Wikipedia:Arbitration/Requests/Case/Magioladitis 2/Workshop

Main case page (Talk) — Evidence (Talk) — Workshop (Talk) — Proposed decision (Talk)

Case clerk: Miniapolis (Talk) Drafting arbitrator: Callanecc (Talk)

Purpose of the workshop: The case Workshop exists so that parties to the case, other interested members of the community, and members of the Arbitration Committee can post possible components of the final decision for review and comment by others. Components proposed here may be general principles of site policy and procedure, findings of fact about the dispute, remedies to resolve the dispute, and arrangements for remedy enforcement. These are the four types of proposals that can be included in committee final decisions. There are also sections for analysis of /Evidence, and for general discussion of the case. Any user may edit this workshop page; please sign all posts and proposals. Arbitrators will place components they wish to propose be adopted into the final decision on the /Proposed decision page. Only Arbitrators and clerks may edit that page, for voting, clarification as well as implementation purposes.

Behaviour on this page: Arbitration case pages exist to assist the Arbitration Committee in arriving at fair, well-informed decisions. 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 or clerk, without further warning, by being banned from further participation in the case, or being blocked altogether. Behavior during a case may be considered by the committee in arriving at a final decision.

Motions and requests by the parties

edit

Template

edit

1)

Comment by Arbitrators:
Comment by parties:
Comment by others:

Template

edit

2)

Comment by Arbitrators:
Comment by parties:
Comment by others:

Template

edit

3)

Comment by Arbitrators:
Comment by parties:
Comment by others:


Proposed temporary injunctions

edit

Template

edit

1)

Comment by Arbitrators:
Comment by parties:
Comment by others:

Template

edit

2)

Comment by Arbitrators:
Comment by parties:
Comment by others:

Template

edit

3)

Comment by Arbitrators:
Comment by parties:
Comment by others:

Template

edit

4)

Comment by Arbitrators:
Comment by parties:
Comment by others:

Questions to the parties

edit
Arbitrators may ask questions of the parties in this section.

Questions from Mkdw

edit

The previous case revealed several dichotomies between policy and practice. In a number of remedies, ArbCom was not in a position to entirely resolve certain aspects of the disputed area and the community at large was asked to assist, particularly in regards to a clarification of policy. Since the previous case, I have the following questions:

  1. All: In the opinion of those involved, has the discussion at Wikipedia talk:WikiProject Check Wikipedia#Identifying cosmetic fixes succeeded in clarifying which fixes can be done alone or in concert with other changes?
  2. All: A March 2017 RFC made changes to WP:COSMETICBOT. It was supported by most parties involved. Has this clarified what constitutes as a cosmetic edit and which type of cosmetic (and/or bot) edits are restricted?
  3. All: The COSMETICBOT RFC was closed as having achieved consensus. Were these policy changes enacted fairly and with broad enough support? (Relatively to other bot policies)
  4. Magioladitis: What was the intention of creating WP:COSMETICEDIT? Why was it necessary following the changes to COSMETICBOT?
  5. All: WP:ADMINCOND and WP:ADMINACCT require administrators to exercise good judgement and communication. Does being the subject of community imposed restrictions, in any form, as well as receiving a recent block, put that administrator in contention with these two policy expectations for conduct?
  6. All: Why should BOTCOND exist (which does not)? By community practice, are the expectations on bot operators different than that of any other editor? Admins for example have increased expectations due to their access to tools. In terms of adhering to consensus, policies, restrictions, etc.?
  7. All: Were the restrictions placed against Magioladitis by the community enacted through broad enough consensus? (Relatively to ANI)
  8. All: Prior to Magioladitis' most recent community imposed restrictions, were there other bot operators conducting identical or similar types of edits (semi or fully automated) other than Magioladitis?

I have not had a chance to review the evidence filings that have already started. If some of these have been answered please point me to the relevant sections. Please also be as brief as possible in answering these questions (3 to 4 sentences each). Will ask follow up questions if required. Thanks, Mkdw talk 02:19, 24 July 2017 (UTC)[reply]

  • BOTCOND does not exist and it's unclear why it should or should not be created. A lot of discussion has been spent on whether these are 'bot operator conduct issues' or 'standard editor conduct issues'. If there are different expectations, I would want to see how these are laid out before determining if ADMINCOND has some jurisdiction. Mkdw talk 16:24, 24 July 2017 (UTC)[reply]
  • @Beetstra: Could you elaborate on what you meant by "I'll add to this that this is more a case of BOTCOND than of ADMINCOND"? I had assumed you meant an issue of bot operator conduct, but your answer below distinguishes BOTCOND and BOTOPCOND (something never previously brought up) as being distinctly two different things. What's makes this case more about BOTCOND (aka BOTPOL) than ADMINCOND? Mkdw talk 03:47, 25 July 2017 (UTC)[reply]
  • @Magioladitis: Will you be responding to the rest of the questions? Also, it seems some of your answers don't align with the question number. For example, answer #4 seems like it's addressing question #5. I don't want to make any assumptions. If you could let me know that would be great before I proceed further. Mkdw talk 16:11, 28 July 2017 (UTC)[reply]
Response from BU Rob13
BU Rob13's Response

In order:

  1. Unsure, but the RfC did.
  2. Yes.
    Noting also that no issues related to COSMETICBOT have been brought up with regard to behavior since the first case. The issues are unrelated. ~ Rob13Talk 19:21, 24 July 2017 (UTC)[reply]
  3. Yes. The RfC had a large participation relative to similar discussions.
  4. Skip.
  5. I'd usually say yes. Here, we also have to consider that this is a continuation of a pattern of behavior going back several years (examined in the first case).
  6. BOTCOND doesn't exist and I don't think it needs to. Note that WP:BOTREQUIRE, WP:BOTCOMM, WP:BOTISSUE, WP:MEATBOT, and WP:BOTASSIST exist already and do spell out some conduct expectations of bot operators and editors using scripts. Beyond what's already there, bots and semi-automated tools are capable of much higher levels of disruption than manual editing, holding all things equal, but we can sanction that the same as high levels of disruption from a manual editor. High disruption warrants more aggressive preventative sanctions no matter how the disruption occurs.
  7. Yes. The topic bans received almost universal support among a large number of uninvolved editors.
  8. It's complicated. Magioladitis' semi-automatic edits leading to his second topic ban mirrored the edits being made by PrimeBOT and Magic links bot. Bot-flagged edits can be hidden from watchlists, but AWB edits cannot without installing an external script. This watchlist spam was the problem. No other editors made similar edits without a bot flag. No other editors continued making edits after promising to stop without obtaining consensus.

Please let me know if you'd like me to elaborate on any. ~ Rob13Talk 04:42, 24 July 2017 (UTC)[reply]

@Mkdw: I've updated my answers above based on your BOTCOND clarification. Having said that, I personally believe almost everything having to do with bot policy is out-of-scope under the scope the Committee has set. Zero edits from Yobot are at issue, only edits from the main account. I could definitely see WP:BOTASSIST or WP:MEATBOT being relevant, but the rest of the bot policy doesn't apply to the edits at issue. ~ Rob13Talk 17:02, 24 July 2017 (UTC)[reply]
Response from Magioladitis
Magioladitis' Response
  1. Not quite. COSMETICBOT gives examples but does not conaint a full list. Moreover, some things seem to be described in the Manual fo Style but till now some people opposed them as sole tasks or most importantly we have not still found the correct procedure to do that. For example Remove/Fix invisible unicode characters from pages was denied while Remove/Fix invisible unicode characters from pages is above to get approval because a different method of discussion was used.
  2. The definition was refined to include examples of edits that do not affect the visual output and still considered not cosmetic. Still that does not the solve the problem of whether this refers only to bots, under which conditions the edits could be made, if they should be made at all, etc. At least this does not solve it in a centralised way. This causes in this task and other mini-debates of whether to implement secondary tasks or not. See also that this task took 4 months to get approval.
  3. The fact that an editor (Andy) already raised questions on the policy and they way the people reacted on this I thin shows the situation that has formed. The community is not fully aware of the new COSMETICBOT policy. I was never convinced that the old was ever really comprehended. This is the reason the policy changed so easily.
  4. A page about Bot policy should only contain things that are about bots. Since edits that do not change the visual output are done manually I think we need a definition for these edits. My intention was to create a info page showing that cosmetic edits are connected to bots but since they can also be done by editors when using a wikilink to a definition we should not directly address people directly to a bot's pages. I have claimed in the past that the term "cosmetic edit" has a meaning only between Wikipedians (mainly those with experience). So an info page would be useful. Especially, since the term is connected to a great group of Wikipedians, the wikignomes. Should a solution be to add a See alo to the COSMETICBOT o wikignomes? I am not sure. I think an info pages was/is a better solution.
  5. If this was true any sanction to any admin would directly lead to desysoping.
  6. I will be amazed if ADMINCOD is here to in fact disallow discussions on policies, implementation of policies, etc. Initiating discussion about the problems I described above don't contradict my judgment. I have not violated the bans. The bans were imposed for two different things. Their connection, as long as I understand it, is that Wikipedia lacks a strategy of dealing with two fractions:: Those who want to watchlist their pages to help the project and those who want to fix things per the Manual of Style to help the project. Promoting one group over the other is not a good solution. As long as we deny to find an optimal solution we will lose editors from both sides. We have lost many bot operators during these years.
  7. I like the idea. For example I don't like the idea that some AWB bot owners have redirected the bot's talk page to their own page. I think we have to formulate a universal strategy for bot operators. I have started a discussion even for asking AWB bot owners to be asked to enable general fixes. Till now these ideas were rejected. Still all these are just ideas. (Banning someone from throwing idea on the table does not make much sense to me.)
  8. I will skip this one. I am not the right person to ask this I guess.
  9. A lot of people perform semi-automated edits, a lot of people perform CHECKWIKI edits (recently we discovered that a task that is considered "purely cosmetic" that of removing duplicated categories is systematically done by editors manually), many editors edit in higher speeds. The edits I did recently do not fall in the term of "cosmetic", they have consensus, two bots are doing them but the bots were not fixing everything and I started fixing in addition to them. Many examples were given. Bgwhite has also received similar complains to me. Ladsgroup received complains too. In the past Rich, Lightmouse and perhaps others. From time to time many people receive complains. (See Primefak getting a complain even for their bot) -- Magioladitis (talk) 15:51, 24 July 2017 (UTC)[reply]

@Mkdw: I did. Please check because I am in the bus and I am working in a small screen. -- Magioladitis (talk) 16:39, 28 July 2017 (UTC)[reply]

