a deadline system where Delta is allowed to revert the unreverted items that are still on the list after a week. The community would then have a week (or whatever other time frame is seen to be appropriate) to either declare the bot edits to be wrongly reverted or to be correctly reverted and then remove such items from the list.
Count Iblis (
talk) 14:00, 16 January 2012 (UTC)
I'm afraid this remedy is an invitation to a festival of wiki-lawyering and perhaps WP:TAGTEAMing. Besides not addressing the issue that was the proximate cause of this arbitration (no consensus on what pattern or similar edits means), what does "both Betacommand and the subsequent user will share editorial responsibility for the edits made in this fashion" mean in practice? If the edits prove controversial in the community, are both editors going to be blocked? ASCIIn2Bme ( talk) 15:58, 19 January 2012 (UTC)
The purpose and intent of proposed remedy 7 is just what I said in my support comment when I proposed it: I think that if Betacommand channels his automated or repeated edits into a list of edits that can be reviewed and implemented by another user, we keep the best of his contributions (identification of pages with problems and a method of addressing the problems) while eliminating the big problems that people have with his contributions (errors and occasional mis-prioritization of tasks). Proposed copyedits to address any perceived ambiguities in the wording or to simplify the interpretation of the remedy would be most welcome here.
I appreciate all the comments in this section (and everywhere else) but I must say I am puzzled by ASCIIn2Bme's remarks, including that "I don't want to speculate why this comes from the Arbitrator who drafted the decision in the Betacommand 2 case." Anyone is free to disagree with any proposal I make or even to try to convince me to modify or withdraw it; following such comments is part of my job. But if the implication is meant to be that I am intentionally trying to create confusion or to stoke drama rather than work toward a fair and productive solution to the case, that is highly offensive. Newyorkbrad ( talk) 00:37, 21 January 2012 (UTC)
I'm sorry that Newyorkbrad felt attacked, but if I were to speculate about his reasons for posting this remedy, it would certainly not include deliberately causing confusion. If I understood the history of the community restrictions following the Betacommand 2 case, they arose because that case did not address the matter effectively. ¶ Alas, the subsection of the community that usually deals with Δ can't agree about the meaning of soup. Whether this happens because it involves many programmers instead of LL.D.s, again I don't want to speculate. Nevertheless, I think there is rough consensus in the community that the restrictions need to be simplified. I don't think remedy 7 manages to do this, rather the opposite, so I expressed my frustration in that regard, perhaps not in the best way. Regards, ASCIIn2Bme ( talk) 05:03, 21 January 2012 (UTC)
In regards to Franamax's comment on the Arb who said "Thanks for your comments"; I think that must be me. I had proposed a clean start option, and had received a number of comments for that, mostly critical of the proposal, and at the way it was done. Franamax left a message for me, which I took to be a rhetorical question designed to make me reflect on my action, and not intended to be fully answered. I took the message in the same spirit as the other comments I had received, and as part of my learning curve, and I thanked Franamax. As regards answering the question why I "didn't place those ideas first at the /Workshop page?" The idea had been raised in a few places before I formally put it on the Proposed Decision page, including Δ's talkpage, this talkpage, and the Workshop page. I felt the proposal was worth serious consideration as I believed it was a viable solution (still do). I feel that all viable proposals are worth considering, even if rejected, as the reasons for rejection can be helpful to a fuller understanding of the issues, even if only to answer the question - "Why wasn't XYZ considered?". As for why the PD page rather than the Workshop page for the formal wording of the proposal. I think that was mainly because the PD page had been opened, and because other proposals had already been directly placed on there rather than being workshopped first. I hope that clarifies matters, but if not, I'm prepared to expand. SilkTork ✔Tea time 01:39, 24 January 2012 (UTC)
For #6, one small tweak is to allow Δ to revert his own edits or edits by one of his bots in any namespace. During the testing phase of bot development, edits must occasionally be reverted, and there is no reason the sanction should prevent him from doing that while it allows him to develop bots. — Carl ( CBM · talk) 14:17, 16 January 2012 (UTC)
The PD refers to Δ by his former username (except in FoF1). This seems to be rather unusual since the usual practice is to use the current username and not a previous one (see, e.g., the Abortion case in which User:Anythingyouwant is referred to as such and not by his previous username Ferrylodge, even though he had a previous arbcom case under that name.) It might be a good idea to either add a sentence in the FoF like "As Δ is best known under his previous username, we will refer to him as Betacommand in the remainder of this decision for ease of reference" or replace all the "Betacommand" with "Δ". T. Canens ( talk) 16:31, 16 January 2012 (UTC)
Would someone please explain why typing Δ's name (7 characters, code without spaces = & Delta ;) is such an onerous burden? His chosen username is shorter to type than many of the arbitrators but no one is proposing forcing them into "simpler" names. Further, it is impolite in my view to insist on using someone's former name. You wouldn't (I hope) refer to a trans woman by her birth name, nor to a divorced woman who chose to revert to her maiden name by her married name. Admittedly courtesy in the online world isn't the same as in face-to-face interactions but I think the general principle of a right to self-identify extends to respecting an editor's choice of username. Now I recognise that there were good reasons to use the Betacommand name in labelling this case as it follows the earlier ones but surely we can respect Δ sufficiently to refer to him by the name he has selected.
If ArbCom does choose to pass a remedy forcing a name change - and forcing an editor back to a name surrounded by controversy - I would like to see a principle, grounded in extant policy, passed that provides authority for this action. Aside from the points that Δ has made about SUL and Hammersoft has made about the explicit policy allowance for non-alphanumeric names, this has the potential to be seen as another example of ArbCom overreaching its mandate and authority, and for little gain that I can see. EdChem ( talk) 04:20, 17 January 2012 (UTC)
The substance of policy is beyond the purview of ArbCom - that is, ArbCom must follow policy, in this case WP:USERNAME#Usernames with non-Latin characters, not change it. Accordingly, this proposal is beyond ArbCom's authority. While I have no opinion on the case as a whole, this particular proposal is itself a WP:POINT violation. -- Philosopher Let us reason together. 06:01, 17 January 2012 (UTC)
There is the matter of the fact that Betacommand started editing under Δ in order to circumvent a block. In my opinion, this has less to do with "triangle guy" and the name he chooses, and more to do with the fact that he violated another policy in order to circumvent his block. Forcing him to go back to his old name is certainly reasonable in that light, but I think it is complete bull$#!t to make him change it just because it is hard to type. Our policy explicitly allows that. Buffs ( talk) 19:44, 17 January 2012 (UTC)
Just a note that a current AN/I thread (<permalink to archive>) is somewhat stalled in anticipation of a decision here. I'll draw attention to my proposed principle on automated/high-speed/scripted editing - I believe this should be held to a much higher standard. Even in normal course editing and review, it's difficult to check over another's edits when many changes are made at one time. When even one error is finally sorted out from the diff, this is a matter for concern and becomes subject to blanket reversion. When there are hundreds of these edits occurring in a short period of time, it becomes simply overwhelming to maintain any standard of quality. To me, the question arises why the humans who are working hard here should be expected to just give up and accept whatever the high-speed editors decide is the right thing to do, even when it's wrong. Franamax ( talk) 06:46, 24 January 2012 (UTC)
I think people mentioning WP:BRD completely misunderstand BRD if they are proposing applying it to bot edits. BRD means, for example, that if I'm considering a good-faith but potentially controversial edit to the Great Pyramids article that adds several paragraphs claiming (say) that the pyramids were actually built by penguins, I shouldn't agonize too much about its possible reception by other editors. I should just boldly make the edit. The rationale is that even if it took me hours to write the paragraphs, if the edit turns out to be unacceptable, another editor can look at my change for a few moments, decide it's not appropriate, and just click "undo" to revert it, so my bad edit hasn't burned much of anyone's time but my own. The whole basis of "the encyclopedia anyone can edit" is similar--we open ourselves up to (among other things) routine vandalism, on the theory that it's easy to revert.
With bots, it's the opposite. Someone rather easily launches a bot operation that makes 1000's of edits, and it's then extremely difficult for human editors to examine or undo them. Please see User talk:75.57.241.57/x (one of several pages related to a single incident) as an example of the level of effort that it takes to revert a bot operation gone wrong. Someone (I won't name him here since this was an isolated error by an otherwise good editor) got over-enthusiastic with Twinkle and protected several thousand templates without consensus. It took days of hassle including N hours of custom programming (by me, with the help of discussion from other editors that consumed a lot of their time as well), and more hours of drudge work by a helpful admin who did the unprotections, to reconstruct exactly what Twinkle had done to each template and undo it. I've been involved in some other messy high-volume cleanups as well, but most often what happens is that the bad edits simply never get cleaned up because it's too much hassle, and the encyclopedia suffers long-lasting damage.
TL;DR: BOLD is only for actions (like most single edits) that are easy to undo ("revert"). It should never apply to difficult-to-undo operations such as large clusters of edits (automated or otherwise), edits that make irreversible disclosures or provoke conflict, etc. Anyone who has never actually reverted a bot rampage themselves (which can mean having to write another bot to clean up after the first bot) shouldn't go around saying that it's easy or expect other people to do it as part of BRD. If you really think BRD applies, let's see you do the revert. 67.122.210.96 ( talk) 02:40, 25 January 2012 (UTC)
Would an arbitrator please indicate the evidentiary basis for FoF 6.1 - Ongoing communication problems? Thank you, -- Hammersoft ( talk) 19:57, 24 January 2012 (UTC)
I have a quibble with
Minor edit is Wikipedia terminology for edits like spelling corrections, that don't add or remove content. They are flagged as "m" in edit histories which stops them from showing up on some watchlists or searches. The controversies around Δ's editing are mostly about edits that generally do add or remove content (usually remove), so they don't qualify as "minor" in that sense. I could go with "insubstantial" (except when they are wrong), but then why bother making them? 67.122.210.96 ( talk) 07:26, 25 January 2012 (UTC)
moved from project page -- Alexandr Dmitri ( talk) 23:06, 25 January 2012 (UTC)
I use two non-public svn systems. I do not plan on changing that. I open my code up for those who I can trust and who I know will not abuse my tools. If anyone is interested in accessing my web based repo they should contact me. However as it stands I will not publicly release my code until such time as I have done extensive testing on it, and it has been proven stable over a very long period and that I have had time to build in the idiot proofing needed for such a public release. Since I rarely add in the needed fail-safes for people unfamiliar with my code, I rarely make anything public. ΔT The only constant 19:16, 25 January 2012 (UTC)
Δ, I'm puzzled--why is your code somehow more dangerous or abusable than the tons of other bot code already out there? Why would someone capable of modifying your code to do something vandalistic not be capable of writing their own vandalbot? The MW:API including the editing interface is reasonably well-documented after all. I agree that vandals shouldn't be allowed to run bots making 5000 edits per hour, but I see that possibility as basically a server bug rather than something enabled by any particular client code getting loose, so I'd support adding a server patch limiting edit rates to 100 edits/hour (or some comparable number) for any accounts without a bot flag. I do understand your feeling of sitting on crufty, unreleasable code (we all have code like that) but maybe you should take the view that long-running bot code shouldn't be deployed unless it's production-ready (i.e. cleaned up, documented, etc). For shorter term tasks, I'd consider it a reasonable compromise if some other editors (BAG members, say) were known to have private access to the code.