Response from non-party Beetstra
Beetstra's Response
  1. No. See Wikipedia:Arbitration/Requests/Case/Magioladitis_2/Evidence#The community did not significantly participate.2C nor endorse the requested review on common fixes (analysis for counts)
  2. It did .. to the bot operators who performed the majority of the analysis (see Wikipedia:Arbitration/Requests/Case/Magioladitis 2/Evidence#The community did not significantly participate nor endorse the requested review on the policy on cosmetic edits; analysis for counts).
  3. I doubt it, this was mainly a bot-operators discussion - see Wikipedia:Arbitration/Requests/Case/Magioladitis 2/Evidence#The community did not significantly participate nor endorse the requested review on the policy on cosmetic edits (analysis for counts). It is however likely the most attended bot related RfC in 5 years. Maybe the community does not have a lot of influence on the bot policy in general.
  4. -
  5. NO, per WP:ADMINCOND. Arguments based on WP:ADMINACCT should be weighed against WP:CIR (I know, policy against essay).
  6. The WP:BOTPOL is WP:BOTCOND, I presume you mean WP:BOTOPCOND - you touch on one of the major issues in this case. You can split editors for this consideration in 4 groups:
    1. editors that do not use automation (AWB/JWB, etc.),
    2. editors that use automated edits (e.g. AWB), but do not operatate bot accounts,
    3. editors that run bots, and strictly do all automation there (minor group),
    4. editors that operate bot-accounts, and run scripts on their own account.
    what a bot operator does on the bot account is bound by WP:BOTPOL, similarly what is allowed using scripts like AWB on your main account is also under WP:BOTPOL. I don't think it is fair to hold editors who are bot-operators to a different standards then those that don't with respect with what they do on their own account: if a bot operator runs an AWB script on their own account, then it should not be judged differently just because they also operate a bot account.
  7. For the first topic ban (Wikipedia:Administrators' noticeboard/Archive290#Proposal to topic ban Magioladitis from COSMETICBOT-related discussions) I would say that also here there was a consensus that was influenced (lead?) by bot operators (see Wikipedia:Arbitration/Requests/Case/Magioladitis_2/Evidence#The_outcome_of_the_first_topic_ban_was_heavily_influenced_by_bot_operators). 7 of the 16 support !votes were cast by bot operators - the first 5 support !votes were cast by bot operators. The topic ban was enacted by another bot operator. (counts).
  8. The answer to your question is "yes", but with a strong emphasis on "similar type of edits".
Please let me know if I need to provide more data for the replies. --Dirk Beetstra T C 15:05, 24 July 2017 (UTC)[reply]
reply to request for clarification
  • @Mkdw: In the request statement BOTOPCOND would have been more correct indeed. Lets drop the non-existing shorthand: I assert that this is more a case of bot operator conduct than of admin conduct. It was not abuse of editing privileges related to the administrator status that brought him here, the use of advanced rights on his account is not questioned (except, but I think that is stretching, that WP:AWB can automatically be used by admins). --Dirk Beetstra T C 03:59, 25 July 2017 (UTC)[reply]
  • Of course, we do not have a 'bot operator conduct guideline/policy' ..
  • Regarding that this is more a case of BOTPOL - because of what I assert in my statement in the request phase of this case: this is the third (at least) bot operator that is here because of 'not listening to the community' - that is striking, the third. I think that that is reflected in the 7 remedies in the first case - those encourage community input on bot policy and CheckWiki. --Dirk Beetstra T C 04:21, 25 July 2017 (UTC)[reply]
Responses from non-party Headbomb
Headbomb's Response
  • 1: It greatly helped in clarifying which is which, compare [1] to the current version at WP:CWERRORS. In particular, the "Priority" column is now more consistently sortable, and we have the "Cosmetic" column. The discussion is however ongoing, and a couple of CHECKWIKI fixes are still unresolved. However, there are other fixes that can be made, but are not covered by CHECKWIKI. No such categorization is available at WP:GENFIXES for instance.
Concerning why Yobot_35 was declined, and why Yobot 55 is getting trial/approved, the difference is that an RFC on the specific question occured between Yobot 35 and Yobot 55. In the previous case, ARBCOM asked BAG to carefully review the scope of BRFAs, and this is exactly what happened.
  • 2. Absolutely yes. We now give details several examples to illustrate what are and aren't cosmetic edits. These are noted to be typical, and bot operators are still required to seek approval for non-cosmetic bot. For non-bots, we note "human editors may also wish to follow this guidance for the reasons given here, especially if making such changes on large scales." That link points to WP:MEATBOT, which states "... it is irrelevant whether high-speed or large-scale edits ... are actually being performed by a [bot or a human]. No matter the method, the disruptive editing must stop or the user may end up blocked."
  • 3. Yes, for instance see Are/should IPs be allowed to run bots? that took place in the same time period. Both the Cosmetic Bot RFC and the IP bot op RFC had about 22 participants. And going by !vote count, the cosmetic bot RFC had a lot more people giving clear opinions on it.
  • 4. Skip
  • 5. Well that's what the whole case is about. Having an ARBCOM case against them, or a topic ban following it could be worth desysopping over, but there's certainly a grey area. But when it comes to an ARBCOM case, plus multiple topic bans, with violations of those topic bans... Well at some point, that crosses WP:ADMINCOND.
  • 6. WP:BOTCOND / WP:BOTOPCOND don't particularly need to exist, since everything is already covered by WP:BOTPOL. For bots, we have WP:BOTREQUIRE, and for bot operators we have WP:BOTACC/WP:BOTCOMM/WP:BOTISSUE. WP:AWB also has rules of use at WP:AWBRULES. We also require bot operators to be "editors in good standing", and deny several tasks per WP:BOTNOTNOW when there are concerns that someone doesn't understand the community and its rules. As a long time veteran, bot operator, former BAG member, and AWB developper, Magioladitis is or ought to be keenly aware of all of these standards, both written and unwritten, and really has no excuses here.
  • 7. They're certainly comparable with any other topic ban imposed at ANI or elsewhere that I can remember.
  • 8. Yes (bots), and that is partly why the ISBN mess was problematic. We have User:Magic links bot, User:PrimeBOT, User:Yobot (operated by Magiolagitis), and my own User:CitationCleanerBot (undergoing trial) that are taking care of ISBN links. There was no need, or no perceived need, for Magioladitis to be making these edits. One could certainly imagine doing cornercases that bots can't handle manually, but most people prefer bots to do them instead because bots can easily be ignored in watchlists, and there's the restrictions against Magioladitis performing cosmetic edits. Now to the extent that these may or may not have been technically cosmetic, they are certainly non-urgent edits that most people would have preferred to be done by bots. So regardless of the noise surrounding the issue, the solution would have been obvious to everyone: Extend Yobot's code to handle those new corner cases, and get approval for the extended code, and only do things that truly cannot be handled by bots via your main account.

Headbomb {t · c · p · b} 15:57, 27 July 2017 (UTC)[reply]

Re to 173.228.123.121

If we take a global warming analogy, the analogy would be that there's a town hall happening, where experts get to make policy decision, asking for comments from the population, placing ads on the TV, on the radio, doing school announcements, ads in the newspaper. 200 people, mostly environmentalists show up. The policy gets drafted, sets emission standards in vehicles, and is sent to be ratified in parliament. The policy passes by an 80% support, 10% against, and 10% abstentions, with 95%+ support from environmental experts.

Then 2 months later, a guy gets arrested for rolling coal shows up in court, making arguments that the law is invalid, that it's unclear weather or not "emissions" cover rolling coal because it's an after-market modification, he's complaining he hadn't been consulted, that global warming is a hoax, and that it's all a conspiracy to kill jobs and take away freedom.

It's on that guy for not having gotten involved in the process that created the law because he didn't think it would affect him. It's not on experts and politicians for failing to force him to attend.

Response from uninvolved 173.228.123.121
173.228.123.121's Response

Note: I commented at some length in the first Magiolaiditis workshop but haven't been on Wikipedia much or followed developments since then, so will skip some questions for now.

1,2,3: No I don't think so: as usual with bot-related discussions, participation was dominated by bot operators. Headbomb wrote in his /Evidence post, "[t]hat non-technically oriented people didn't want to participate is on them". I very much disagree with that philosophy. It's like blaming global warming on the public which was mostly trying to get on with its own life, rather than calling it a failure of governance resulting from self-serving politicians being swayed by COI-soaked industry money etc.

5: It's evidence that the admin has a problem in the specific area of the restriction and should tread lightly in or avoid that area. If someone is under multiple unrelated restrictions, that suggests a problem in their whole approach to editing; but if they're under a restriction because they have a persistent weak spot in one area, they can still be good in others. I'm fine with the general concept of an admin being placed under a restriction while keeping their sysop bit. I'll stay silent on whether this specific case goes further into negative territory to the point where Mag's bit has to be reconsidered.

6: I believe high-speed editing (HSE) should be treated differently from and have different accountability standards than regular editing, and that editors should have to demonstrate general competence in editing before being allowed to engage in HSE (more discussion). I support Principle 11.1 from Betacommand 2. (Fwiw, Rich F thought that principle was ok in general but shouldn't have applied to Betacommand.[2] I don't agree with Rich on that.) I believe that:

  1. Wikipedia policy and DR should distinguish between HSE and non-HSE but not refer specifically to bots, since these cases are always unbelievably clouded by useless wikilawyering about what constitutes automation. What matters is whether the editor's eyes were open (i.e. they were working on the context of the article content) or closed when they made the edit. We should not care one whit about whether automation was used, and we should presumptively treat all editing above a given speed as eyes-closed HSE (MEATBOT), just as we now presumptively treat above 3 reverts as edit warring (the 3RR policy was highly controversial when it was first proposed and enacted).
  2. HSE should be enabled by a permission flag similar to rollback, that can be given by admin discretion subject to some guidelines, and can be withdrawn through admin discretion or ANI in case of disruptive use. That makes HSE an advanced permission partway between regular editing and adminship. The wiki software should limit the editing speed of non-HSE accounts to some number of pages per day, enough to handle most regular editing, though some productive (even non-automated) wikignomes might have to ask for and receive the HSE bit. Note: This is obviously not a proposal for an arb remedy, but rather just a viewpoint presented to help Arbcom weigh community sentiment in an area that it asked about.
  3. I believe the above would have prevented or simplified at least ten messy arbitration cases (Betacommand 1,2,3; MzMcBride 1,2; Date Delinking; Rich F; Magioladitis 1 and 2, OccultZone, and I'm sure several others that don't come to my mind right now) and would have been of no detriment to the project.

7,8: Don't know, will have to research.

173.228.123.121 (talk) 22:35, 30 July 2017 (UTC)[reply]

Responses from non-parties

Proposed final decision

edit

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.


Proposals by User:Magioladitis

edit

Proposed principles

edit

Thank system shows community appreciation of work

edit

1) Magioladiti's work is appreciated by part of the community. His work on ISBNs received a high number of thanks.

Comment by Arbitrators:
Comment by parties:
Headbomb I reworded. -- Magioladitis (talk) 07:11, 5 August 2017 (UTC)[reply]
Comment by others:
Strike "highly", and change that to change it to "some of " the community. As for a "high number of thanks", make enough edits, and 0.01% will become a "high number". I'd put my own 'thank' rate at roughly 1-in-25 edits when I do citation cleanup. And I don't receive complaints either. Headbomb {t · c · p · b} 14:13, 31 July 2017 (UTC)[reply]

2) The thank system community appreciation of editors' work

Comment by Arbitrators:
Comment by parties:
Comment by others:

3) Sending thanks publicly shows endorsement

Comment by Arbitrators:
Comment by parties:
Comment by others:
As a general remark, this shows one thing only: Someone, somewhere, clicked on "thank". It shows that some people appreciated the edit, for one reason or another. If you make thousands of edits, you're bound to get a few thanks. One also needs to take into account the complaints for those edit when judging general endorsement.. Headbomb {t · c · p · b} 14:13, 31 July 2017 (UTC)[reply]

Template

edit

2) {text of Proposed principle}

Comment by Arbitrators:
Comment by parties:
Comment by others:


Proposed findings of fact

edit

Developers are encouraged to find a way to hide certain edits from watchlists

edit

1) Wikipedia provides functionality to hide edits from registered accounts and bots. There is also a script to hide AWB edits. This gives an idea of how things could be implemented. Community is encouraged to create a better watchlist enviroment that will allow some edits to be hidden based on editors' preference.