How does the following modification sound: code doesn't have be released for bots where the BRFA process establishes that the approval is for a non-recurring, one-off task that is expected to be completed (i.e. over and done with, no follow-ons expected) in 30 days or less from initial deployment. Longer-running bots are expected to have clean, documented, release-quality code in a public place. 67.122.210.96 ( talk) 00:30, 26 January 2012 (UTC)
It appears that the BAG-approved-only-tasks resolution is what is going to pass. I don't think that the open-source-only-tasks is going to pass, but I have no idea. How about a middle of the road solution then. Betacommand, is there any active/semi-active member of the Bot Approvals Group (there are eight active and six semi-active) that you would not trust in showing the code to? If not, find a secure way to make the code available to those people. To be brutally honest, I think that all of this concern over the code is nonsense, all we need to care about is the end product, but for the people who are concerned about the code, this seems like a good compromise. Sven Manguard Wha? 02:15, 26 January 2012 (UTC)
All,
I removed a comment as it insinuated that delta committed a crime or could possibly be fired from his job from misuse of company property (if proven true). Since this allegation had no link or backup to the claim, I removed it IAW WP:BLP and explained it on the user's talk page. It should also be noted that this comment was the user's SOLE contribution. Perhaps it isn't a WP:SOCK, but since the vast majority of WPians don't start at editing in ArbCom, I think I'm hearing a WP:DUCK quacking... Buffs ( talk) 21:05, 26 January 2012 (UTC)
OK, I guess I can go along with the idea that we don't really need to see this, er, polydimensional code. Per my "general discussion" workshop comment, program bugs per se haven't particularly been the issue in most of these incidents: the programs usually do work as intended, and the problems have mostly been in task selection, large editing runs without enough testing and feedback, problems dealing with other users, etc. It's not even that helpful to do anything sounding like pushing responsibility for Δ's software problems onto other users or expecting us to do the work of reviewing it. Δ, you've mentioned that you've given access to the code to a few other users. Would you be ok with saying who a couple of them are? That might be good enough for the auditing purposes discussed in the previous thread. 69.111.193.96 ( talk) 07:12, 27 January 2012 (UTC) (address changed)
Requiring that his bot be open source does not require that the framework is open source as well. This polydimensional framework has been in development for years. The polydimensional framework API should be stable. If it was cutting edge, there will be research papers about it already. The proposed remedy forces Betacommand to use a bot framework which does have a stable published API that permits linking. The only problem would be if the polydimensional framework API is a trade secret, and Betacommand decides that he doesnt want to write his bots (existing and future) on a bot framework which does have a published API that permits linking. Some bot operators have been able to develop closed-source cutting edge frameworks which are research (the Cobi vandal bots?), and we've seen the research papers come out of this. Sadly it isnt a good idea to let Betacommand continue to do this. John Vandenberg ( chat) 11:22, 27 January 2012 (UTC)
An editor "frustrating" others is not sufficient for an ArbCom sanction, e.g. admonishment. In general, Wikipedia discussions of users should be phrased in terms of behaviors, rather than in terms of emotional states of editors, e.g. "frustration", which are more subjective.
An editor violating behavioral policies, e.g., by failing to reply to reasonable requests for explanations of edits or constructively to engage in discussions, is a legitimate concern for ArbCom.
Thus, proposal 6.2 is superior to proposal 6.1. Kiefer. Wolfowitz 23:24, 26 January 2012 (UTC)
Re Kirill "I don't believe we have the authority to force a user to release material under a particular type of license": right, arbcom can't force the release of code. But it can certainly decline to let the bots keep running if Δ doesn't choose to release the code. Whether it's important in this particular situation is a different question, but IMHO it is a legitimately available remedy.
67.119.12.141 (
talk) 08:55, 27 January 2012 (UTC) Looks like Jclemens and Coucrelles already made this same point before I did.
67.119.12.141 (
talk) 20:43, 28 January 2012 (UTC)
This case has been dragging on for an excessive period of time now. At this stage, we have ArbCom agreeing via proposed principles that:
Ok, moving forward we have the following findings of fact being agreed upon;
In short; so far ArbCom hasn't identified the problem. If you can't identify the problem, tossing around remedies isn't going to solve anything. So not surprisingly, in its remedies ArbCom decides
And here's the kicker; 16 remedies haven't passed or have failed outright. No small wonder; again, if you can't identify and support with evidence the actual problem, you can't craft a remedy to fix it. 2 1/2 months ago I noted ArbCom's incompetence in the establishment of this case. Move forward to today, and the fumbling about is still happening.
Rather than focusing on remedies at this point, ArbCom needs to go back to the drawing board and try to ascertain what in fact the problem was when this all started.
Lost in all of this is the fact that ArbCom screwed up. See FoF 5. In July of 2009, ArbCom assigned two mentors to Δ (see unban decision), who failed in their appointed tasks, and neither of whom informed Δ they would no longer continue to mentor him. In fact a year after they'd been made mentors, they still had the mentorship note on the top of his talk page [58] that one of them had placed there the year before [59]. Worse, ArbCom didn't even take notice they were not receiving monthly reports. Where's the FoF that ArbCom screwed up in their self appointed duty? Where's the FoF that the mentors who had been assigned to Δ failed in their appointed duties?
Here's another good one. 7 January 2012, ArbCom proposes Δ be limited to WP:BAG approved tasks ( Remedy 6). Did they bother to ask BAG about this, or even notify them they were being included as a possible remedy? Nope. I had to notify them myself, five days later [60]. Of course when someone from BAG did show up less than a day later, they opposed it and begged them not to do it.
So in summary;
It's high time ArbCom took a good long look in the mirror.
I would really, really like some members of ArbCom to respond to these queries. -- Hammersoft ( talk) 19:07, 27 January 2012 (UTC)
ArbCom should recognize the very serious failure they made in not upholding its self assigned tasks they agreed to in the 2009 unban decision. In dropping the ball, they set the stage for a series of unfortunate events. Refusing to recognize their role in this failure is tantamount to allowing Δ to fail, and blaming him for the troubles directly descendant from their joint mismanagement of this case. Whether such recognition is made public is immaterial. But, subsequent to that recognition, ArbCom should allow themselves, the community, and Δ a do-over to the effect that they go back to their agreed upon resolution as of 11 July 2009. Thus, I propose Remedy 9;
9) ArbCom vacates the current Δ3 case, and reinstates its 11 July 2009 decision, including the assignment of two mentors. This set of restrictions is amended to permit operation of Δbot, and the development of code to be used by other people, recognizing that such other editors bear full responsibility for the execution of said code. Further, the edit throttle is set to the community sanctioned average of 4 edits per minute over any 10 minute time period.
I would sternly advise that any would-be mentors understand their roles are voluntary, but they must notify ArbCom and Δ if they withdraw from their duties. Additionally, ArbCom should sternly advise itself that it needs to maintain contact with the case and follow through with their role in it. -- Hammersoft ( talk) 19:07, 27 January 2012 (UTC)
I remain unconvinced that a second mentor is going to magically materialize just in time to resolve the case. I'm also not sure if ArbCom will allow My76Strat to mentor given that their view of things seems to be at odds with ArbCom's view of things; since a mentor's job is to fix things rather than sweep them under the rug, ArbCom might want someone who agrees with them about the existence of various problems. -- N Y Kevin @145, i.e. 02:28, 3 February 2012 (UTC)
AGK wrote above that "the committee is hopelessly divided in how to dispose of this case." Is that really true, though? Eight arbitrators have supported Remedy #6 already, which means it will pass with the support or formal abstention of just one of the four arbs who haven't voted on it yet. Newyorkbrad, Risker, Coren, and Mailer Diablo should weigh in on this remedy. A Stop at Willoughby ( talk) 19:39, 4 February 2012 (UTC)
Several people appear to be of the mistaken impression that Δ is not allowed to run bots or automated programs. This is in fact false.
If anyone has any evidence proving he is currently forbidden from running bots or automated processes, please do advise. I'm well aware of the edit throttle. That's not my point in this thread. -- Hammersoft ( talk) 19:59, 31 January 2012 (UTC)
I am hatting this off, in my capacity as case clerk, for the second time. Do not revert or continue this discussion here. |
---|
The following discussion has been closed. Please do not modify it. |
Hammersoft / Buffs dialogue
|
There has been very little activity on the proposed decision in the last week and I want to encourage the arbiters, especially those who have not yet voted on the decision to weigh in. Remedies 2.1 and 6 in particular are both only one vote away from passing. I do not think that spending additional time trying to craft an ideal compromise sanction is worthwhile at this point. Arbitration is not magic pixie dust™. The advantage of an Arb case over an ANI discussion is that one or the other can pass (even if it is 9-7) and we can move on from the status quo, which is widely considered unacceptable. Eluchil404 ( talk) 20:27, 5 February 2012 (UTC)
Informal tally. Feel free to format better and update
(2.1) (6) Kirill sup opp SirFozzie sup opp Jclemens sup opp Coren sup --- Fuchs 1st 2nd Davies sup opp Vandenberg sup sup Courcelles 2nd 1st Risker sup opp PhilKnight opp sup Casliber opp sup AGK opp sup Elen opp sup SilkTork opp sup Mailer sup --- Newyorkbrad opp opp
I think John Vandenberg should state a preference between 2.1 and 6. Right now that doesn't matter, but if another Arbitrator were to vote for (6), it would put that over the 9 votes bar as well. ASCIIn2Bme ( talk) 22:45, 5 February 2012 (UTC)
Mailer Diablo has now voted on (6). Only Coren is missing in action there. ASCIIn2Bme ( talk) 07:15, 7 February 2012 (UTC)
One more vote on remedy 6 could make a difference. Just a standard user requesting a hold on closure for an additional 24 hrs. Buffs ( talk) 21:53, 6 February 2012 (UTC)
@Jclemens: Are there prior cases where ArbCom violated its own stated principals in how to close a case? At the time your clerk said they were going to close the case, there were three net closes. We're now down to one net support to close. Your "advice" is chilling; in effect you're saying I am out of line for raising an issue with how ArbCom (in this case a clerk) is doing its business. I remind you that you operate at the behest of the community, not the other way around. This is no different than an administrator telling a non-administrator "I advise you to not tell me what to do". -- Hammersoft ( talk) 14:27, 8 February 2012 (UTC)
How about we have a calming effect by both sides disengaging? SirFozzie ( talk) 18:35, 10 February 2012 (UTC)
I propose Article 35 to an ArbCom code, specifically that 34 articles be developed to restructure ArbCom into an institutional element worthy to serve as a means of last resort. I am deeply concerned upon observing this process, my first; and respectfully advise the committee that I believe they, as a body, have failed their charter, seriously under-serving their largest constituent; Wikipedia herself! I'll leave the distinguished members to develop the other 34. My76Strat ( talk) 09:14, 7 February 2012 (UTC)
Hammersoft, if there really has been sockpuppetry in this case, why haven't you opened a WP:SPI (or if you did, where is it)? Have you informed the committee of this? Shall we believe you on a bare assertion alone? -- N Y Kevin @078, i.e. 00:52, 8 February 2012 (UTC)
NYKevin, I agree that a sock-puppetry investigation would be the most transparent venue in which to resolve concerns of case-page socking, but it is acceptable to e-mail an arbitrator privately; and moreover, such a solution could be more elegant. On a more general note, I don't think the issue of socking in this case is really worth significant focus, and I would prefer that we spend our time refining the proposed decision. AGK [•] 11:52, 8 February 2012 (UTC)
My76Strat ( talk) 10:23, 8 February 2012 (UTC)
@AGK: So the sockpuppetting is not worth significant focus? At what point does it become unacceptable for sockpuppets to abuse the ArbCom process? We've got IPs making unfounded accusations that Δ is responsible for a denial of service attack against another person involved in this case ( submitted 'evidence'). We've got IPs making an accusation of 112 edits in a ten minute period ( submitted 'evidence') on 18 May 2011 when in actuality there were three edits [69] in the stated time period. Oh but not to worry, this is all perfectly acceptable because we've got an ArbCom member using it as evidentiary basis. And really, it's ok for sockpuppets to submit evidence. Afterall, the 500 word limit is really just a minor irritant to those willing to sockpuppet to submit evidence. -- Hammersoft ( talk) 14:16, 8 February 2012 (UTC)
Hey, Hammersoft, put up or shut up. Post to WP:SPI or stop posting about this at all. Jtrainor ( talk) 14:21, 8 February 2012 (UTC)
HS, back to your query, I'm just not seeing any evidence of sockpuppetry. As for ArbCom allegedly using this sockpuppet's evidence, it only garnered a single vote, so it had no actual effect. Let's say for the sake of argument that there was a sockpuppet. It didn't have any real effect, so there's nothing to be done here. I concur that a SPI is probably a good route to follow for these accusations. If they were sockpuppeting, then they deserve to be held to account for their actions. Buffs ( talk) 15:09, 8 February 2012 (UTC)
The California ones are probably all me (my address changed several times during the case for technical reasons mostly outside my control). I have no idea about the TX one or the story with that "112 edits" diff (even after I spent a while trying to figure out what went wrong with it). Some of the other diffs got results with discrepancies from my own checks, maybe because of some difference in how things were calculated. It's also possible that article or image deletions caused some gaps in Δ's contrib history that made some of the discrepancies in the counts.
In my own submission, I posted a diff indicating Tristessa had been DDOS'd, but Hammersoft should note that I didn't make any claims of any sort about how it might have happened. I simply thought it was a weird enough incident that I decided to call it to arbcom's attention so they could check into it further if they decided it seemed relevant.
I don't know why Hammersoft is pursuing these conspiracy theories anyway. For quite a long time, Δ-related drama seemed to in part be proxy battles about NFCC enforcement, so it was impossible to say anything about the bots without appearing to take sides in the NFCC battle. (On NFCC philosophy I'm actually mostly on Hammersoft's side; I'd just prefer for WP to deal with NFCC by much different methods than Δ used). However, with Δ banned from the NFCC area, we're left with what I'd have thought was a less "political" issue of straightforward recurring bot misuse. So I don't understand why Hammersoft is so angry. 67.117.145.9 ( talk) 19:13, 8 February 2012 (UTC)
I did my own edit speed calculation and added it to evidence. I think the issue is not in any real doubt. 67.117.145.9 ( talk) 07:34, 9 February 2012 (UTC)
I've done some background reading on the MSK case, and this struck me as relevant here as well:
As a result of this history I believe the original mentorship was a failure, and that the Mentorship Committee is not up to the task of monitoring problem users. So far as I can tell, there are no follow-up procedures, no functioning communications, etc. Fundamental problems are that the job requires one to be a policeman/therapist/teacher, while most of us came here to be editors, and that it is a bureaucracy requiring procedures and maintenance. Those deficiencies should be addressed, in another forum, before new cases are assigned to it by the ArbCom or by goodwill ambassadors.
— User:Will Beback 08:30, 2 June 2006 (UTC)
Are there any cases where ArbCom-mandated mentorship has succeeded in the long run in reforming editors? ASCIIn2Bme ( talk) 09:48, 8 February 2012 (UTC)
http://wikipediareview.com/index.php?showtopic=34323&st=20 — Preceding unsigned comment added by 76.69.228.210 ( talk) 04:48, 11 February 2012 (UTC)
So, what the ArbCom is now doing, is first deciding to ban, but upon realising that they did not have a finding of fact supporting such a decision, they then find a FoF to vote upon so that they have a reason to ban? Do you guys have any clue how this is appearing now? -- Dirk Beetstra T C 04:19, 13 February 2012 (UTC)
What I want, is either that you, even if that takes months and months of deliberations, reach a proper decision on wiki. But here, you obviously first decided to ban, but did not have a significant FoF that would accomplish such a decision (1 is general, 2.2 and 4 are things which are by far not necessarily Betacommands fault if they are a fault at all - part of the failure is the community and even ArbCom's fault, 6.1 is the closest to finding a problem due to Δ himself - one that actually could have been solved (if it is not already solved), if the mentoring by ArbCom would not have failed. You used those FoF's to decide to ban (some even vote to close because there is a ban decision ..), and then it stopped for some time. Now, in a burst of a couple of hours, you see Arbs vote on a last FoF, the only one that actually defines a problem with Δ. Yes, my question remains: do you guys realize how this looks? Either it looks like you just decide to ban and then realize that there is no FoF actually sufficiently supporting that, or you fail to have enough time to actually look properly into cases. Yes, I want want proper decisions, just as what I expect of a judge deciding over life sentences (or worse). Yes, there will always be people unhappy, and yes, they will always make mistakes, that will happen in all processes. But if a judge decides such things, at least the severity of the punishment is in line with the severity of the crime. It is good to read that you, Jclemens, are completely comfortable with first deciding to ban ( diff), but only last night found reason (enough?) to do so:
... apparently not, because it takes then a couple of days of (off-wiki?) discussion (or you are just busy with another case or real life), and suddenly a realisation that the case needs to be closed (PhilKnight suggests that it needs a fast close because of time constraints ..), and that you actually need a proper reason, which, obviously, you did not have. And even this one (FoF 3.1), is citing the weasel word 'often' quite a number of times. Is 0.1% of the time over the speed limit 'often'? If the total average speed is less than 1 edit per 10 minutes over the last year then that suddenlty becomes 'a few'. One single case of incivility (where I can show cases of (mild) incivility against Δ by, low and behold, even an ArbCom member) - without even a timeframe, IIRC this restriction is there for 2-3 years? And a number of sub-optimal edits / mistakes (no violations not of the type of POV violations, CopyVio's etc.), and related to my last point, some cases of 'saving without reviewing' (I wonder how you can even prove such a thing, you must have been looking over his shoulder, or have a way of defining that edits must have been not properly reviewed). Those two points would be 'carelessness' .. at worst. And that in 10s of thousands of edits! And that is what you decide to ban over in the end? That is utterly out of perspective, and calling it 'ignoring the restrictions' is then a complete misinterpretation of the facts.
You consider this 'trying to find a middle ground where various options are explored and consensus is sought for a lesser remedy before banning is the default outcome'? No, this still looks like you were finding a way of banning Δ, not caring about the improvements he made (since last cases, or generally over time), looking at the accusations/evidence out of perspective (again, 30 speed violations in a year is less than 0.1% of the time, only one slight violation in the last 6 months, with an average editing speed which is significantly below 1 edit per 10 minutes (taking 20.000 edits in a year; far, far from the 40 per minute)).
I am convinced that there is a problem here, but your actions here certainly do not convince me that Δ is the (only) problem (IMHO, he is just a part of the problem - you may want to look at the current discussion about about how WP:AN/I works in this respect, and NFCC work in perspective of the SOPA/PIPA Wikipedia blackout), and whether this is the only or best solution. I remain, this ArbCom case is an utter failure, and this remedy is far, far too harsh for the passing finding of facts, especially before 3.1 passed. -- Dirk Beetstra T C 07:07, 13 February 2012 (UTC)
Silktork, you here say: "What is true is that the sanctions have failed, and without agreement on an alternative to the sanctions, then a ban is viable even without this finding." Without this finding, you present no evidence that there is any disruption, or that it is Δ's fault that the sanctions have failed (actually, ArbCom itself failed some of its own sanctions of the last case), or that Δ is actually the cause of the disruption. You basically suggest 'because we can't find anything else, we just ban Δ and we are ready.' Could you please elaborate on that remark? -- Dirk Beetstra T C 10:32, 13 February 2012 (UTC)
The point of the FoF is for the arbitrators to tell us why they voted on the ban. They are not required to give any such explanation, strictly speaking, but it is certainly a good practice for them to do so and I'm glad they are making the effort. A similar situation often happens in committee meetings in real life in my experience. After a long discussion, the committee may come to a decision and take a vote to record it. But if they have to also write a short paragraph expressing the official opinion of the committee about why they reached that decision, this paragraph is often harder than the decision itself, because different committee members may have different reasons, and not be willing to agree on the reasons even though they agree on the outcome. — Carl ( CBM · talk) 12:18, 13 February 2012 (UTC)
According to the motion to unban Δ [70]:
It seems like this is relevant to the current remedy. — Carl ( CBM · talk) 12:32, 13 February 2012 (UTC)
PhilKnight says "we have other cases waiting, and we need to close this in the near future". Why? Is there some deadline we are not aware of? Perhaps ArbCom isn't aware that Δ hasn't edited in more than a week. There is no ongoing disruption, no outstanding complaints, no questionable actions happening. Hell, he hasn't even edited mainspace in more than a month! But now we must rush to close this case? Why?