Comment by Arbitrators:
Comment by parties:
This issue tried to be solved in various occassions. I discussed this with Reedy via Skype and in Wikimedia Hackathon. I also opened a discussion in Wikipedia. (See: Wikipedia:Arbitration/Requests/Case/Magioladitis_2/Evidence#Various_discussions_to_help_systematic_work_on_Wikipedia_started). It's interesting to see the discussion there. I believe there is a way to satisfy everyone. We don't want to disatisfy the people who use watchlists and we could provide them even more advanced filters to work with. I believe WMF programmers should also look at this. Claiming that hiding AWB edits is not a solution is nonsense the same way that someone claims that hiding bot edits or registered edits for New Page Patrollers is not helpful. Providing a better watchlist system will be a greta benefit for the Community. ight now the main reason that people complain on certain edits is the watchlist disruption. -- Magioladitis (talk) 20:34, 8 August 2017 (UTC)[reply]
@Headbomb: I am not discussing hiding AWB edits here. This is only an example of implementation. ArbCom can't take that specific decision. They can give a general guideline for a better watchlist system though to help those who use it become more efficient. -- Magioladitis (talk) 07:37, 12 August 2017 (UTC)[reply]
Headbomb makes the point I made months ago; hiding AWB edits is not a solution because AWB edits cannot be presumed ignorable. Many AWB edits are quite substantive and may change content on the page in non-trivial manners. I would never want to hide all AWB edits, and many of the editors who have complained have said the same. ~ Rob13Talk 17:15, 4 August 2017 (UTC)[reply]
Comment by others:
This sounds more like a remedy than an FoF, and I'm not sure what it's supposed to mean. I'd sure like a better way to distinguish bot from human edits than what we have now. I'm told that the bot assertion in watchlists isn't stored in the article history, but only in the recent changes table, which is only retained for 30 days. So there's no way to tell whether a 3 year old edit was made with a bot assertion. There are some Phabricator discussions addressing this, but I haven't looked at them yet. 173.228.123.121 (talk) 05:22, 1 August 2017 (UTC)[reply]
I agree that this is more a remedy than a FoF. It would be in line with the FoF in the first case, where the AWB developers were encouraged to implement a system to cope with problematic edits. --Dirk Beetstra T C 07:07, 1 August 2017 (UTC)[reply]
This is the same remedy as the first ARBCOM case. It's rather weird for Magioladitis to propose this as a finding of fact/remedy/whatever, given he's an AWB developper, and well aware of longstanding issues that would help with this such as T100443 and many others listed here and here that received no attentions from AWB devs. Headbomb {t · c · p · b} 12:50, 1 August 2017 (UTC)[reply]
The FoF has now changed, so some of those comments no longer apply. However, watchlist clogging is only one of the problematic aspects, not the only problematic aspect. T11790 will help with this once it's deployed, and while other technical solutions can be found to help lessen the burden of near-trivial edits on the community, it would still in no way invalidate the underlying concerns, or make compliance with policy optional. Headbomb {t · c · p · b} 18:46, 3 August 2017 (UTC)[reply]
Re the rewritten version: it would be great to have an idea how to hide the AWB edits. In the last case I learned that AWB had made around 110M edits at the time, but only 32M (iirc) had the string AWB in the edit summary. Any suggestions how to hide the other 68M? Also this still reads like a remedy. Maybe of relevance: see my chat with Xaoxflux of a few days ago[3] where I realized that recognizing bot edits in software amounted to implementing a Turing test on edits. I wrote up an amusing example here a while back. 173.228.123.121 (talk) 16:21, 4 August 2017 (UTC)[reply]
There is WP:HIDEAWB, and that script / watchlists could make use of T111663 down the line, if AWB tags its own edits. But as I've said elsewhere, even if the option to hide all AWB edits was somehow available, this is still no excuse for the original behaviour. Plus, unlike bots, AWB edits cannot be assumed to be ignorable by default, as they are not community vetted. Headbomb {t · c · p · b} 17:02, 4 August 2017 (UTC)[reply]
Thanks, the tag will help if it happens, though it would also be great if there was a tag or metadata flag for every edit received through the API (as opposed to the web UI). I don't use watchlists at all and rarely look at recent changes, so I'm mostly interested in ignoring bots when looking at article histories. 173.228.123.121 (talk) 19:16, 4 August 2017 (UTC)[reply]
@Magioladitis:, as I have explained to you before, THAT IS EXACTLY THE MINDSET THAT WP:HIDEAWB IS NOT MEANT TO ENCOURAGE. To quote "However I want to be clear that unlike approved bots, we absolutely cannot approach AWB disputes with a "if someone pollutes your watchlist with crap, just hide it" mentality, because a) AWB edits haven't been community-vetted b) even if there is a core of edits that are vetted, running custom scripts, and performing further manual cleanup is common with AWB." emphasis in original. To this you replied "[a] very nice point indeed". Are you now going back on your word? Do you now claim that AWB edits ought to be treated like bot edits and ought to be assumed ignorable, despite the fact that they are not community vetted? Headbomb {t · c · p · b} 23:27, 8 August 2017 (UTC)[reply]
Headbomb is right about the issues he discusses above, but I still would very much appreciate a reliable technical way to tell bot/script edits (AWB would be a very good start) from manual human edits. Preserving the bot assertion and noting the editing interface (web vs. API) in article histories would be helpful. 173.228.123.121 (talk) 03:54, 9 August 2017 (UTC)[reply]

Community is encouraged to find a method to perform maintance edits with minimum disruption

edit

2) Editing a page to match the Manual of Style is welcome. Still the community is encouraged to find a method to perform these edits with minimum disruption to the existing watchlist system.

Comment by Arbitrators:
Comment by parties:
I propose this but I am still not satisfied with the wording since I think the decision of something like that should be part of a Request for Comment discussion because I am confident the community will come with better ideas. Using watchlists to follow thousands of pages seems ineffective to me but since people still use it we have to do two things together: Perform the neccessary edits with no interuptions, minimise disruption for the people who use watchlists to follow a large amount of pages. I note that may people are new page patrollers but we do not requite a maximum amount of new pages created to assist them. -- Magioladitis (talk) 20:34, 8 August 2017 (UTC)[reply]
Comment by others:
I very much oppose using bots to enforce conformance with the MoS, especially for small issues. It's fine when human editors do it while actively working on an article's contents, or if they make occasional stylistic edits in articles they're not active in but that aren't part of a mass editing spree. WP:RFAR/DDL showed that bots combined with MOS obsession are a combustible enough mixture that those ingredients shouldn't be allowed near each other. 173.228.123.121 (talk) 04:19, 9 August 2017 (UTC)[reply]

BAG is encouraged to find a way to speed up BRFA process

edit

3) BRFAs often get backlog and BRFA is sometimes really slow making it inefficient.

Comment by Arbitrators:
Comment by parties:
Based in the evidence I gave. -- Magioladitis (talk) 20:32, 8 August 2017 (UTC)[reply]
This isn't accurate for any bot operator but Magioladitis. Yes, Magioladitis' bot tasks take a while to approve because he's convinced the majority of BAG that they aren't worth the trouble to deal with them or driven BAG members to outright recuse. Other bot operators that haven't had behavioral issues have had tasks approved rather quickly recently. ~ Rob13Talk 23:34, 8 August 2017 (UTC)[reply]
Comment by others:
You made like 20 BRFAs or whatever number it was all at once. So yes you were asked to slow down. The slowness of processing your BRFAs is caused by the issues brought forward during both arbcom cases and two topic bans that have followed, and the Magioladitis/Yobot fatigue it brought on BAG. Otherwise BRFAs get handled at their normal rate, which heavily depends on the bot/trial complexity, and the operator's good standing in the community. Headbomb {t · c · p · b} 23:15, 8 August 2017 (UTC)[reply]

Template

edit

4) {text of proposed finding of fact}

Comment by Arbitrators:
Comment by parties:
Comment by others:

Proposed remedies

edit

Note: All remedies that refer to a period of time, for example to a ban of X months or a revert parole of Y months, are to run concurrently unless otherwise stated.

BU Rob13 is prohibited from any interaction with user:Magioladitis

edit

1) BU Rob13 is banned from replying to messages to Magioladitis' talk page and vice versa.

Comment by Arbitrators:
Comment by parties:
In February 2017, Magioladitis asked for "Interaction near-zero" from me. As a courtesy to him, I agreed that I would refrain from speaking to him directly about issues with his edits where I felt he was violating community standards. As an alternative, he requested that I take him directly to ANI. Since then, I have edited his talk page only a handful of times, all when I did not believe I was being critical in the way he asked me not to be directly toward him.
  1. [4] I briefly clarified my opinion on something that I had been vocal about in the past. I avoided directly expressing my disagreement with these edits out of courtesy, instead opting to state that I did not defend them as I had similar previous edits.
  2. [5] I supported Magioladitis' solution to a perceived problem, defending it against others who disagreed with it.
  3. [6] [7] I asked for clarification about an edit Magioladitis had made, then thanked him when he provided it.
  4. [8] I asked Magioladitis to seek consensus for a type of edit which appeared to violate his restriction at AN, noting that I both supported the edit and would support him in such a discussion. I made very clear that I wasn't at all interested in litigating the violation, just in making sure he didn't have trouble down the road. (I later removed this because I realized ISBN linked to a different place than in the previous revision; this is the same edit Beetstra mistook as cosmetic-only earlier in this case, so it's an easy mistake to make.)
  5. [9] I helped an editor who was struggling with a feature in AWB. Someone asked me to take a look and help with this on IRC, if I recall correctly.
(Note that I also commented a couple times in a discussion about his block during the unblock request on his talk page; there was no better place to comment because unblock requests must occur on talk pages.) This is the total extent of my comments on Magioladitis' talk page over the course of 5 months. One contact per month is exceedingly normal for two active editors participating in the same area. I've been very careful never to come to Magioladitis' talk page with a negative request out of courtesy, even though that is most certainly not required of me. Further, note that I haven't started a single discussion about Magioladitis at any community venue or project space talk page before this case. Still, Magioladitis thinks we have far too much contact. Where is this contact coming from and why is it perceived negatively?

It's really simple. I'm a BAG member. I'm interested in bot policy. Every time Magioladitis spammed a discussion about COSMETICBOT or some other facet of the bot policy, I participated, as would be expected based on my interests. I've seen Magioladitis around quite infrequently since he was topic banned from discussing COSMETICBOT because the discussion spam stopped. Magioladitis doesn't want me to stop interacting with him when it benefits him: [10] [11] [12] He just wants me to stop opposing his endless proposals to weaken the bot policy and stop voicing my opinion when his behavior is addressed in community venues I have watchlisted. It's about silencing a critic. ~ Rob13Talk 18:02, 28 July 2017 (UTC)[reply]

Please provide a diff that shows I gave "false advice on bots" or strike the personal attack. As for the constant claim that I refused to handle his BRFAs, this is correct. I recused from handling his bot tasks per WP:INVOLVED: [13]. I publicly stated support for that particular task multiple times, but I can't pick and choose when I abide by my recusal. I've explained this to Magioladitis in the past (see second diff, for instance). [14] [15] ~ Rob13Talk 18:19, 28 July 2017 (UTC)[reply]
As an aside, Headbomb mentioned a one-way interaction ban on Magioladitis as a possibility. I'd like to very strongly state my opposition to that. I have a philosophical problem with interaction bans in general, but one-way interaction bans are particularly bad. They create a no-win situation for all involved, since they more-or-less allow the non-banned editor to pseudo-topic ban the other from any discussion they decide to participate in. I do want the harassment to stop, but I'd be a hypocrite if I went along with that particular instrument being used to stop it. ~ Rob13Talk 21:23, 28 July 2017 (UTC)[reply]
I provided evidence in the Evidencee subpage. BU Rob13 is abusing the process here by provigin Evidence diffs and text as comments. -- Magioladitis (talk) 16:23, 11 August 2017 (UTC)[reply]
Comment by others:
I see no evidence of this being warranted. What can at best be summed up as one-way paranoia is not a reason to impose an interaction ban on the recipient of that paranoia. BU Rob 13 has been cordial in all his recent interactions with Magioladitis (peak 'non-cordiality' happened during the ARBCOM/Magioladitis 1 proceedings, but has greatly subsided after that), but he's certainly the one Magioladitis focuses on the most, for some unknown reason. The first ARBCOM case was started by Ramaksoud2000 (talk · contribs), the first topic ban was initiated by myself, the second topic ban initiated by Xaosflux (talk · contribs). Several BAG members, of which BU Rob 13 is a member of, have interacted with Magioladitis in much harsher terms. In the interest of avoiding drama and ensuring fairness, BU Rob 13 has recused himself from approving Magioladitis-operated and Checkwiki-related bots in his RfBAG. You can't ask him to approve your bots, then turn around accuse him of opposing them / hounding you when he expressed support but remains recused.
The rest of Magioladitis claims', like BU Rob 13 gives "false/bad advice on bots" are blatant misrepresentation of facts (see what actually happened). If anything, the interaction ban should be on Magioladitis casting bad faith aspersions against BAG members, and BU Rob13 in particular. But at this point, an admonishment/reminder to WP:AGF would be a more measured response, than an interaction ban. Headbomb {t · c · p · b} 20:37, 28 July 2017 (UTC)[reply]

Proposed enforcement

edit

Template

edit

1) {text of proposed enforcement}

Comment by Arbitrators:
Comment by parties:
Comment by others:

Template

edit

2) {text of proposed enforcement}

Comment by Arbitrators:
Comment by parties:
Comment by others:

Proposals by User:Headbomb

edit

Proposed principles

edit

Decisions should be based on concerns about Magioladitis, and actions and responses from Magioladitis

edit
This is not something ARBCOM needs to vote on per se, but it should be read.
0) Sometimes allies do more damage to a cause than they help with it. The concept of a useful fool sort of illustrates this. Alice who would benefit from A, is brought in by Bob who opposes A because Alice can be used to argue against A without her realizing it.