And still, as mentioned in various threads above we have serious problematic issues in play;
Not to worry though, just make veiled threats to not question ArbCom and its clerks, and then close the case without any basis in fact. Just sweep the whole unfortunate mess under the rug, and everything will quiet down in time.
ArbCom, we KNOW you're divided on this case, and from statements by various members of your body we know you are polarized by opinions you brought to the table when this case opened. But, we didn't elect you to adjudicate cases based on opinions of an editor you bring to the case before it opens. You are compelled to dispassionately review the evidence presented, and make findings based on that evidence. You aren't doing that. This case severely lacks any evidentiary basis from anything over the last six months to support findings against Δ, and the result of this case is to ban Δ from the project.
ArbCom, what are you going to do in a year when Δ appeals to be unbanned, and you fail to provide any evidence of behaviors he is supposed to correct that have not been corrected in the last six months? Hmm? What then? -- Hammersoft ( talk) 15:39, 13 February 2012 (UTC)
See [71]. So, the concerns that BAG isn't going to address issues with non-consensual edits by a bot after approval, but that they would punt them back to the community, seem valid. I floated an idea to SlikTork a while back about a "BotCom", but I guess we're far away from a consensus for an institution that is able to deal with bot-related disputes. ASCIIn2Bme ( talk) 21:20, 14 February 2012 (UTC)
Case clerk: AlexandrDmitri ( Talk) Drafting arbitrators: Kirill Lokshin ( Talk) & Elen of the Roads ( Talk)
Wikipedia Arbitration |
---|
|
Track related changes |
Behaviour on this page: Arbitration case pages exist to assist the Arbitration Committee in arriving at a fair, well-informed decision. You are required to act with appropriate decorum during this case. While grievances must often be aired during a case, you are expected to air them without being rude or hostile, and to respond calmly to allegations against you. Accusations of misbehaviour posted in this case must be proven with clear evidence (and otherwise not made at all). Editors who conduct themselves inappropriately during a case may be sanctioned by an arbitrator, clerk, or functionary, without further warning, by being banned from further participation in the case, or being blocked altogether. Personal attacks against other users, including arbitrators or the clerks, will be met with sanctions. Behavior during a case may also be considered by the committee in arriving at a final decision.
Active:
Inactive:
Recused
Excuse me for being blunt, but the proposed decision date has already passed, and there hasn't been any movement at all on the proposed decision page. Now I know that most of your discussion takes place where no one can see it, but I worry that this case might not get closed out by the time that the ArbCom changeover takes place, and that this is percisely the kind of mess that shouldn't be handled over the course of two terms.
Finally, while I'm not accusing anyone of doing this, if action on this is being delayed because it might affect the ongoing elections, I'd like to say that such behavior would be extremely inappropriate. Sven Manguard Wha? 04:07, 3 December 2011 (UTC)
According to "Majority reference" table: 1–2 (Abstentions) result in 7 (Support votes needed for majority) Bulwersator ( talk) 17:34, 13 December 2011 (UTC)
I think there's a bit of cognitive dissonance in holding those positions simultaneously. I think it's fine to say that the sanctions were grounded in real concerns, but on the other hand the whole pattern identification business turned out impractical when participants in the drama couldn't agree that a script-generated sequence of edits was a pattern. The restrictions while well-meaning didn't work as intended. So, it makes little sense for ArbCom to confirm or "take[s] possession" of them as-is. ASCIIn2Bme ( talk) 04:15, 4 January 2012 (UTC)
Picking up on the comment by Courcelles, one possible alternative remedy would be to ban Betacommand from everything except running automated tasks via a flagged bot account approved by the Bot Approvals Group. He would be permitted to edit from his main account only on talk pages or noticeboards and only to discuss his bot operations (with an exception for his or his bots' userspace). Any other edits would have to come through his bots, and must be approved tasks. This would allow, for example, User:Δbot to continue clerking the SPI venue (I've heard no complaints as to his operations there). – xeno talk 13:56, 4 January 2012 (UTC)
When we turn to the proposed decision and consider how to resolve the issue with Betacommand's approach to interaction, I rejected the one-year ban, because the problem is with Betacommand's interactions - which a one-year break from Wikipedia would not solve. An indefinite ban, with a return only upon the presentation of a satisfactory plan for contributing constructively, would be a solution in that it would ensure Betacommand could only return if he became capable of contributing without the generation of a plethora of heated discussion. However, if we could agree that Betacommand was only problematic inasmuch as he runs (and apparently messes up) unapproved cleanup-style runs (not vetted bots), then a sensible solution could be to ban him from making any edit that was not approved in advance by BAG. Of those remedies that would meaningfully resolve the "Betacommand problem", I am inclined to pass the least severe proposal; if we do so and the problem persists, a siteban could be instituted by motion with little difficulty (and a remedy resolving that we will siteban in the event of continued issues with Betacommand may need to be included). Apologies for the verbosity, but I wanted to set my views down clearly before we proceed. AGK [•] 15:16, 4 January 2012 (UTC)
I'd like to explore Xeno's suggestion, if Betacommand believes this would allow him to continue to be useful; I've started two proposals for article-space restrictions on the workshop. See Workshop#Article edit rate restriction. John Vandenberg ( chat) 22:01, 5 January 2012 (UTC)
This seems like a terrible idea. If someone has a long and intractable history of editing disruptively in some area (say articles about penguins), usually they get topic-banned so they have to stop editing about penguins but can still edit any other topic they like. The proposal here is the opposite, something like "so-and-so may henceforth only edit about penguins, subject to the endorsement of the penguin approval group". I don't have any problem with Δ editing like a normal editor (i.e. completely manually and on a normal human scale) but IMHO he should be completely removed for a while from bots and scripts of any sort (thus the 25 page/day limit proposal). To Silktork re "saving useful bots": the bots involved are mostly not that useful (I'd say they're worse than useless, but that's just me), and any useful ones can be reimplemented by others. We should not let "vendor lock-in" dictate our actions in any case. 67.122.210.96 ( talk) 05:30, 9 January 2012 (UTC)
Give how discussions seem to work at BAG [3] I don't know if this is actually going to solve anything with respect to drama. ASCIIn2Bme ( talk) 17:46, 11 January 2012 (UTC)
I am glad this comity is considering clean start as a possible remedy. I beg members to be objective in weighing this option. There are assertions that this form of remedy implies an external blame is at core. Please read the policy again and if I may be so bold as to ask, please also show the prose that supports the assertion. It is problematic that Betacommand is under current sanctions, because that would preclude availability of this remedy. The proposal to vacate sanctions could be adopted and the clean start immediately ensue. With well crafted word, and the prohibitive sanction against evading scrutiny, this remedy does no harm to the encyclopedia, nor place jeopardy upon it. With respectful intent, thank you for considering this case, and this request as well. - My76Strat ( talk) 03:16, 5 January 2012 (UTC)
@Jtrainor - What do you find difficult? Can I ease the difficulty? I do stand available to answer any direct question you may have. In general, I do oppose the comparison of Betacommand's conduct with criminal conduct. This because generally I see no comparison. And I am just naive enough to believe it is not necessary. We have a southern saying: "you don't kick a man when he is down" which I have come to understand, and agree. Perhaps we could see a bit less "kicking", or not. My76Strat ( talk) 16:56, 11 January 2012 (UTC)
I know I found them when reviewing evidence a month ago, but I am not entirely certain that the version of community sanctions I have found is the most current one, particularly given all the amendments and modifications done over time. Could someone please link me to the *current* iteration of Betacommand/Δ's community sanction? Risker ( talk) 04:50, 5 January 2012 (UTC)
There absolutely are ways forward instead of back. And Xeno's proposal above is one such. The attempt to follow the "community imposed sanctions" was another, which resulted in more problems than it resolved (perhaps suggesting that the "community sanctions" were a complete failure, in terms of drafting, implementation and observance).
Regardless, if the result of this case is a ban, or something that destroys the value (either net, or potential) that Δ can bring to the project, it will be a sad day.
Rich
Farmbrough, 16:34, 5 January 2012 (UTC).
Just adopt my proposal about mentoring proposed some time ago on the Workshop page. So, some editor in good standing who has the relevant experience with operating bots, will be in charge of setting restrictions on what Delta is allowed to do. This solves two problems. Not only are the problems that Delta has caused contained, but also the drama generated when Delta is alleged to have violated some restriction will end, because Delta's restrictions are now off topic to the community. This is now the sole task of the mentor. The community may raise some perceive problem about Delta to the mentor, but it's always the mentor who takes some decision w.r.t. restrictions. If Delta violates a restriction set by the mentor, then that's also an issue between Delta and the mentor only. The mentor does have the duty to report Delta back to ArbCom if Delta were to misbehave and/or not stick to the rules set by the mentor. ArbCom can then revoke the mentorship agreement, which automatically bans Delta from undertaking any automated tasks. Count Iblis ( talk) 23:14, 7 January 2012 (UTC)
Assuming the community restrictions are continued/imported as AE-remedies, who is going to be considered uninvolved for the purpose of AE discussion and decision-making? This is a serious question because these are not typical content disputes focused on a topic area, but are about systematic changes to many unrelated articles. Reading the proceeding here and also the previous AN[I] threads, one cannon escape the feeling that over time Δ has engendered a dedicated fan club who shows up with regularity at most discussions about him. And the positions of the participants in previous disputes involving Δ are a very good predictor of their position in subsequent ones. A rather obvious and prolonged factionalism has developed around the "Δ-problem". ASCIIn2Bme ( talk) 06:57, 6 January 2012 (UTC)
Also, the recently but inconclusively closed discussion Wikipedia:AE#Suggested_change_to_practice:_comments_by_non-neutral_editors should be of relevance here. ASCIIn2Bme ( talk) 07:02, 6 January 2012 (UTC)
I think there were some CU-confirmed socks besides the "official" list given in this Arbitration. See Category:Wikipedia sockpuppets of Betacommand. Amusingly, one of these was doing in 2009 the kind of "clean up" that was again in focus in the autumn of 2011: Special:Contributions/Shimane Kanagi. ASCIIn2Bme ( talk) 10:18, 8 January 2012 (UTC)
The antecedent is stated
here. I.E. the case was a request for clarification as to whether a proposed task fell under the aegis of a previous ARBCOM imposed restriction.