I feel User:Beetstra has been so clueless about bot policy, so argumentative, and so downright poisonous to the whole case that I would feel bad if Magioladitis was judged based on discussions initiated by Beetstra, or on the resulting arguments and general clusterfuck. Sanctions and remedies concerning Magioladitis should be based on concerns about Magioladitis, and actions and responses from Magioladitis. I don't think ARBCOM would punish one editor because one of their 'allies' was bringing more heat than light on an issue, but I do think this reminder is needed for everyone that judging someone based on other people's action isn't a good thing.

Comment by Arbitrators:
Comment by parties:
@Headbomb and Hasteur: As a point of practicality, the Committee will neither sanction a non-party nor add a party to the case this late in the game. I agree with the sentiments and general chagrin that this case devolved into trying to invalidate every technical policy we have on-wiki, but the arbitrators do not need an admonishment to make clear they aren't buying it. A final decision that focuses on the actual problems, not the imagined ones, will do. ~ Rob13Talk 21:55, 17 August 2017 (UTC)[reply]
Comment by others:
Concur to the point of wondering if Beetstra deserves a warning or admonishment for generally disrupting the proceedings by attempting to point blame everywhere and elsehwere besides the locus of the case. Hasteur (talk) 19:36, 17 August 2017 (UTC)[reply]
I'd support such a proposal. I'm not really sure on what the exact phrasing would be, so I'll let someone else draft it, but I'm definitely behind the sentiment. Headbomb {t · c · p · b} 19:40, 17 August 2017 (UTC)[reply]

Edits made on a large scale require consensus, regardless of the method by which they are made

edit
Adapted from WP:MEATBOT

1) For the purpose of dispute resolution, it is irrelevant whether high-speed or large-scale edits that involve errors an attentive human would not make are actually being performed by a bot, by a human assisted by a script, or even by a human without any programmatic assistance. No matter the method, the disruptive editing must stop or the user may end up blocked. Merely editing quickly, particularly for a short time, is not by itself disruptive.

Comment by Arbitrators:
Comment by parties:
Yes. -- Magioladitis (talk) 09:33, 31 July 2017 (UTC)[reply]
Comment by others:
I like this a lot, especially the emphasis on "irrelevant". And I still support objective speed limits (e.g. 20/pages/day) if someone is restricted as bot remedy, instead of writing remedies that create murky questions of how the edits are being done. 173.228.123.121 (talk) 23:04, 30 July 2017 (UTC)[reply]
This does look useful. We consistently talk about focusing on the edits not the editor, so focusing on the edits not how the edits were made is a logical extension of that principle. Thryduulf (talk) 08:24, 31 July 2017 (UTC)[reply]
I completely agree. --Dirk Beetstra T C 09:30, 31 July 2017 (UTC)[reply]

Cosmetic and minor edits can be controversial, and require consensus to be done on a large scale

edit
Adapted from WP:COSMETICBOT

2) Cosmetic changes to the wikitext are sometimes the most controversial, either in themselves or because they clutter page histories, watchlists, and/or the recent changes feed with edits that are not worth the time spent reviewing them. When done by bots, such changes should not usually be done on their own, but may be allowed in an edit that also includes a substantive change. Changes that are typically considered substantive affect something visible to readers and consumers of Wikipedia, while changes that do not are typically considered cosmetic. Minor edits are not usually considered cosmetic but still need consensus to be done by bots.

While WP:COSMETICBOT applies only to bots, human editors may also wish to follow this guidance for the reasons given above, especially if making such changes on large scales.

Comment by Arbitrators:
Comment by parties:
It's not, actually, Headbomb. Those edits weren't cosmetic-only; they introduced a link to ISBN. The only issue was disruptively making them over the many objections of others without seeking consensus. ~ Rob13Talk 15:06, 31 July 2017 (UTC)[reply]
Not quite sure what exactly you're objecting to here. The point is that both cosmetic and minor edits require consensus. Regardless of whether changing the magic word version to the template version is cosmetic or minor, consensus is still needed. Headbomb {t · c · p · b} 15:33, 31 July 2017 (UTC)[reply]
Comment by others:
Irrelevant, Magioladitis did not perform an excessive number of cosmetic edits, and that is not questioned here. The first case remedies asked the community to review WP:COSMETICBOT, and Magioladits was commenting a lot on that, that was what got him topic banned - not that he was performing them on purpose. --Dirk Beetstra T C 09:32, 31 July 2017 (UTC)[reply]
Most definitely relevant, as this is [in part] what lead to the second topic ban. Headbomb {t · c · p · b} 14:17, 31 July 2017 (UTC)[reply]
No, that is not what lead to the second ban - what lead to the second ban was the 'high speed replacement of ISBN magic words'. Not whether those were cosmetic, nor whether they were minor. --Dirk Beetstra T C 18:16, 31 July 2017 (UTC)[reply]
As has been explained multiple times to you, editing rate was not the concern, the bot-like flooding of watchlists was. Headbomb {t · c · p · b} 18:18, 31 July 2017 (UTC)[reply]
Yes, I know that is the circular reasoning that we constantly have .. but it had nothing to do with cosmetic or minor, and that is what we discuss in this principle. --Dirk Beetstra T C 18:21, 31 July 2017 (UTC)[reply]
It does, because both cosmetic and minor of edits require consensus to be done on a large scale, and this details the reasons for why this is the case. Headbomb {t · c · p · b} 12:36, 1 August 2017 (UTC)[reply]
You see, circular. Headbomb, Magioladitis was in the timeframe between the two cases banned from doing cosmetic edits. If he would have done cosmetic edits we would not be here. There were no problems with minor edits either (well, the edits were marked minor). But whether they are cosmetic or minor is not to the point. If I go around doing 20.000 major edits using AWB then that qualifies as large scale, and if I do 100 major edits in 10 minutes, then that qualifies as high speed. Where you guys seem to hint at, is that anything that is large scale, or is high speed needs consensus to be run (and then, more so if they are minor or cosmetic). --Dirk Beetstra T C 13:28, 1 August 2017 (UTC)[reply]
You are actually saying that in the above principle, and I agree with that one. --Dirk Beetstra T C 13:30, 1 August 2017 (UTC)[reply]
"anything that is large scale [...] consensus to be run" indeed that's exactly what I'm saying. And cosmetic/minor edits are no exceptions. Headbomb {t · c · p · b} 15:41, 1 August 2017 (UTC)[reply]
Yes, that is exactly what you are saying in the above principle. --Dirk Beetstra T C 15:17, 2 August 2017 (UTC)[reply]
There's a weird concept of consensus that sometimes comes up in these discussions, maybe more often in connection with the WP:MOS than with bots, but similar. Basically if 5 people on a bot operator page or some obscure WP:MOS talk page discuss a proposal to do X, and they agree 5-0 that X is good so they start doing it, and then 10 people who notice it after the fact start opposing it at ANI, that's not 10 people acting against consensus and it's not consensus changing. It shows there's actually a 10-5 majority against X and there was never real consensus for X to begin with. There might have appeared to be one, because of the confounding factor implicit in where the original measurement took place. But the biased sample didn't actually reflect the (latent) sense of the wiki, because it was concentrated in a place mostly habituated by bot ops, MOS zealots, or whatever. Silence of others is not assent, especially if the person speaks up afterwards. 173.228.123.121 (talk) 08:08, 8 August 2017 (UTC)[reply]

The distinction between whether edits are cosmetic or minor is ultimately irrelevant

edit
Adapted from Wikipedia:COSMETICBOT, WP:MEATBOT

3) While as a rule of thumb, cosmetic edits should not be done on a large scale, consensus can create exceptions for particular cosmetic edits. Minor edits are not usually considered cosmetic but still need consensus to be done on a large scale. Both types of edits require consensus to be done on a large scale.

Comment by Arbitrators:
Comment by parties:
Comment by others:
I think this boils down to, "If someone, in good faith, believes that consensus should be obtained before making a given (type of) edit on a large scale, do not make that (type of) edit without first obtaining consensus". Anyone repeatedly ignoring this is liable to be censured. Thryduulf (talk) 08:28, 31 July 2017 (UTC)[reply]
I don't see how this principle is relevant. --Dirk Beetstra T C 09:36, 31 July 2017 (UTC)[reply]
Because a good chunk of the dispute is focused on whether somethings are cosmetic or not. And the point is that it doesn't matter, because both types of edits require consensus anyway. Headbomb {t · c · p · b} 14:19, 31 July 2017 (UTC)[reply]
No, there is no dispute whether his edits were cosmetic. If his edits were cosmetic we would not be here. --Dirk Beetstra T C 18:22, 31 July 2017 (UTC)[reply]

How large-scale edits are made matters

edit
Adapted from WP:COSMETICBOT, WP:BOTFLAG, WP:HIDEBOTS, Help:Watchlist#Options

4) Edits performed on a large scale can clutter page histories, watchlists, and/or the recent changes feed with edits that are not worth the time spent reviewing them. Both minor manual edits and bot-flagged edits can be hidden from watchlists, but as minor edits are not community-vetted, they cannot be assumed to actually be harmless/desired/actually minor. The WP:BOTFLAG is designed to mark edits as community-vetted, which will both hide them from the recent change feed, and mark them as general safe to ignore on watchlists. For this reason, many editors choose to hide bot edits on their watchlists, but nonetheless display minor edits.

Comment by Arbitrators:
Comment by parties:
Comment by others:
I agree with this in essence - though cluttering is a very subjective term. (User:Headbomb, you have a typo: '.. many editors choose .. '. --Dirk Beetstra T C 18:28, 31 July 2017 (UTC)[reply]
It's not just about watchlists, but the entire article history forever. Also the bot assertion is only kept for 30 days. There's no way to hide the bot-flagged edits when looking at years of edit history of an article. Alternatively, if you're actively editing an article, you might want to look at all the recent edits just in case a bot made a mistake. So that's more reviewing time looking at often-gratuitous changes. Here is an example that I wrote up last time. 173.228.123.121 (talk) 16:20, 1 August 2017 (UTC)[reply]

Administrators are not expected to be perfect, but sustained or serious disruption of Wikipedia is incompatible with adminship

edit
Adapted from WP:ADMINCOND

5) Administrators are expected to lead by example and to behave in a respectful, civil manner in their interactions with others. Administrators are expected to follow Wikipedia policies and to perform their duties to the best of their abilities. Occasional mistakes are entirely compatible with adminship; administrators are not expected to be perfect. However, sustained or serious disruption of Wikipedia is incompatible with the expectations and responsibilities of administrators, and consistent or egregious poor judgment may result in the removal of administrator tools.

Comment by Arbitrators:
Comment by parties:
Comment by others:
Yes. It is not relevant whether the disruption or poor judgement is related to actions that require administrator tools or not - if you are not meeting the expectations of an admin then you should not be an administrator. Thryduulf (talk) 08:30, 31 July 2017 (UTC)[reply]
Standard. --Dirk Beetstra T C 09:37, 31 July 2017 (UTC)[reply]

Proposed findings of fact

edit

Magioladitis is one of the AWB developers

edit

1) Magioladitis is one of the AWB developers, and has a developer-level permission to edit with AWB, which supercedes admin-level access, or the whitelist.

Comment by Arbitrators:
Comment by parties:
Comment by others:

Desysopping will not alter or remove Magioladitis' AWB developer status

edit

1) While desysopping to remove AWB access could normally be done for most admins, this is not the case here. If desysopped, Magioladitis' AWB developer status would remain unaffected.

Comment by Arbitrators:
Comment by parties:
True, but not relevant to the arguments being made. WP:ADMINCOND is not related to AWB access, and personally I think Magioladitis has lost the trust of the community (see both topic ban discussions and the opening statements in this case for numerous examples of this). ~ Rob13Talk 15:45, 3 August 2017 (UTC)[reply]
Comment by others:
This is mostly to note that if Magioladitis is desysopped, it has to be for reasons OTHER than revoking AWB access. I don't particularly care to see it formally adopted, but ARBCOM does need to be aware won't change the AWB situation. Headbomb {t · c · p · b} 16:27, 3 August 2017 (UTC)[reply]

Template

edit

1) {text of proposed finding of fact}

Comment by Arbitrators:
Comment by parties:
Comment by others:

Proposed remedies

edit

Note: All remedies that refer to a period of time, for example to a ban of X months or a revert parole of Y months, are to run concurrently unless otherwise stated.

Magioladitis encouraged to address longstanding AWB issues

edit

1) Magioladitis is encouraged to address longstanding AWB issues (e.g. these and many of these) that would make it easier for both bots and non-bots to avoid inconsequential or minor edits (e.g. T138977), as well as provide increased clarity in edit summaries (e.g T161467).

Comment by Arbitrators:
Comment by parties:
Comment by others:

Magioladitis is to seek consensus before taking on large scale semi-automated tasks

edit

2) Magioladitis is to seek explicit consensus from the community before making any semi-automated editing. This can be done in the following ways

  1. Any WP:CWERRORS with a priority of "Mid" or higher is assumed to have consensus unless objections arise.
  2. On his own talk page, by getting the thumbs up from a WP:BAG member directly. (Limit: 7 such requests per week)
  3. With a 7-day notice at WP:BOTN, getting the thumbs up from at least 2 BAG members. (Limit: 1 such request per week)
  4. Via a formal WP:RFC at the WP:VP. (Limit: 1 such request per month)

If the edits draw complaints, Magioladitis must immediately stop his semi-automated editing until given the thumbs up to resume by BAG. At any point in the process, BAG may ask for a stronger consensus to be established.

This restriction is in addition to any other existing restriction.

Comment by Arbitrators:
Comment by parties:
I do not see this as feasible due to the number of involved BAG members. I don't want to risk burning out the 1-2 uninvolved admins with requests like these. Personally, I think the community has been handling the behavioral issues properly and should be allowed to continue doing so. The one aspect of the dispute that the community can't handle – the sysop flag – is what the Committee should focus on. ~ Rob13Talk 15:47, 3 August 2017 (UTC)[reply]
Comment by others:
I don't think WP:INVOLVED is too much of a concern here. Most of us don't deal with Magioladitis because we're just tired/fed up of the drama because it seems our advice is mostly ignored, more than a case of not being able to be impartial. If Magioladitis can get one (or two) BAG members to give him the thumbs up when most of us are annoyed, I'm pretty sure that would curb a lot of the problematic behaviour. As for the 'burning out', this is why I put limits on the number of request that can be made. I'm certainly open to tweaking this, but I think the general idea has merit. Headbomb {t · c · p · b} 16:38, 3 August 2017 (UTC)[reply]
Pseudo-edit, added that WP:CWERRORS of mid priory/higher can be assumed to have consensus. Headbomb {t · c · p · b} 16:38, 3 August 2017 (UTC)[reply]

Proposed enforcement

edit

Template

edit

1) {text of proposed enforcement}

Comment by Arbitrators:
Comment by parties:
Comment by others:

Template

edit

2) {text of proposed enforcement}

Comment by Arbitrators:
Comment by parties:
Comment by others:

Proposals by User:BU Rob13

edit

Proposed principles

edit

Purpose of Wikipedia

edit

1) The purpose of Wikipedia is to create a high-quality, free-content encyclopedia in an atmosphere of camaraderie and mutual respect among contributors. Contributors whose actions are detrimental to that goal may be asked to refrain from them, even when these actions are undertaken in good faith; and good faith actions, where disruptive, may still be sanctioned.

Comment by Arbitrators:
Comment by parties:
From Wikipedia:Arbitration/Requests/Case/The Rambling Man. Note many of my proposed principles mirror those I proposed in the first case. ~ Rob13Talk 04:46, 20 August 2017 (UTC)[reply]
Comment by others:

Consensus

edit

2) Disagreements concerning patterns of edits are to be resolved by seeking to build consensus through the use of polite discussion – involving the wider community, if necessary. The dispute resolution process is designed to assist consensus-building when normal talk page communication has not worked. When there is a good-faith dispute, editors are expected to participate in the consensus-building process and to carefully consider other editors' views, rather than simply continuing to edit over objections. Continuing to edit over objections without seeking consensus is often disruptive.

Comment by Arbitrators:
Comment by parties:
Modified from Wikipedia:Arbitration/Requests/Case/American politics 2. ~ Rob13Talk 04:46, 20 August 2017 (UTC)[reply]
Comment by others:

Administrator conduct

edit

3) Administrators are expected to lead by example and to behave in a respectful, civil manner in their interactions with others. Administrators are expected to follow Wikipedia policies and to perform their duties to the best of their abilities. Occasional mistakes are entirely compatible with adminship; administrators are not expected to be perfect. However, sustained or serious disruption of Wikipedia is incompatible with the status of administrator, and consistently or egregiously poor judgment may result in the removal of administrator status.

Comment by Arbitrators:
Comment by parties:
From Wikipedia:Arbitration/Requests/Case/The Rambling Man. ~ Rob13Talk 04:46, 20 August 2017 (UTC)[reply]
Comment by others:

Role of the Arbitration Committee

edit

4) The Arbitration Committee may not create, alter, or destroy policies or guidelines established by community consensus. The community may alter existing policies as laid out in the procedural policy for doing so.

Comment by Arbitrators:
Comment by parties:
Originally written by me at Wikipedia:Arbitration/Requests/Case/Magioladitis/Workshop, but this is far more important this time around. Dirk's contributions in Evidence and Workshop in particular seriously confused the role of ArbCom, and a principle clearly stating that the Committee will not destroy a policy is warranted. ~ Rob13Talk 04:46, 20 August 2017 (UTC)[reply]
Comment by others:

Proposed findings of fact

edit

COSMETICBOT

edit

1) Following the Committee's recommendation to clarify the WP:COSMETICBOT policy in the previous case, the community discussed changes to the wording in a well-publicized RfC and found consensus for the new wording. [16]

Comment by Arbitrators:
Comment by parties:
Well, next time I'll be sure to kidnap editors who don't care about technical matters in the dead of night and force them to comment, Dirk. ~ Rob13Talk 14:46, 1 August 2017 (UTC)[reply]
A local consensus advertised at WP:CENT? That's one of the genuinely funniest things I've heard in a long, long time. ~ Rob13Talk 18:04, 1 August 2017 (UTC)[reply]
And there it is. It's no longer a slippery slope argument; Dirk is literally stating that the bot policy does not have broad community consensus. This is being used to argue that bot operators should be able to do whatever they please, with no apparent road to community oversight. The bar that Dirk is attempting to set for a sufficiently "broad" consensus to provide any oversight to bot operators is quite literally unattainable, as even an RfC advertised at CENT, the village pump, and in various technical-related areas isn't "broad enough" under this argument. If there's one common thread in all the botop cases that have come before ArbCom, this is it. Rich, Betacommand, Magioladitis – they all feel that they can do whatever they wish while ignoring and trampling over the community. Dirk apparently shares in this philosophy, as he showed clearly in the case request statements: "Maybe it is time that the community starts to get their act together" (emphasis his) [17]. ~ Rob13Talk 21:54, 1 August 2017 (UTC)[reply]
Comment by others:
Whatever the reasons of the community behind it - I do not agree that the community (at large) discussed the changes in the wording. --Dirk Beetstra T C 18:30, 31 July 2017 (UTC)[reply]
@BU Rob13: You don't need to kidnap them. A simple rewording of this would suffice: "Following the Committee's recommendation to clarify the WP:COSMETICBOT policy in the previous case, the changes to the wording were discussed in a well-publicized RfC, which resulted in a consensus for the new wording." One could consider to link the word 'consensus' to WP:LOCALCONSENSUS. --Dirk Beetstra T C 16:49, 1 August 2017 (UTC)[reply]
Well, the advertising was not too successful.... What else would you call a policy change in the bot policy that was mainly !voted on by bot operators .. broad community consensus .. that is even funnier. --Dirk Beetstra T C 18:20, 1 August 2017 (UTC)[reply]
I'd call it unqualified consensus. 22 participants is a significant amount, comparable to most BOTPOL-related RFCs. That the majority of participants were technically-minded or bot-related is no surprise, but rather a good sign, since this is a policy that affect technical aspects and bots. Had bot-ops not participated, that would be worrisome. Headbomb {t · c · p · b} 19:51, 1 August 2017 (UTC)[reply]
Of course it is, it got likely more outside participants than the RfCs in the last years .. it was the first one that was broadly announced in years. Says a lot about the community endorsement of the policy. --Dirk Beetstra T C 20:48, 1 August 2017 (UTC)[reply]
Complete nonsense. The RFC on IP bot operators took place a month before. Both the Cosmetic Bot RFC and the IP bot op RFC had about 22 participants. The previous on, from December 2016 (on activity requirements) has 11 participants, and one on mandating Assertion had 6 participants, but those covered rather uncontroversial updates of BOTPOL that would affect pretty much no one. And going by !vote count, the cosmetic bot RFC had a lot more people giving clear opinions on it. Headbomb {t · c · p · b} 21:32, 1 August 2017 (UTC)[reply]
@BU Rob13: What slippery slope, BU Rob13? I've said since the beginning that it is a community problem. --Dirk Beetstra T C 22:31, 1 August 2017 (UTC)[reply]

Magioladitis bludgeoned the process

edit

2) Magioladitis created a large number of discussions relating to COSMETICBOT in a period of a few days. [18] [19] [20] [21] These discussions were started after the community had already clarified the policy in an RfC, and they had the effect of bludgeoning the process.

Comment by Arbitrators:
Comment by parties:
Comment by others:

Topic ban (1)

edit

3) Magioladitis received a topic ban from discussing WP:COSMETICBOT in late June 2017. [22]

Comment by Arbitrators:
Comment by parties:
Comment by others:

Topic ban violation

edit

4) Magioladitis was blocked for violating his topic ban on discussing WP:COSMETICBOT in mid July 2017 as a result of multiple comments at WT:BOTPOL (e.g. [23] [24]). Magioladitis further tested the limits of his topic ban by creating an information page heavily related to COSMETICBOT. [25]

Comment by Arbitrators:
Comment by parties:
Comment by others:

Mass editing over complaints

edit

5) Magioladitis made large quantities of semi-automatic edits from his main account related to a pending bot task (e.g. [26]). These edits continued after receiving multiple complaints with no attempt to seek consensus for continuing to make these edits at high volume without a bot flag. [27] [28] The Arbitration Committee previously reminded Magioladitis against making contentious semi-automatic edits, especially relating to bot tasks.

Comment by Arbitrators:
Comment by parties:
Comment by others:

Topic ban (2)

edit

6) Magioladitis received a topic ban from semi-automatic editing in early July 2017. [29]

Comment by Arbitrators:
Comment by parties:
Comment by others:

Proposed remedies

edit

Note: All remedies that refer to a period of time, for example to a ban of X months or a revert parole of Y months, are to run concurrently unless otherwise stated.

Magioladitis desysopped

edit

1) For consistent poor judgement and failure to follow Wikipedia's policies and guidelines, Magioladitis is desysopped. He may regain adminship at any time through a successful Request for adminship.

Comment by Arbitrators:
Comment by parties:
Comment by others:

Magioladitis reminded

edit

2) Magioladitis is reminded that editing over the opposition of other editors without seeking consensus is often disruptive. When editors contest Magioladitis' edits, he is encouraged to stop editing until consensus is clearly demonstrated at an appropriate venue.

Comment by Arbitrators:
Comment by parties:
Comment by others:

Committee's future role

edit

3) The Committee reserves the right to act by motion if issues with Magioladitis' editing continue despite community sanctions.

Comment by Arbitrators:
Comment by parties:
I'm not sure if this is needed as a remedy or not, but my point is that I do not think the Committee should sanction Magioladitis for his behavior beyond desysopping. The community handled that with the topic bans, and hopefully the issues will not recur. The Committee is good at a number of things, but I don't think crafting technical remedies is one of them, if I may be a bit blunt. The current topic bans are containing the issues properly, so no reason to rock the boat. If the Committee wishes to give more teeth to community sanctions, I've proposed a remedy for that below. The only thing the Committee must consider right now is adminship, and only because no community-based process is available to handle that. If things go to shit down the road, a motion could always be used in lieu of a third case. ~ Rob13Talk 05:35, 20 August 2017 (UTC)[reply]
Comment by others:

Magioladitis restricted

edit

4) Magioladitis is restricted from violating community sanctions indefinitely. This includes but is not limited to topic bans, interaction bans, and other editing restrictions. This applies to both current and future community sanctions.