Rich
Farmbrough, 18:16, 11 January 2012 (UTC).
Actually I realise there is an antecedent, in that the complexity and finickiness of the restrictions meant that there was alot of squabbling and lots of people unhappy. Hopefully the new remedy sorts this out. Casliber ( talk · contribs) 11:16, 12 January 2012 (UTC)
Some months ago, I began developing a list of Δ's accomplishments on Wikipedia. This was in response to a rather vocal group of editors who despise Δ, and seem to think everything he has done is sort of a inverse Midas; 'anything he touches turns to crap'. Given the committee seems to be hung up on trying to decide whether Δ is a benefit to the project, it would be in their interest to consider this non-exhaustive list of his accomplishments here. See User:Hammersoft/Δ vitae. -- Hammersoft ( talk) 23:08, 11 January 2012 (UTC)
Why is it the business of Arbcom to hand out back-pattings and attaboys? Jtrainor ( talk) 23:34, 12 January 2012 (UTC)
Thanks Hammersoft; can you finish adding the dates, order it by date, and put it in /Evidence. Thanks. John Vandenberg ( chat) 10:00, 24 January 2012 (UTC)
I have followed this case as a completely uninvolved observer. I expected a choice between a ban and something like this (Bot Group approved tasks only) - I think they are choices that would address the issue rather than skirting it or nibbling at it, maybe the only choices that would do so. However, if Beta's to stay active, it is not enough to allow interaction with users - the quality of that interaction should be addressed as well. I'm not workshopping a proposal for you - I think it needs to be handled more skillfully and gently than I could - but there needs to be a remedy directed at the quality of Beta's interaction with other users in conjunction with the BAG-approved only remedy. Jd2718 ( talk) 11:50, 12 January 2012 (UTC)
I'm basically saying SPI bot is irrelevant to his ability to operate an article-editing bot. As to the last question, we can look at the history of arbitration cases, community sanctions, bans, and the evidence in this case, and everyone can come to their own conclusions.
My opinion, which I have previously expressed, is that if we don't want a full ban, the best option is to impose much firmer sanctions that do not rely on the ability of Δ to modulate his own editing. The BAG-only option seems like it has a possibility of achieving that, I think. It's parallel to the previous requirement that he has to post on VPR, but it no longer assumes that he will follow the rules of his own accord. And, by moving enforcement to AE, it makes the penalties more likely to stick if Δ violates the restrictions. — Carl ( CBM · talk) 16:50, 13 January 2012 (UTC)
Remedies:
I feel if we are to allow bots, that there should be a limit for at least a year to see how it works. Perhaps limit to three? The community sanctions would not apply to the bots, though the topic ban should still be in place. All forms of editing sanctions should come under ArbCom (which is what PR 1.5 is saying). SilkTork ✔Tea time 15:11, 12 January 2012 (UTC)
I've looked in the evidence page and don't see evidence for recent poor interactions. Would someone please supply such evidence, or we'll have to assume that Betacommand is responding within reasonable parameters. SilkTork ✔Tea time 15:30, 12 January 2012 (UTC)
Anything could be interesting when taken out of context. In reference to the "blinking text"; what actually happened is this:
There's the context. I find it interesting that Δ is taken to task for making some text red and blinking, yet Intoronto1125 is never remonstrated for calling Δ an idiot or for edit warring in direct violation of WP:NFCC #10c. -- Hammersoft ( talk) 15:32, 13 January 2012 (UTC)
The above example dif, and the explanation, and the discussion surrounding it have been useful. I have got a good feel for the situation, and it confirms other impressions I have picked up. Thank you. SilkTork ✔Tea time 17:02, 13 January 2012 (UTC)
Is this a joke? The case languished in the Workshop phase with little movement for just short of two months. TotientDragooned ( talk) 20:10, 12 January 2012 (UTC)
For whatever it is worth, I am very willing to mentor Betacommand through whatever sanction, short of a site ban the committee may willingly adopt. I am generally available, and have an established rapport with him, such that I am confident "intense mentorship" could succeed.
In offering this service, I am willing to accept a tandem block, if and whenever Betacommand may warrant one. I would agree to be bound by a clause equivocating his failure, as a failure of my own. I would like to establish advance benchmarks that Betacommand and I can work in good faith to achieve which would clearly enunciate the path to normal editing. It would be a great advantage to success if a streamlined process was in place to request and perhaps effect temporary modification along the way if and whenever circumstances may warrant. And I assure with the weight of promise and the bond of word that I will monitor Betacommand and maintain a record of every potentially relevant action, discussion, or third opinion sought to show, demonstrably, the proactive steps taken to achieve "productive editing rehabilitation" and analysis of progress made. Respectfully submitted with prayerful hope and regard. My76Strat ( talk) 16:33, 11 January 2012 (UTC)
The above offer as well as some recent edits to the workshop page may be impossible to consider for procedural errors, missed deadlines, or any other reason, that being very new, I would reasonable not know. I hope where I may have been out of process I haven't been out of line. If the burden of my being tardy is too great, please advise me that I am too late. If there is any reason otherwise that I should compile what I believe is compelling, please advise so that I may begin in earnest, the proper construction of a document I do wish to respectfully present. My76Strat ( talk) 06:25, 13 January 2012 (UTC)
Please don't do this. It's hard enough recruiting and keeping BAG members. Instructing us to wade into this field of figuring out what Betacommand can do with a bot, both in light of the community consensus and this case's restrictions, is simply sliding the headache to us. MBisanz talk 13:00, 13 January 2012 (UTC)
Like, say, running unauthorized tasks with the bot? Given the years of drama surrounding this, ArbCom needs to spell out the enforcement procedure, if any. And most importantly, who is authorized to make this determination: ArbCom, BAG, or any "uninvolved" admin? And what are the appeal venues? ASCIIn2Bme ( talk) 13:52, 13 January 2012 (UTC)
The main remedy currently passing almost passing is 6 (
3.3.6 Betacommand limited to BAG approved tasks). There is also a proposal to vacate current sanctions. (
3.1.6 Community sanctions vacated). Most of the tangle of previous sanctions would be superseded by 3.3.6.
However, the interaction sanctions stand apart; they would be voided, and not replaced or superseded. There is no finding currently proposed that identifies an improvement in the quality of Beta's interactions with other editors, nor am I aware of evidence that would support such a finding. Arbitrator AGK, above, offered that the interaction problems occurred around scripts, not bots. Arbitrator Silk Tork, two comments further down, offered above that he would prefer look at only current evidence from a bot account. Neither reason adequately supports an affirmative action to void interaction restrictions.
A remedy addressing the quality of Beta's interactions with other editors needs to be crafted (or continued). If, on the other hand, the arbitrators choose to eliminate sanctions related to Beta's interactions with other editors, they should do so explicitly by remedy, and such remedy should be supported by findings of fact. Jd2718 ( talk) 15:18, 14 January 2012 (UTC)
Since this was another editor restricted to use automation only in BAG-approved cases (and was mentioned by Kirill), the discussion surrounding the final block should be informative here: [41] ASCIIn2Bme ( talk) 15:27, 15 January 2012 (UTC)
I am afraid Remedy 7 ("Betacommmand restricted") will leave us in essentially the same position.
There is a small advantage that #7 would be enforced at AE instead of ANI, but overall #7 seems like a very small tweak to the existing sanctions. — Carl ( CBM · talk) 12:25, 16 January 2012 (UTC)
Actually, creating such lists could also be used to deal with Delta's communication problems. If we implement something along the lines of Remedy 6, then Delta could be put on a 0RR or 1RR restriction. Then he could run a bot to check on reversions of his bot edits and then post them on some file which others can deal with. You can also imagine a deadline system where Delta is allowed to revert the unreverted items that are still on the list after a week. The community would then have a week (or whatever other time frame is seen to be appropriate) to either declare the bot edits to be wrongly reverted or to be correctly reverted and then remove such items from the list.
Count Iblis (
talk) 14:00, 16 January 2012 (UTC)
I'm afraid this remedy is an invitation to a festival of wiki-lawyering and perhaps WP:TAGTEAMing. Besides not addressing the issue that was the proximate cause of this arbitration (no consensus on what pattern or similar edits means), what does "both Betacommand and the subsequent user will share editorial responsibility for the edits made in this fashion" mean in practice? If the edits prove controversial in the community, are both editors going to be blocked? ASCIIn2Bme ( talk) 15:58, 19 January 2012 (UTC)
The purpose and intent of proposed remedy 7 is just what I said in my support comment when I proposed it: I think that if Betacommand channels his automated or repeated edits into a list of edits that can be reviewed and implemented by another user, we keep the best of his contributions (identification of pages with problems and a method of addressing the problems) while eliminating the big problems that people have with his contributions (errors and occasional mis-prioritization of tasks). Proposed copyedits to address any perceived ambiguities in the wording or to simplify the interpretation of the remedy would be most welcome here.