Comment by Arbitrators:
Comment by parties:
This is offered up as a possibility if the Committee wants to give more teeth to community sanctions. This would allow admins to perform AE blocks if a community sanction is violated by Magioladitis. This may be necessary given the amount of kerfuffle raised by the last block for a very clear topic ban violation. [30] (see comments especially by Dirk). I'm rather neutral on this; I merely offer it as an option. ~ Rob13Talk 05:35, 20 August 2017 (UTC)[reply]
Comment by others:

Proposed enforcement

edit

Template

edit

1) {text of proposed enforcement}

Comment by Arbitrators:
Comment by parties:
Comment by others:

Template

edit

2) {text of proposed enforcement}

Comment by Arbitrators:
Comment by parties:
Comment by others:

Proposals by Beetstra

edit

Proposed principles

edit

Template

edit

1) {text of Proposed principle}

Comment by Arbitrators:
Comment by parties:
Comment by others:

Template

edit

2) {text of Proposed principle}

Comment by Arbitrators:
Comment by parties:
Comment by others:


Proposed findings of fact

edit

There is no proof that the community has lost patience with Magioladitis

edit

1) All discussions relating to the encouragement by the ArbCom regarding 'Magioladitis 1', and all discussions and actions relating to (alleged) misbehavior by Magioladitis were heavily influenced by bot operators.

Comment by Arbitrators:
The text of the proposal does not line up well with the header. Bot operators are part of the community just like everyone else. If the substance of what you are saying is that only bot operators have lost patience with Magioladitis, then the relevant questions would include why this is so, and whether their feelings are justified. Newyorkbrad (talk) 14:26, 7 August 2017 (UTC)[reply]
Comment by parties:
I think it's true that technically-inclined editors are the ones that have lost the most patience with Magioladitis because many of his activities have affected us most severely. See the activities behind topic ban #1, for instance, which involved spamming discussions that technically-inclined editors would be most likely to participate in. We are not exclusively the ones who lost patience, though. See my analysis below that 75% of non-bot operators who participated in topic ban #2's discussion supported a topic ban. When the actions have affected the community at large, the non-technically-inclined editors have been just as fed up. ~ Rob13Talk 14:41, 7 August 2017 (UTC)[reply]
Comment by others:
One ARBCOM case following several years of "stop making those edits", two topic bans, and a second ARBCOM case is not evidence? Headbomb {t · c · p · b} 17:03, 3 August 2017 (UTC)[reply]
I'm technically inclined but not a bot operator. I'm unwilling to personalize my comments in this case towards Magioladitis, but am mostly here to express a loss of patience with bot-related misbehavior on Wikipedia in general, going all the way back to Betacommand 1 or maybe earlier. I see this case and its predecessor as more iterations of "same lousy editing, different bot" that we already have plenty of experience with. I'm actually liking the bot operator comments more here than in earlier cases though. I think we're all coming to understand that our previous DR approaches haven't been working. 173.228.123.121 (talk) 04:02, 9 August 2017 (UTC)[reply]

Template

edit

2) {text of proposed finding of fact}

Comment by Arbitrators:
Comment by parties:
Comment by others:

Template

edit

3) {text of proposed finding of fact}

Comment by Arbitrators:
Comment by parties:
Comment by others:

Template

edit

4) {text of proposed finding of fact}

Comment by Arbitrators:
Comment by parties:
Comment by others:

Proposed remedies

edit

Note: All remedies that refer to a period of time, for example to a ban of X months or a revert parole of Y months, are to run concurrently unless otherwise stated.

The bot operators are advised that they should work more closely with the community

edit

1) Bot operators are advised to more actively engage members of the community into their processes.

Comment by Arbitrators:
This would be better worded as "... other members of the community". And the question it raises is: How? Newyorkbrad (talk) 14:27, 7 August 2017 (UTC)[reply]
Comment by parties:
A pattern of poor logic, maybe. Again, unless we kidnap people who don't care about technical matters and force them to participate in RfCs or actively prohibit technically-inclined editors from becoming bot operators, the two sets "editors who are knowledgeable about technical matters and care to participate in technical RfCs" and "editors who possess the technical skills to be a bot operator" will remain very closely related. That's not the eighth wonder of the world; it's patently obvious. ~ Rob13Talk 14:19, 5 August 2017 (UTC)[reply]
Comment by others:
There is zero evidence there is a pattern here that generalizes to bot operators in general. Headbomb {t · c · p · b} 16:53, 3 August 2017 (UTC)[reply]
Sure .. read carefully through my evidence. In every single case involving the previous ArbCom case on Magioladitis, and all actions taken towards Magioladitis there is heavy involvement by bot operators. Maybe I should use the word 'encouraged'? --Dirk Beetstra T C 16:59, 3 August 2017 (UTC)[reply]
Again, this is either Magioladitis specific, or based on nonsensical statements that the WP:COSMETICBOT update wasn't done in collaboration with the community, or that bot ops are second class members of the community. Headbomb {t · c · p · b} 17:05, 3 August 2017 (UTC)[reply]
Are we seeing a pattern yet? --Dirk Beetstra T C 18:11, 3 August 2017 (UTC)[reply]
That you keep arguing things that aren't true based on no evidence? Yes I'm quite getting tired of that. Headbomb {t · c · p · b} 18:57, 3 August 2017 (UTC)[reply]
@Newyorkbrad: The ‘how’ question is something to throw back at the community - answering it might be the concern below that ArbCom does not advice/write policy, they can however state that things in procedures and guidelines are not well established, see previous remedies in ‘Magioladitis 1’. —Dirk Beetstra T C 16:29, 7 August 2017 (UTC)[reply]
Which neglects the fact that BAG nomination procedures, Bot trials, and bot policies are well established. Headbomb {t · c · p · b} 19:03, 7 August 2017 (UTC)[reply]
..by the bot operator community. We have the mandate, but are now in near isolation from them. BAG elects BAG, BAG sets policy. We have to take a step back. —Dirk Beetstra T C 06:51, 8 August 2017 (UTC)[reply]
No, the community elects BAG, just as the community sets policy. Headbomb {t · c · p · b} 17:25, 8 August 2017 (UTC)[reply]

Community encouraged

edit

2) The community is encouraged to review the Wikipedia:Bot policy and clearly implement what they expect from bot operators, and from editors who perform tasks on large scales or at high speeds. They are also encouraged to (re-)define the boundaries of 'large scale tasks', 'high speed editing', 'tasks that can be performed by a bot', in line what is common practice.

Comment by Arbitrators:
Comment by parties:
Comment by others:
This amounts to deciding when something is a pile or a heap. If you annoy people with several edits done against consensus, it doesn't matter how you do those edits, you need to stop. WP:MEATBOT is crystal clear on that. Headbomb {t · c · p · b} 16:57, 3 August 2017 (UTC)[reply]
No, it is quite easy to define that 'if your task involves more than 500 pages', or 'if you edit at faster than 10 edits per minute', then you need to consider that you need to discuss that task with the community, whatever you do. At the moment, there is nothing that stops me from stepping up to an editor who performed two AWB edits that both show up in my watchlist and state that they are flooding my watchlist. I could even step up to an editor who does one AWB editor and ask him to not do that again, because that job is better suited to be done by a bot. The problem is indeed a heap paradox, because there are no rules defined whatsoever, and they are consistently flouted by the community at large - several of the editors in my list of 'high speed editors' perform tasks that easily span pages of 500 contribs. The problem is not that there is not that there is one editor purportedly not listening to the community (well, one complaint), it is that there are no guidelines. To stay with your paradox: you are removing one grain, and think that there is no heap anymore. --Dirk Beetstra T C 17:12, 3 August 2017 (UTC)[reply]
And then someone will go gna gna gna, I've only made 400 edits at 9 edits per minute! WP:BOTPOL/WP:MEATBOT doesn't apply to me! Headbomb {t · c · p · b} 17:30, 3 August 2017 (UTC)[reply]
And that remark is perfectly in line with what you expect of the community. A total lack of good faith towards them. --Dirk Beetstra T C 18:17, 3 August 2017 (UTC)[reply]
And to address your concern - yes, that community member has not edited at high speed, and that community member did not do a large scale task. Currently the status is 'Gna gna gna, you've made 2 similar edits in 5 minutes, you were both doing a large scale task, ánd editing at high speed. Moreover, you are flooding my watchlist.' And now you have properly applied your heap paradox. --Dirk Beetstra T C 18:21, 3 August 2017 (UTC)[reply]

Members of the community are to seek consensus before taking on large scale tasks, or tasks at high speed

edit

3) Members of the community are to seek explicit consensus from the community before making any large scale or high speed edits as defined in the bot policy.

Editors who perform any tasks that involve a large number of pages, or are operated at high speeds, have to stop on complaints, and achieve consensus before continuing their edits.

Comment by Arbitrators:
The second sentence does have overtones of the "fait accompli" principle that we've adopted in the past. Newyorkbrad (talk) 14:29, 7 August 2017 (UTC)[reply]
Comment by parties:
Comment by others:
Adapted from Headbomb. --Dirk Beetstra T C 18:35, 3 August 2017 (UTC)[reply]
Complete WP:CREEP beyond ARBCOM's jurisdiction, which does not have the mandate to dictate policy. Consensus for basic/common task can often be assumed without any problem, having hundreds of editors create posts at the WP:VP to fix basic stuff like misuse of citation parameters is as unworkable as it is unneeded. Headbomb {t · c · p · b} 21:25, 5 August 2017 (UTC)[reply]

@Newyorkbrad: yes, ‘’fait accompli’’. If you go through the ‘high speed edits’ (some of which part of rather large sets of edits), there are many which are perfectly doable by bots. They have been done, they did not receive complaints. Concern is nonetheless that this is writing policy, unless we word it differently. —Dirk Beetstra T C 06:49, 8 August 2017 (UTC)[reply]

edit

4) Members of the BAG and other bot operators are indefinite restricted from closing RfCs and community discussions that relate to the bot policy, or other bot operators, or BAG members (and their election).

Comment by Arbitrators:
If the viewpoint behind this proposal is that BAG members or bot operators are automatically too "involved" to close these discussions, I am not persuaded. (And if that were the rationale, then the authorization below for a future appeal would not make much sense.) If the viewpoint is that BAG members and bot operators, as a group, are guilty of some sort of misconduct that warrants a "ban", it is unsupported, plus it would impose a sanction on editors who are not parties to the case (and as for as-yet-unknown future BAG members and bot operators, could not be made parties even if we wanted to). Newyorkbrad (talk) 14:32, 7 August 2017 (UTC)[reply]
Comment by parties:
Comment by others:
Appeal no earlier than 2 years after start of the ban. —Dirk Beetstra T C 20:37, 5 August 2017 (UTC)[reply]
Utterly ridiculous, and completely pointless restriction, devoid of any basis whatsoever in any BAG member's behaviour. That's like saying admins can't close RFCs related to deletion policy.Headbomb {t · c · p · b} 21:17, 5 August 2017 (UTC)[reply]
What magic happens when an editor becomes a bot operator or BAG that spontaneously reduces the ability to evaluate the consensus of the community and render a closing statement? I'd really like to know because it seems youre premise with this remedy and the following two is that BotOps and BAG aren't members of the community at large and should be excluded from consensus discussions and cant be trusted. Hasteur (talk) 21:58, 6 August 2017 (UTC)[reply]
@Hasteur: Reading comments here from some here clearly shows what ‘magic happens’ ... the previous case showed problems with parts of the bot policy, and they have clearly not been resolved. WP:INVOLVED comes to mind. My concern is that the bot policy and their operations arenot reflecting a broad community consensus, because bot operators understand the technical parts and the general public does not or does not care, and a tendency to ‘police [their] own’. Sure, you can take Magioladitis out of the equation, but that is not going to solve the issues, we will be back here within no-time with yet another BotOp (as I say in my opening statement, this is the third, that will be number 4 .. at least). I am sorry, it is not that I don’t trust the BotOps or BAG, but they affect the community, and we need to find ways to have the community decide. In my opinion, there is a clear ‘we know what is best for the community’ going on here, and taking out Magioladitis is NOT the solution. —Dirk Beetstra T C 07:12, 7 August 2017 (UTC)[reply]

@Newyorkbrad:I see your point. What I am trying to address is similar as above, trying to get rid of a situation where only/mainly BotOps decide ‘what is best for the community’. I know it is unintended, but the situation becomes autocratic. —Dirk Beetstra T C 16:36, 7 August 2017 (UTC) @Newyorkbrad: But is that not what ArbCom has done with problem areas on Wikipedia before, strict 1RR on pages in a certain area. That affects editors who were not party to the case, and affects editors who have never edited said pages. I see this as a variety of Wikipedia:General_sanctions#Types_of_sanctions. —Dirk Beetstra T C 13:28, 10 August 2017 (UTC)[reply]

Amendments to the bot policy

edit

5) Amendments to the bot policy will be discussed in RfCs which will be strictly held on WP:VPP. Bot operators and BAG members are allowed to comment, discuss and recommend in those discussions, but their !votes will be ignored by the closer of the RfC.