I appreciate all the comments in this section (and everywhere else) but I must say I am puzzled by ASCIIn2Bme's remarks, including that "I don't want to speculate why this comes from the Arbitrator who drafted the decision in the Betacommand 2 case." Anyone is free to disagree with any proposal I make or even to try to convince me to modify or withdraw it; following such comments is part of my job. But if the implication is meant to be that I am intentionally trying to create confusion or to stoke drama rather than work toward a fair and productive solution to the case, that is highly offensive. Newyorkbrad ( talk) 00:37, 21 January 2012 (UTC)
I'm sorry that Newyorkbrad felt attacked, but if I were to speculate about his reasons for posting this remedy, it would certainly not include deliberately causing confusion. If I understood the history of the community restrictions following the Betacommand 2 case, they arose because that case did not address the matter effectively. ¶ Alas, the subsection of the community that usually deals with Δ can't agree about the meaning of soup. Whether this happens because it involves many programmers instead of LL.D.s, again I don't want to speculate. Nevertheless, I think there is rough consensus in the community that the restrictions need to be simplified. I don't think remedy 7 manages to do this, rather the opposite, so I expressed my frustration in that regard, perhaps not in the best way. Regards, ASCIIn2Bme ( talk) 05:03, 21 January 2012 (UTC)
In regards to Franamax's comment on the Arb who said "Thanks for your comments"; I think that must be me. I had proposed a clean start option, and had received a number of comments for that, mostly critical of the proposal, and at the way it was done. Franamax left a message for me, which I took to be a rhetorical question designed to make me reflect on my action, and not intended to be fully answered. I took the message in the same spirit as the other comments I had received, and as part of my learning curve, and I thanked Franamax. As regards answering the question why I "didn't place those ideas first at the /Workshop page?" The idea had been raised in a few places before I formally put it on the Proposed Decision page, including Δ's talkpage, this talkpage, and the Workshop page. I felt the proposal was worth serious consideration as I believed it was a viable solution (still do). I feel that all viable proposals are worth considering, even if rejected, as the reasons for rejection can be helpful to a fuller understanding of the issues, even if only to answer the question - "Why wasn't XYZ considered?". As for why the PD page rather than the Workshop page for the formal wording of the proposal. I think that was mainly because the PD page had been opened, and because other proposals had already been directly placed on there rather than being workshopped first. I hope that clarifies matters, but if not, I'm prepared to expand. SilkTork ✔Tea time 01:39, 24 January 2012 (UTC)
For #6, one small tweak is to allow Δ to revert his own edits or edits by one of his bots in any namespace. During the testing phase of bot development, edits must occasionally be reverted, and there is no reason the sanction should prevent him from doing that while it allows him to develop bots. — Carl ( CBM · talk) 14:17, 16 January 2012 (UTC)
The PD refers to Δ by his former username (except in FoF1). This seems to be rather unusual since the usual practice is to use the current username and not a previous one (see, e.g., the Abortion case in which User:Anythingyouwant is referred to as such and not by his previous username Ferrylodge, even though he had a previous arbcom case under that name.) It might be a good idea to either add a sentence in the FoF like "As Δ is best known under his previous username, we will refer to him as Betacommand in the remainder of this decision for ease of reference" or replace all the "Betacommand" with "Δ". T. Canens ( talk) 16:31, 16 January 2012 (UTC)
Would someone please explain why typing Δ's name (7 characters, code without spaces = & Delta ;) is such an onerous burden? His chosen username is shorter to type than many of the arbitrators but no one is proposing forcing them into "simpler" names. Further, it is impolite in my view to insist on using someone's former name. You wouldn't (I hope) refer to a trans woman by her birth name, nor to a divorced woman who chose to revert to her maiden name by her married name. Admittedly courtesy in the online world isn't the same as in face-to-face interactions but I think the general principle of a right to self-identify extends to respecting an editor's choice of username. Now I recognise that there were good reasons to use the Betacommand name in labelling this case as it follows the earlier ones but surely we can respect Δ sufficiently to refer to him by the name he has selected.
If ArbCom does choose to pass a remedy forcing a name change - and forcing an editor back to a name surrounded by controversy - I would like to see a principle, grounded in extant policy, passed that provides authority for this action. Aside from the points that Δ has made about SUL and Hammersoft has made about the explicit policy allowance for non-alphanumeric names, this has the potential to be seen as another example of ArbCom overreaching its mandate and authority, and for little gain that I can see. EdChem ( talk) 04:20, 17 January 2012 (UTC)
The substance of policy is beyond the purview of ArbCom - that is, ArbCom must follow policy, in this case WP:USERNAME#Usernames with non-Latin characters, not change it. Accordingly, this proposal is beyond ArbCom's authority. While I have no opinion on the case as a whole, this particular proposal is itself a WP:POINT violation. -- Philosopher Let us reason together. 06:01, 17 January 2012 (UTC)
There is the matter of the fact that Betacommand started editing under Δ in order to circumvent a block. In my opinion, this has less to do with "triangle guy" and the name he chooses, and more to do with the fact that he violated another policy in order to circumvent his block. Forcing him to go back to his old name is certainly reasonable in that light, but I think it is complete bull$#!t to make him change it just because it is hard to type. Our policy explicitly allows that. Buffs ( talk) 19:44, 17 January 2012 (UTC)
Just a note that a current AN/I thread (<permalink to archive>) is somewhat stalled in anticipation of a decision here. I'll draw attention to my proposed principle on automated/high-speed/scripted editing - I believe this should be held to a much higher standard. Even in normal course editing and review, it's difficult to check over another's edits when many changes are made at one time. When even one error is finally sorted out from the diff, this is a matter for concern and becomes subject to blanket reversion. When there are hundreds of these edits occurring in a short period of time, it becomes simply overwhelming to maintain any standard of quality. To me, the question arises why the humans who are working hard here should be expected to just give up and accept whatever the high-speed editors decide is the right thing to do, even when it's wrong. Franamax ( talk) 06:46, 24 January 2012 (UTC)
I think people mentioning WP:BRD completely misunderstand BRD if they are proposing applying it to bot edits. BRD means, for example, that if I'm considering a good-faith but potentially controversial edit to the Great Pyramids article that adds several paragraphs claiming (say) that the pyramids were actually built by penguins, I shouldn't agonize too much about its possible reception by other editors. I should just boldly make the edit. The rationale is that even if it took me hours to write the paragraphs, if the edit turns out to be unacceptable, another editor can look at my change for a few moments, decide it's not appropriate, and just click "undo" to revert it, so my bad edit hasn't burned much of anyone's time but my own. The whole basis of "the encyclopedia anyone can edit" is similar--we open ourselves up to (among other things) routine vandalism, on the theory that it's easy to revert.
With bots, it's the opposite. Someone rather easily launches a bot operation that makes 1000's of edits, and it's then extremely difficult for human editors to examine or undo them. Please see User talk:75.57.241.57/x (one of several pages related to a single incident) as an example of the level of effort that it takes to revert a bot operation gone wrong. Someone (I won't name him here since this was an isolated error by an otherwise good editor) got over-enthusiastic with Twinkle and protected several thousand templates without consensus. It took days of hassle including N hours of custom programming (by me, with the help of discussion from other editors that consumed a lot of their time as well), and more hours of drudge work by a helpful admin who did the unprotections, to reconstruct exactly what Twinkle had done to each template and undo it. I've been involved in some other messy high-volume cleanups as well, but most often what happens is that the bad edits simply never get cleaned up because it's too much hassle, and the encyclopedia suffers long-lasting damage.
TL;DR: BOLD is only for actions (like most single edits) that are easy to undo ("revert"). It should never apply to difficult-to-undo operations such as large clusters of edits (automated or otherwise), edits that make irreversible disclosures or provoke conflict, etc. Anyone who has never actually reverted a bot rampage themselves (which can mean having to write another bot to clean up after the first bot) shouldn't go around saying that it's easy or expect other people to do it as part of BRD. If you really think BRD applies, let's see you do the revert. 67.122.210.96 ( talk) 02:40, 25 January 2012 (UTC)
Would an arbitrator please indicate the evidentiary basis for FoF 6.1 - Ongoing communication problems? Thank you, -- Hammersoft ( talk) 19:57, 24 January 2012 (UTC)
I have a quibble with
Minor edit is Wikipedia terminology for edits like spelling corrections, that don't add or remove content. They are flagged as "m" in edit histories which stops them from showing up on some watchlists or searches. The controversies around Δ's editing are mostly about edits that generally do add or remove content (usually remove), so they don't qualify as "minor" in that sense. I could go with "insubstantial" (except when they are wrong), but then why bother making them? 67.122.210.96 ( talk) 07:26, 25 January 2012 (UTC)
moved from project page -- Alexandr Dmitri ( talk) 23:06, 25 January 2012 (UTC)
I use two non-public svn systems. I do not plan on changing that. I open my code up for those who I can trust and who I know will not abuse my tools. If anyone is interested in accessing my web based repo they should contact me. However as it stands I will not publicly release my code until such time as I have done extensive testing on it, and it has been proven stable over a very long period and that I have had time to build in the idiot proofing needed for such a public release. Since I rarely add in the needed fail-safes for people unfamiliar with my code, I rarely make anything public. ΔT The only constant 19:16, 25 January 2012 (UTC)
Δ, I'm puzzled--why is your code somehow more dangerous or abusable than the tons of other bot code already out there? Why would someone capable of modifying your code to do something vandalistic not be capable of writing their own vandalbot? The MW:API including the editing interface is reasonably well-documented after all. I agree that vandals shouldn't be allowed to run bots making 5000 edits per hour, but I see that possibility as basically a server bug rather than something enabled by any particular client code getting loose, so I'd support adding a server patch limiting edit rates to 100 edits/hour (or some comparable number) for any accounts without a bot flag. I do understand your feeling of sitting on crufty, unreleasable code (we all have code like that) but maybe you should take the view that long-running bot code shouldn't be deployed unless it's production-ready (i.e. cleaned up, documented, etc). For shorter term tasks, I'd consider it a reasonable compromise if some other editors (BAG members, say) were known to have private access to the code.