Comment by Arbitrators:
This would be a policy change outside the scope of ArbCom in general and this case in particular. If proposed elsewhere, I would oppose it as counterproductive and divisive. Newyorkbrad (talk) 14:35, 7 August 2017 (UTC)[reply]
Comment by parties:
Not only ludicrous but also against WP:ARBPOL. ~ Rob13Talk 22:14, 5 August 2017 (UTC)[reply]
Comment by others:
Obvious in combination with above, RfC not to be closed by bot operators or BAG members. Bot operators are obviously allowed to bring the RfC if they think it ichanges in the bot policy are necessary. Again, appeal no earlier than 2 years from instatementof the restriction. —Dirk Beetstra T C 20:45, 5 August 2017 (UTC)[reply]
Again, an utterly ridiculous, and completely pointless restriction. I have no words to describe how profoundly stupid this idea even is. Excluding the most knowledgeable and most affected editiors from the discussion is both perverse and moronic.Headbomb {t · c · p · b} 21:21, 5 August 2017 (UTC)[reply]
Great flying NO. The amount of discoussions that will flood VPP should take this will make VPP want to bifrucate and make a VPP-BOTS and VPP-NOBOTS. I think considered policy changes that can affect the wider community should be advertised on VPP, but that's not a requirement. And as to dismissing Bot-ers votes you're removing the right for Bot-ers to participate in the consensus discussion. Hasteur (talk) 21:58, 6 August 2017 (UTC)[reply]
@Newyorkbrad: Again, I see your point (I will never make a good drafter). I meant this as a way of making the community more in power of the policies and procedures .. I still believe that is needed to break the autocracy. I know that I have troubles explaining bot technicalities, we have to do a better hjob at that. —Dirk Beetstra T C 16:41, 7 August 2017 (UTC)[reply]

BAG membership

edit

6) New BAG members are to receive at least 2/3rds of their support !votes from non bot-operators or other BAG members.

BAG members who have been elected in the last 2 years, and who have not received a community support of 2/3rd of the !votes or more in their election shall be removed from BAG, and start a new procedure if they wish to be re-elected.

Comment by Arbitrators:
This proposed policy change is way outside the role of the ArbCom in general, and the scope of this case in particular. Newyorkbrad (talk) 14:36, 7 August 2017 (UTC)[reply]
Comment by parties:
Not only ludicrous but against WP:ARBPOL. ~ Rob13Talk 22:14, 5 August 2017 (UTC)[reply]
So by your argument, all admins are WP:INVOLVED with respect to other admins? Obviously not. I recommend reading that policy. ~ Rob13Talk 13:33, 7 August 2017 (UTC)[reply]
That may be your belief, Beetstra, but it is not supported by fact. We regularly list discussions in many extremely visible venues to encourage additional input from editors who may not watchlist technical pages. They just don't show up. I would love for more people to become involved in technical issues, and if you have a way to make that happen without setting up Wikipedia's equivalent of apartheid, let's hear it. The problem is not that bot operators are rejecting input from non-bot operators. The problem is that non-bot operators aren't interested in the discussions (if that is even a problem). ~ Rob13Talk 17:07, 7 August 2017 (UTC)[reply]
@Beetstra: A dictatorship of everyone who chooses to vote is usually called a "democracy". ~ Rob13Talk 17:45, 8 August 2017 (UTC)[reply]
Comment by others:
Consensus to be determined by a non-bot operator. Appeal after no less than 2 years by community RfC establishing an incorporation into the bot-policy of a community established ruling regarding BAG membership determination rules. This RfC is to be held on WP:VPP, no bot operators or BAG members to !vote, and closed by any admin of the community who is not a bot operator or BAG member, whatever the status of the above remedies. —Dirk Beetstra T C
Complete WP:CREEP beyond ARBCOM's jurisdiction, which does not have the mandate to dictate policy. Consensus for BAG membership is within the mandate of WP:CRATs, and excluding the most knowledgeable and most affected from the discussion is both perverse and moronic. Which also ignores the fact that every active BAG member did get more than 2/3 of the vote in support. Don't even recall one BAG member that ever got below 80% support. Headbomb {t · c · p · b} 21:28, 5 August 2017 (UTC)[reply]
2/3rd of !votes by editors ‘’’who are not bot operators’’’. I have given 3 examples of BAG members who did not get that. —Dirk Beetstra T C 16:47, 6 August 2017 (UTC)[reply]
Even more ludricrous if you change it to kicking out BAG members from the BAG based that even more insane criterion. And And I've yet to see those 3 examples. If you're referring to myself/Rob/Cyberpower's RfBAGs, Rob's was the only one with any opposes, both coming from bot operators that ended up at ARBCOM. You literally have no case whatsoever against BAG or any of its member. Headbomb {t · c · p · b} 18:47, 6 August 2017 (UTC)[reply]
Who is talking about opposes? I am only talking about support. —Dirk Beetstra T C 07:16, 7 August 2017 (UTC)[reply]
Since when did being confirmed for BAG become as high a threshold as Administrator? Furthermore, why does it have to be 2/3 of non botters? What ability to judge consensus or the ability to evaluate issues do we loose when we become a bot Operator or BAG? Furthermore retroactively dis-BAGing members already in is only going to cause more problems in being able to review BRFAs. In short, NO. Hasteur (talk) 21:58, 6 August 2017 (UTC)[reply]
We ‘police our own’, WP:INVOLVED. Clearly we don’t take that step back by ourselves. —Dirk Beetstra T C 07:15, 7 August 2017 (UTC)[reply]
@BU Rob13: That is a standard that I keep onto myself, which is related. As standalone situations I do not think that we are too involved, but the pattern goes too much the other way. My main worry is that the group is becomin/has become too closed to community input, a ‘we know what is best for the community’, ‘we willtake care of our own’. You often said that your only option is to drag non botops into the discussion, the opposite would be to ‘enforce’ significant non botop support. Maybe we botops do, with the best of intentions, a bad job in engaging them. We have our procedures, but is that really in line with the community? I don’t think so. —Dirk Beetstra T C 16:15, 7 August 2017 (UTC)[reply]
@Newyorkbrad: with that I can agree(in remedy form). —Dirk Beetstra T C 16:15, 7 August 2017 (UTC)[reply]
@BU Rob13: I fully understand the problem, BU Rob13. This is not about being at fault. The thing is, there are parts codified in the bot policy that the community does not care about, only the BotOps. Editors edit manually at bot-like speeds, and automated all over it. If those taks hit 20-30 edits a minute then they are not carefully checked, that is just fully automated running on a main account. Those are thus better done by a bot. That is a rule that the community, even outside of bot operators, regularly breaks. It is unreasonable to pinpoint that on one editor. —Dirk Beetstra T C 18:23, 7 August 2017 (UTC)[reply]
I know what I, as bot operator and fervent scripter, or as admin, would like to see, but I take a step back, as I do with some admin-editer rights. Not because I am involved, per WP:INVOLVED, but because I think that I am too involved as a person. There are certain consensusses that I uttery agree or disagree with, but where I do not !vote, because I am not the person who is affected. Too far to the other side, and you run nearly an autocracy. —Dirk Beetstra T C 18:31, 7 August 2017 (UTC)[reply]
Wikipedia, like any web 2.0 project, is at its very core an autocracy. But I've yet to see anywhere in BAG anything that resembles a hint of "we know best, therefore we can ignore others" attitude. Like anywhere else, ignorant opinions will be ignored (for example if someone objects to a task based on a misunderstanding of the task), but legitimate concerns are always heard and taken into account, regardless of who says them. All of our proceedings are open to anyone and everyone in the community, with the most important ones advertised in many places. Are we perfect? Of course not. But you also have to remember that BAG is actually empowered by the community to act on its behalf for the routine stuff, and required to ensure consensus in more touchy areas (WP:BAGG is extremely clear on this). The problem you describe doesn't actually exist. Headbomb {t · c · p · b} 19:22, 7 August 2017 (UTC)[reply]
that is then where we disagree, we bot operators are at the service of the community. We do not decide for them. We should take that step back, we are not aligned with the community. —Dirk Beetstra T C 20:25, 7 August 2017 (UTC)[reply]
I said BAG, not bot ops, was empowered by the community to act on its behalf, much like ARBCOM, CRATS, and ADMINS are all empowered to act on behalf of the community. No one has ever claimed any of those entities weren't servicing the community. Headbomb {t · c · p · b} 04:36, 8 August 2017 (UTC)[reply]
OK, I know where we are then and where we end. Those are exactly the standards that I mean, and where I have seen problems coming all the time. —Dirk Beetstra T C 06:44, 8 August 2017 (UTC)[reply]
Headbomb, re at its very core..., is autocracy really the word you wanted? I wonder if you meant something more like do-ocracy (article now deleted but you can gooble the term). It basically means WP:BOLD, people individually going off and doing stuff. That can work when there isn't too much factionalism, and the huge disparity in editing speed between bots and humans isn't used to create or amplify m:Wikimedia_power_structure#Technocracy too much. 173.228.123.121 (talk) 10:26, 8 August 2017 (UTC)[reply]
I'll admit that I thought Dirk meant it in the sense of self-governance, rather than in the sense of dictators. Anyway, self-governance is the way I intended my use of the word to come across, although I'm now unsure what the correct term would be. Headbomb {t · c · p · b} 17:36, 8 August 2017 (UTC)[reply]

Template

edit

7) {text of proposed remedy}

Comment by Arbitrators:
Comment by parties:
Comment by others:

Template

edit

8) {text of proposed remedy}

Comment by Arbitrators:
Comment by parties:
Comment by others:

Proposed enforcement

edit

Template

edit

1) {text of proposed enforcement}

Comment by Arbitrators:
Comment by parties:
Comment by others:

Template

edit

2) {text of proposed enforcement}

Comment by Arbitrators:
Comment by parties:
Comment by others:
The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

Proposals by Callanecc (drafting arbitrator)

edit

Proposed principles

edit

Purpose of Wikipedia

edit

1) The purpose of Wikipedia is to create a high-quality, free-content encyclopedia in an atmosphere of camaraderie and mutual respect among contributors. Contributors whose actions are detrimental to that goal may be asked to refrain from them, even when these actions are undertaken in good faith; and good faith actions, where disruptive, may still be sanctioned.

Comment by Arbitrators:
Comment by parties:
Comment by others:

Communication

edit

2) Editors should use their best efforts to communicate with one another, particularly when disputes arise. When an editor's input is consistently unclear or difficult to follow, the merits of his or her position may not be fully understood by those reading the communication. An editor's failure to respond to concerns with sufficient clarity, conciseness and detail, or failure to focus on the topic being discussed, can impede both collaborative editing and dispute resolution. It is a condition of operating a bot that the operator communicates cordially, promptly, and appropriately.

Comment by Arbitrators:
Comment by parties:
Comment by others:

Administrator conduct

edit

3) Administrators are trusted members of the community who are expected to lead by example. They are expected to follow Wikipedia's policies and restrictions which are placed upon them. Occasional mistakes are entirely compatible with this; administrators are not expected to be perfect. However, sustained or serious disruption of Wikipedia is incompatible with the expectations and responsibilities of administrators, and consistent or egregious poor judgement may result in the removal of administrator tools.

Comment by Arbitrators:
Comment by parties:
Comment by others:

Accountability for conduct

edit

4) Editors are accountable for their conduct. As such, they are expected to respond appropriately to queries about their actions and to justify them where needed. Where the Arbitration Committee, the community or other authorised person imposes a sanction, editors are expected to comply with both the letter and spirit of the sanction.