How does the following modification sound: code doesn't have be released for bots where the BRFA process establishes that the approval is for a non-recurring, one-off task that is expected to be completed (i.e. over and done with, no follow-ons expected) in 30 days or less from initial deployment. Longer-running bots are expected to have clean, documented, release-quality code in a public place. 67.122.210.96 ( talk) 00:30, 26 January 2012 (UTC)
It appears that the BAG-approved-only-tasks resolution is what is going to pass. I don't think that the open-source-only-tasks is going to pass, but I have no idea. How about a middle of the road solution then. Betacommand, is there any active/semi-active member of the Bot Approvals Group (there are eight active and six semi-active) that you would not trust in showing the code to? If not, find a secure way to make the code available to those people. To be brutally honest, I think that all of this concern over the code is nonsense, all we need to care about is the end product, but for the people who are concerned about the code, this seems like a good compromise. Sven Manguard Wha? 02:15, 26 January 2012 (UTC)
All,
I removed a comment as it insinuated that delta committed a crime or could possibly be fired from his job from misuse of company property (if proven true). Since this allegation had no link or backup to the claim, I removed it IAW WP:BLP and explained it on the user's talk page. It should also be noted that this comment was the user's SOLE contribution. Perhaps it isn't a WP:SOCK, but since the vast majority of WPians don't start at editing in ArbCom, I think I'm hearing a WP:DUCK quacking... Buffs ( talk) 21:05, 26 January 2012 (UTC)
OK, I guess I can go along with the idea that we don't really need to see this, er, polydimensional code. Per my "general discussion" workshop comment, program bugs per se haven't particularly been the issue in most of these incidents: the programs usually do work as intended, and the problems have mostly been in task selection, large editing runs without enough testing and feedback, problems dealing with other users, etc. It's not even that helpful to do anything sounding like pushing responsibility for Δ's software problems onto other users or expecting us to do the work of reviewing it. Δ, you've mentioned that you've given access to the code to a few other users. Would you be ok with saying who a couple of them are? That might be good enough for the auditing purposes discussed in the previous thread. 69.111.193.96 ( talk) 07:12, 27 January 2012 (UTC) (address changed)
Requiring that his bot be open source does not require that the framework is open source as well. This polydimensional framework has been in development for years. The polydimensional framework API should be stable. If it was cutting edge, there will be research papers about it already. The proposed remedy forces Betacommand to use a bot framework which does have a stable published API that permits linking. The only problem would be if the polydimensional framework API is a trade secret, and Betacommand decides that he doesnt want to write his bots (existing and future) on a bot framework which does have a published API that permits linking. Some bot operators have been able to develop closed-source cutting edge frameworks which are research (the Cobi vandal bots?), and we've seen the research papers come out of this. Sadly it isnt a good idea to let Betacommand continue to do this. John Vandenberg ( chat) 11:22, 27 January 2012 (UTC)
An editor "frustrating" others is not sufficient for an ArbCom sanction, e.g. admonishment. In general, Wikipedia discussions of users should be phrased in terms of behaviors, rather than in terms of emotional states of editors, e.g. "frustration", which are more subjective.
An editor violating behavioral policies, e.g., by failing to reply to reasonable requests for explanations of edits or constructively to engage in discussions, is a legitimate concern for ArbCom.
Thus, proposal 6.2 is superior to proposal 6.1. Kiefer. Wolfowitz 23:24, 26 January 2012 (UTC)
Re Kirill "I don't believe we have the authority to force a user to release material under a particular type of license": right, arbcom can't force the release of code. But it can certainly decline to let the bots keep running if Δ doesn't choose to release the code. Whether it's important in this particular situation is a different question, but IMHO it is a legitimately available remedy.
67.119.12.141 (
talk) 08:55, 27 January 2012 (UTC) Looks like Jclemens and Coucrelles already made this same point before I did.
67.119.12.141 (
talk) 20:43, 28 January 2012 (UTC)
This case has been dragging on for an excessive period of time now. At this stage, we have ArbCom agreeing via proposed principles that:
Ok, moving forward we have the following findings of fact being agreed upon;
In short; so far ArbCom hasn't identified the problem. If you can't identify the problem, tossing around remedies isn't going to solve anything. So not surprisingly, in its remedies ArbCom decides
And here's the kicker; 16 remedies haven't passed or have failed outright. No small wonder; again, if you can't identify and support with evidence the actual problem, you can't craft a remedy to fix it. 2 1/2 months ago I noted ArbCom's incompetence in the establishment of this case. Move forward to today, and the fumbling about is still happening.
Rather than focusing on remedies at this point, ArbCom needs to go back to the drawing board and try to ascertain what in fact the problem was when this all started.
Lost in all of this is the fact that ArbCom screwed up. See FoF 5. In July of 2009, ArbCom assigned two mentors to Δ (see unban decision), who failed in their appointed tasks, and neither of whom informed Δ they would no longer continue to mentor him. In fact a year after they'd been made mentors, they still had the mentorship note on the top of his talk page [58] that one of them had placed there the year before [59]. Worse, ArbCom didn't even take notice they were not receiving monthly reports. Where's the FoF that ArbCom screwed up in their self appointed duty? Where's the FoF that the mentors who had been assigned to Δ failed in their appointed duties?
Here's another good one. 7 January 2012, ArbCom proposes Δ be limited to WP:BAG approved tasks ( Remedy 6). Did they bother to ask BAG about this, or even notify them they were being included as a possible remedy? Nope. I had to notify them myself, five days later [60]. Of course when someone from BAG did show up less than a day later, they opposed it and begged them not to do it.
So in summary;
It's high time ArbCom took a good long look in the mirror.
I would really, really like some members of ArbCom to respond to these queries. -- Hammersoft ( talk) 19:07, 27 January 2012 (UTC)
ArbCom should recognize the very serious failure they made in not upholding its self assigned tasks they agreed to in the 2009 unban decision. In dropping the ball, they set the stage for a series of unfortunate events. Refusing to recognize their role in this failure is tantamount to allowing Δ to fail, and blaming him for the troubles directly descendant from their joint mismanagement of this case. Whether such recognition is made public is immaterial. But, subsequent to that recognition, ArbCom should allow themselves, the community, and Δ a do-over to the effect that they go back to their agreed upon resolution as of 11 July 2009. Thus, I propose Remedy 9;
9) ArbCom vacates the current Δ3 case, and reinstates its 11 July 2009 decision, including the assignment of two mentors. This set of restrictions is amended to permit operation of Δbot, and the development of code to be used by other people, recognizing that such other editors bear full responsibility for the execution of said code. Further, the edit throttle is set to the community sanctioned average of 4 edits per minute over any 10 minute time period.
I would sternly advise that any would-be mentors understand their roles are voluntary, but they must notify ArbCom and Δ if they withdraw from their duties. Additionally, ArbCom should sternly advise itself that it needs to maintain contact with the case and follow through with their role in it. -- Hammersoft ( talk) 19:07, 27 January 2012 (UTC)
I remain unconvinced that a second mentor is going to magically materialize just in time to resolve the case. I'm also not sure if ArbCom will allow My76Strat to mentor given that their view of things seems to be at odds with ArbCom's view of things; since a mentor's job is to fix things rather than sweep them under the rug, ArbCom might want someone who agrees with them about the existence of various problems. -- N Y Kevin @145, i.e. 02:28, 3 February 2012 (UTC)
AGK wrote above that "the committee is hopelessly divided in how to dispose of this case." Is that really true, though? Eight arbitrators have supported Remedy #6 already, which means it will pass with the support or formal abstention of just one of the four arbs who haven't voted on it yet. Newyorkbrad, Risker, Coren, and Mailer Diablo should weigh in on this remedy. A Stop at Willoughby ( talk) 19:39, 4 February 2012 (UTC)
Several people appear to be of the mistaken impression that Δ is not allowed to run bots or automated programs. This is in fact false.
If anyone has any evidence proving he is currently forbidden from running bots or automated processes, please do advise. I'm well aware of the edit throttle. That's not my point in this thread. -- Hammersoft ( talk) 19:59, 31 January 2012 (UTC)
I am hatting this off, in my capacity as case clerk, for the second time. Do not revert or continue this discussion here. |
---|
The following discussion has been closed. Please do not modify it. |
Hammersoft / Buffs dialogue
|
There has been very little activity on the proposed decision in the last week and I want to encourage the arbiters, especially those who have not yet voted on the decision to weigh in. Remedies 2.1 and 6 in particular are both only one vote away from passing. I do not think that spending additional time trying to craft an ideal compromise sanction is worthwhile at this point. Arbitration is not magic pixie dust™. The advantage of an Arb case over an ANI discussion is that one or the other can pass (even if it is 9-7) and we can move on from the status quo, which is widely considered unacceptable. Eluchil404 ( talk) 20:27, 5 February 2012 (UTC)
Informal tally. Feel free to format better and update
(2.1) (6) Kirill sup opp SirFozzie sup opp Jclemens sup opp Coren sup --- Fuchs 1st 2nd Davies sup opp Vandenberg sup sup Courcelles 2nd 1st Risker sup opp PhilKnight opp sup Casliber opp sup AGK opp sup Elen opp sup SilkTork opp sup Mailer sup --- Newyorkbrad opp opp
I think John Vandenberg should state a preference between 2.1 and 6. Right now that doesn't matter, but if another Arbitrator were to vote for (6), it would put that over the 9 votes bar as well. ASCIIn2Bme ( talk) 22:45, 5 February 2012 (UTC)
Mailer Diablo has now voted on (6). Only Coren is missing in action there. ASCIIn2Bme ( talk) 07:15, 7 February 2012 (UTC)
One more vote on remedy 6 could make a difference. Just a standard user requesting a hold on closure for an additional 24 hrs. Buffs ( talk) 21:53, 6 February 2012 (UTC)
@Jclemens: Are there prior cases where ArbCom violated its own stated principals in how to close a case? At the time your clerk said they were going to close the case, there were three net closes. We're now down to one net support to close. Your "advice" is chilling; in effect you're saying I am out of line for raising an issue with how ArbCom (in this case a clerk) is doing its business. I remind you that you operate at the behest of the community, not the other way around. This is no different than an administrator telling a non-administrator "I advise you to not tell me what to do". -- Hammersoft ( talk) 14:27, 8 February 2012 (UTC)
How about we have a calming effect by both sides disengaging? SirFozzie ( talk) 18:35, 10 February 2012 (UTC)
I propose Article 35 to an ArbCom code, specifically that 34 articles be developed to restructure ArbCom into an institutional element worthy to serve as a means of last resort. I am deeply concerned upon observing this process, my first; and respectfully advise the committee that I believe they, as a body, have failed their charter, seriously under-serving their largest constituent; Wikipedia herself! I'll leave the distinguished members to develop the other 34. My76Strat ( talk) 09:14, 7 February 2012 (UTC)
Hammersoft, if there really has been sockpuppetry in this case, why haven't you opened a WP:SPI (or if you did, where is it)? Have you informed the committee of this? Shall we believe you on a bare assertion alone? -- N Y Kevin @078, i.e. 00:52, 8 February 2012 (UTC)
NYKevin, I agree that a sock-puppetry investigation would be the most transparent venue in which to resolve concerns of case-page socking, but it is acceptable to e-mail an arbitrator privately; and moreover, such a solution could be more elegant. On a more general note, I don't think the issue of socking in this case is really worth significant focus, and I would prefer that we spend our time refining the proposed decision. AGK [•] 11:52, 8 February 2012 (UTC)
My76Strat ( talk) 10:23, 8 February 2012 (UTC)
@AGK: So the sockpuppetting is not worth significant focus? At what point does it become unacceptable for sockpuppets to abuse the ArbCom process? We've got IPs making unfounded accusations that Δ is responsible for a denial of service attack against another person involved in this case ( submitted 'evidence'). We've got IPs making an accusation of 112 edits in a ten minute period ( submitted 'evidence') on 18 May 2011 when in actuality there were three edits [69] in the stated time period. Oh but not to worry, this is all perfectly acceptable because we've got an ArbCom member using it as evidentiary basis. And really, it's ok for sockpuppets to submit evidence. Afterall, the 500 word limit is really just a minor irritant to those willing to sockpuppet to submit evidence. -- Hammersoft ( talk) 14:16, 8 February 2012 (UTC)
Hey, Hammersoft, put up or shut up. Post to WP:SPI or stop posting about this at all. Jtrainor ( talk) 14:21, 8 February 2012 (UTC)
HS, back to your query, I'm just not seeing any evidence of sockpuppetry. As for ArbCom allegedly using this sockpuppet's evidence, it only garnered a single vote, so it had no actual effect. Let's say for the sake of argument that there was a sockpuppet. It didn't have any real effect, so there's nothing to be done here. I concur that a SPI is probably a good route to follow for these accusations. If they were sockpuppeting, then they deserve to be held to account for their actions. Buffs ( talk) 15:09, 8 February 2012 (UTC)
The California ones are probably all me (my address changed several times during the case for technical reasons mostly outside my control). I have no idea about the TX one or the story with that "112 edits" diff (even after I spent a while trying to figure out what went wrong with it). Some of the other diffs got results with discrepancies from my own checks, maybe because of some difference in how things were calculated. It's also possible that article or image deletions caused some gaps in Δ's contrib history that made some of the discrepancies in the counts.
In my own submission, I posted a diff indicating Tristessa had been DDOS'd, but Hammersoft should note that I didn't make any claims of any sort about how it might have happened. I simply thought it was a weird enough incident that I decided to call it to arbcom's attention so they could check into it further if they decided it seemed relevant.
I don't know why Hammersoft is pursuing these conspiracy theories anyway. For quite a long time, Δ-related drama seemed to in part be proxy battles about NFCC enforcement, so it was impossible to say anything about the bots without appearing to take sides in the NFCC battle. (On NFCC philosophy I'm actually mostly on Hammersoft's side; I'd just prefer for WP to deal with NFCC by much different methods than Δ used). However, with Δ banned from the NFCC area, we're left with what I'd have thought was a less "political" issue of straightforward recurring bot misuse. So I don't understand why Hammersoft is so angry. 67.117.145.9 ( talk) 19:13, 8 February 2012 (UTC)
I did my own edit speed calculation and added it to evidence. I think the issue is not in any real doubt. 67.117.145.9 ( talk) 07:34, 9 February 2012 (UTC)
I've done some background reading on the MSK case, and this struck me as relevant here as well:
As a result of this history I believe the original mentorship was a failure, and that the Mentorship Committee is not up to the task of monitoring problem users. So far as I can tell, there are no follow-up procedures, no functioning communications, etc. Fundamental problems are that the job requires one to be a policeman/therapist/teacher, while most of us came here to be editors, and that it is a bureaucracy requiring procedures and maintenance. Those deficiencies should be addressed, in another forum, before new cases are assigned to it by the ArbCom or by goodwill ambassadors.
— User:Will Beback 08:30, 2 June 2006 (UTC)
Are there any cases where ArbCom-mandated mentorship has succeeded in the long run in reforming editors? ASCIIn2Bme ( talk) 09:48, 8 February 2012 (UTC)
http://wikipediareview.com/index.php?showtopic=34323&st=20 — Preceding unsigned comment added by 76.69.228.210 ( talk) 04:48, 11 February 2012 (UTC)
So, what the ArbCom is now doing, is first deciding to ban, but upon realising that they did not have a finding of fact supporting such a decision, they then find a FoF to vote upon so that they have a reason to ban? Do you guys have any clue how this is appearing now? -- Dirk Beetstra T C 04:19, 13 February 2012 (UTC)
What I want, is either that you, even if that takes months and months of deliberations, reach a proper decision on wiki. But here, you obviously first decided to ban, but did not have a significant FoF that would accomplish such a decision (1 is general, 2.2 and 4 are things which are by far not necessarily Betacommands fault if they are a fault at all - part of the failure is the community and even ArbCom's fault, 6.1 is the closest to finding a problem due to Δ himself - one that actually could have been solved (if it is not already solved), if the mentoring by ArbCom would not have failed. You used those FoF's to decide to ban (some even vote to close because there is a ban decision ..), and then it stopped for some time. Now, in a burst of a couple of hours, you see Arbs vote on a last FoF, the only one that actually defines a problem with Δ. Yes, my question remains: do you guys realize how this looks? Either it looks like you just decide to ban and then realize that there is no FoF actually sufficiently supporting that, or you fail to have enough time to actually look properly into cases. Yes, I want want proper decisions, just as what I expect of a judge deciding over life sentences (or worse). Yes, there will always be people unhappy, and yes, they will always make mistakes, that will happen in all processes. But if a judge decides such things, at least the severity of the punishment is in line with the severity of the crime. It is good to read that you, Jclemens, are completely comfortable with first deciding to ban ( diff), but only last night found reason (enough?) to do so:
... apparently not, because it takes then a couple of days of (off-wiki?) discussion (or you are just busy with another case or real life), and suddenly a realisation that the case needs to be closed (PhilKnight suggests that it needs a fast close because of time constraints ..), and that you actually need a proper reason, which, obviously, you did not have. And even this one (FoF 3.1), is citing the weasel word 'often' quite a number of times. Is 0.1% of the time over the speed limit 'often'? If the total average speed is less than 1 edit per 10 minutes over the last year then that suddenlty becomes 'a few'. One single case of incivility (where I can show cases of (mild) incivility against Δ by, low and behold, even an ArbCom member) - without even a timeframe, IIRC this restriction is there for 2-3 years? And a number of sub-optimal edits / mistakes (no violations not of the type of POV violations, CopyVio's etc.), and related to my last point, some cases of 'saving without reviewing' (I wonder how you can even prove such a thing, you must have been looking over his shoulder, or have a way of defining that edits must have been not properly reviewed). Those two points would be 'carelessness' .. at worst. And that in 10s of thousands of edits! And that is what you decide to ban over in the end? That is utterly out of perspective, and calling it 'ignoring the restrictions' is then a complete misinterpretation of the facts.
You consider this 'trying to find a middle ground where various options are explored and consensus is sought for a lesser remedy before banning is the default outcome'? No, this still looks like you were finding a way of banning Δ, not caring about the improvements he made (since last cases, or generally over time), looking at the accusations/evidence out of perspective (again, 30 speed violations in a year is less than 0.1% of the time, only one slight violation in the last 6 months, with an average editing speed which is significantly below 1 edit per 10 minutes (taking 20.000 edits in a year; far, far from the 40 per minute)).
I am convinced that there is a problem here, but your actions here certainly do not convince me that Δ is the (only) problem (IMHO, he is just a part of the problem - you may want to look at the current discussion about about how WP:AN/I works in this respect, and NFCC work in perspective of the SOPA/PIPA Wikipedia blackout), and whether this is the only or best solution. I remain, this ArbCom case is an utter failure, and this remedy is far, far too harsh for the passing finding of facts, especially before 3.1 passed. -- Dirk Beetstra T C 07:07, 13 February 2012 (UTC)
Silktork, you here say: "What is true is that the sanctions have failed, and without agreement on an alternative to the sanctions, then a ban is viable even without this finding." Without this finding, you present no evidence that there is any disruption, or that it is Δ's fault that the sanctions have failed (actually, ArbCom itself failed some of its own sanctions of the last case), or that Δ is actually the cause of the disruption. You basically suggest 'because we can't find anything else, we just ban Δ and we are ready.' Could you please elaborate on that remark? -- Dirk Beetstra T C 10:32, 13 February 2012 (UTC)
The point of the FoF is for the arbitrators to tell us why they voted on the ban. They are not required to give any such explanation, strictly speaking, but it is certainly a good practice for them to do so and I'm glad they are making the effort. A similar situation often happens in committee meetings in real life in my experience. After a long discussion, the committee may come to a decision and take a vote to record it. But if they have to also write a short paragraph expressing the official opinion of the committee about why they reached that decision, this paragraph is often harder than the decision itself, because different committee members may have different reasons, and not be willing to agree on the reasons even though they agree on the outcome. — Carl ( CBM · talk) 12:18, 13 February 2012 (UTC)
According to the motion to unban Δ [70]:
It seems like this is relevant to the current remedy. — Carl ( CBM · talk) 12:32, 13 February 2012 (UTC)
PhilKnight says "we have other cases waiting, and we need to close this in the near future". Why? Is there some deadline we are not aware of? Perhaps ArbCom isn't aware that Δ hasn't edited in more than a week. There is no ongoing disruption, no outstanding complaints, no questionable actions happening. Hell, he hasn't even edited mainspace in more than a month! But now we must rush to close this case? Why?
And still, as mentioned in various threads above we have serious problematic issues in play;
Not to worry though, just make veiled threats to not question ArbCom and its clerks, and then close the case without any basis in fact. Just sweep the whole unfortunate mess under the rug, and everything will quiet down in time.
ArbCom, we KNOW you're divided on this case, and from statements by various members of your body we know you are polarized by opinions you brought to the table when this case opened. But, we didn't elect you to adjudicate cases based on opinions of an editor you bring to the case before it opens. You are compelled to dispassionately review the evidence presented, and make findings based on that evidence. You aren't doing that. This case severely lacks any evidentiary basis from anything over the last six months to support findings against Δ, and the result of this case is to ban Δ from the project.
ArbCom, what are you going to do in a year when Δ appeals to be unbanned, and you fail to provide any evidence of behaviors he is supposed to correct that have not been corrected in the last six months? Hmm? What then? -- Hammersoft ( talk) 15:39, 13 February 2012 (UTC)
See [71]. So, the concerns that BAG isn't going to address issues with non-consensual edits by a bot after approval, but that they would punt them back to the community, seem valid. I floated an idea to SlikTork a while back about a "BotCom", but I guess we're far away from a consensus for an institution that is able to deal with bot-related disputes. ASCIIn2Bme ( talk) 21:20, 14 February 2012 (UTC)