Comment by Arbitrators:
Comment by parties:
Comment by others:

Recidivism

edit

5) Editors who have been sanctioned, whether by the Arbitration Committee or the community, for improper conduct are expected to avoid conduct which is below Wikipedia's expectations. Failure to demonstrate appropriate conduct may result in the editor being subject to increasingly severe sanctions.

Comment by Arbitrators:
Comment by parties:
Comment by others:

Automated and semi-automated editing

edit

6) Fully automated bot editing and semi-automated editing scripts perform an important and valuable function on Wikipedia. To facilitate the regulation and coordination of automated editing, the community has a long-established bot policy and a Bot Approvals Group responsible for reviewing potential bot operators' requests for bot approval.

Comment by Arbitrators:
Comment by parties:
Comment by others:

Bot-like editing

edit

7) For the purpose of dispute resolution, it is irrelevant whether high-speed or large-scale edits that involve errors an attentive human would not make are actually being performed by a bot, by a human assisted by a script, or even by a human without any programmatic assistance. Where editors have made a number of similar edits in a short time space and other editors have raised concerns about those edits, the editor is to stop making the edits and engage in discussion. However, merely editing quickly, particularly for a short time, is not by itself disruptive.

Comment by Arbitrators:
Comment by parties:
Sensible. ~ Rob13Talk 16:26, 21 August 2017 (UTC)[reply]
Comment by others:
Just commenting that 'high-speed' and 'large-scale' (or a combination thereof) are very ill defined. --Dirk Beetstra T C 10:58, 21 August 2017 (UTC)[reply]
I think here part of the problem is the question of what was the concern about those edits. Note that per WP:CONSENSUS not every concern needs to be addressed (and see also my above concern about the definition of both 'high-speed' and 'large-scale' which is relevant to this). --Dirk Beetstra T C 11:04, 21 August 2017 (UTC)[reply]

Cosmetic edits and AWB general fixes

edit

8) Changes are typically considered substantive if they affect something visible to readers and consumers of Wikipedia, while changes that do not are typically considered cosmetic. Minor edits are not usually considered cosmetic but still need consensus to be done by bots. Bots should not make edits which are purely cosmetic, however, when making an approved substantive change, bots may make edits which would otherwise be considered cosmetic. Exceptions for bots to make a purely cosmetic edit must be approved by consensus. While WP:COSMETICBOT applies only to bots, human editors may also wish to follow this guidance for the reasons given above, especially if making such changes on large scales.

AWB AWB general fixes (genfixes) are a package of common fixes which can be enabled in bulk in AWB by the user/bot operator. Some general fixes are substantive, while others are cosmetic. It is the responsibility of the bot operator or editor using AWB to ensure that their editing falls within policy, including the bot policy and BRFA (if applicable).

Comment by Arbitrators:
Comment by parties:
Probably not needed due to scope. COSMETICBOT is only relevant due to discussions about COSMETICBOT, so I don't think this is needed as a principle. No strong objection; it just kind of bloats things. ~ Rob13Talk 16:27, 21 August 2017 (UTC)[reply]
I think it's important background for those who aren't aware of the specifics of bots, AWB and BOTPOL. Callanecc (talkcontribslogs) 02:07, 23 August 2017 (UTC)[reply]
Comment by others:
I've tweaked the grammar a bit. However, I'll point out that WP:GENFIXES are a set of fixes made by WP:AWB, which can be enabled (or not) by the user/bot. Several WP:GENFIXES are substantive/non-cosmetic, and whether they are so does not depend on whether or not they are made at the same time of a substantive change.
I would personally simply strike "Cosmetic edits, made at the same time as a substantive change are sometimes called "general fixes" or "genfixes". If something needs to be said about general fixes, I would suggest a seperate FOF/Principle/Whatever: "AWB general fixes are a package of common fixes which can be enabled in bulk in AWB by the user/bot operator. Some general fixes are substantive, while others are cosmetic. If some AWB genfixes are miscategorized as non-cosmetic / non-minor by the AWB, it is the responsibility of the bot operator to ensure that the bot operation remains within the scope of its BRFA and bot policy. Unlike for what has happened at WP:CWERRORS following Magioladitis 1, no discussion of the categorization of AWB genfixes as cosmetic/non-cosmetic/minor has occurred." Headbomb {t · c · p · b} 13:25, 21 August 2017 (UTC)[reply]
Gonna ping Callanecc (talk · contribs) on this, because the remark on what genfixes are might have slipped through the crack. Headbomb {t · c · p · b} 01:00, 23 August 2017 (UTC)[reply]
@Headbomb: How is what I've added? To clarify what you've said, AWB has different categories, but what's in those categories hasn't been discussed onwiki? Callanecc (talkcontribslogs) 02:07, 23 August 2017 (UTC)[reply]
Looks fine to me now. As for the second part, that is correct, unlike for WP:CHECKWIKI, where categorization is documented at WP:CWERRORS, and discussed on the talk page, there's been no equivalent effort for AWB Genfixes. Most of them are categorized correctly, but there's no documentation on how they are categorized, nor venue to discussion the categorization of those general fixes. Headbomb {t · c · p · b} 04:21, 23 August 2017 (UTC)[reply]
Didn't realise this was duplicated. Beetstra, I've moved your comment up to #7. Callanecc (talkcontribslogs) 11:24, 21 August 2017 (UTC)[reply]

High-speed editing

9) For the purposes of dispute resolution, whether an editor is engaging in "high-speed editing" (that is, the number of edits per minute) is irrelevant. Where editors have made a number of similar edits in a short time space and other editors have raised concerns about those edits, the editor is to stop making the edits and engage in discussion.

Comment by Arbitrators:
Comment by parties:
Comment by others:

Proposed findings of fact

edit

Locus of dispute

edit

1) This dispute centres on the conduct of Magioladitis (talk · contribs), which has led to sanctions being imposed by the Arbitration Committee and the by community.

Comment by Arbitrators:
Comment by parties:
I don't think "especially" onward is quite accurate. When sanctions have been imposed, he hasn't violated them terribly often. The problem is that, for a good while, every week seemed to bring a new behavior requiring a new sanction. Perhaps "The dispute centres on the conduct of Magioladitis (talk · contribs), which has led to sanctions imposed by the Arbitration Committee and the community"? ~ Rob13Talk 16:29, 21 August 2017 (UTC)[reply]
Changed, thanks. Callanecc (talkcontribslogs) 09:24, 22 August 2017 (UTC)[reply]
Comment by others:
I would mainly (or even only) say the first community sanction after the first case. Magioladitis stopped early on with AWB on his main account before the second sanction was applied. There is of course relation with the Remedies of the first case. --Dirk Beetstra T C 11:08, 21 August 2017 (UTC)[reply]
I agree, however the fact that the community still passed it (because he didn't stop when asked) is relevant. Callanecc (talkcontribslogs) 09:24, 22 August 2017 (UTC)[reply]

Arbitration Committee remedy: Unblocking own bot

edit

2) As a result of the first arbitration case, Magioladitis was restricted from:

Unblocking their own bot when it has been blocked by another administrator. After discussion with the blocking administrator and/or on the bot owners' noticeboard, the blocking administrator or an uninvolved administrator may unblock the bot.

Comment by Arbitrators:
Comment by parties:
True, but irrelevant. Perhaps you can wrap up several of these "previous case" findings of fact into a single finding of fact referencing the previous case? It is relevant to a pattern of behavior, but it probably isn't important to recap the previous case's results step-by-step rather than just point there. ~ Rob13Talk 16:32, 21 August 2017 (UTC)[reply]
This is probably more a leftover of when I had all three remedies in the same finding. I separated them out under the theme which was most relevant. Though, when I move this to the PD page I might combine them all at the start (under this finding). Callanecc (talkcontribslogs) 09:24, 22 August 2017 (UTC)[reply]
Comment by others:
Example of the previous - he has not unblocked his own bot, Remedy not violated. --Dirk Beetstra T C 11:09, 21 August 2017 (UTC)[reply]
While true, I don't particularly see the relevance, unless it's to establish Magioladitis did not misuse admin-specific tools [which is a fine thing to establish, since it's true]. Headbomb {t · c · p · b} 13:27, 21 August 2017 (UTC)[reply]

ISBN edits

edit
Arbitration Committee remedy: Semi-automated editing
edit

3) As a result of the first arbitration case, Magioladitis was reminded that:

Performing the same or similar series of edits in an automated fashion using a bot and in a semi-automated fashion on his main account is acceptable only as long as the edits are not contentious. Should Yobot be stopped or blocked for a series of edits, Magioladitis may not perform the same pattern of edits via semi-automated tools from his main account where this might reasonably be perceived as evading the block. In this circumstance, Magioladitis (like any other editor) should await discussion and consensus as to whether or not the edits are permissible and useful, and resume making such edits through any account only if and when the consensus is favorable.

Comment by Arbitrators:
Comment by parties:
Highly relevant to the ISBN edits. ~ Rob13Talk 16:33, 21 August 2017 (UTC)[reply]
Comment by others:
Again the same problem - I think the committee would do good in establishing what is actually the concern that are raised about Magioladitis' semi-automated editing on their main account. --Dirk Beetstra T C 11:11, 21 August 2017 (UTC)[reply]
ISBN edits allowed by consensus
edit

4) A request for comment, held in March 2017, was closed by BU Rob13 (talk · contribs) with consensus in favour of allowing bots to replace the ISBN magic link with the ISBN template.

On 22 May 2017, Magic links bot (talk · contribs) was approved to change ISBN magic links to the template. PrimeBOT 13 (talk · contribs) was approved, on 18 June 2017, to make the same changes. Neither bot made other "general fixes" (or "genfixes") to the page while changing the ISBNs.

Comment by Arbitrators:
Comment by parties:
I'll just echo Headbomb. "Cosmetic edit" is a technical definition. There can be consensus to allow a cosmetic edit, but that doesn't stop it from being cosmetic. (Though the ISBN edits were never cosmetic, because the template introduces an extra wikilink. So perhaps abandon the second sentence altogether?) ~ Rob13Talk 16:34, 21 August 2017 (UTC)[reply]
Comment by others:
A correction: The categorization of "switching magic word ISBNs to a template ISBNs" as cosmetic vs substantive didn't change as a result of the RFC. What happen is simply that there was consensus to replace magic words ISBNs with template ISBNs, and to do so by bots. Whether this edit is technically cosmetic or technically substantive is irrelevant, consensus established bots (not humans) should make those conversions. I'll point that someone, could have in good faith believed it implied humans could be making those edits as well. Certainly on a small scale, or bundled with other edits, humans can do that too, but the whole point of doing it with bots is that they can easily be ignored on watchlists (whereas humans cannot be easily ignored). When you have several hundred thousand pages affected (or maybe even millions), that is not a trivial concern.
I would simply strike "As a result of this RfC, the ISBN change became a substantive edit, in contrast to a cosmetic edit."
In addition, there were other bots than Magic links bot approved to this change. I believe there was also PrimeBOT, Yobot, and my own CitationCleanerBot (this one still under trial). I'll dig the BRFAs later today. Headbomb {t · c · p · b} 13:40, 21 August 2017 (UTC)[reply]
@Headbomb: I've removed that sentence and emphasised "bots". Links to the BRFAs would be great! Callanecc (talkcontribslogs) 09:24, 22 August 2017 (UTC)[reply]
@Callanecc:, by date of approval
Headbomb {t · c · p · b} 09:48, 22 August 2017 (UTC)[reply]
Thanks. Callanecc (talkcontribslogs) 11:19, 22 August 2017 (UTC)[reply]
Magioladitis and ISBN fixes
edit

5) After the RfC referred to above, Magioladitis submitted a bot approval request, requesting approval for Yobot (talk · contribs), run by Magioladitis, to change the ISBN magic link to templates and to make "genfixes". This task was not approved until 22 July 2017. Beginning from 18:11, 21 June 2017 (UTC), Magioladitis used AWB from his main account to change ISBN magic links to templates and to make "genfixes" while still waiting for his bot to be approved. It appears that Magioladitis was using the code from his unapproved bot for this editing.[reply]

Comment by Arbitrators:
Comment by parties:
Comment by others:
Concerns raised regarding ISBN edits
edit

6) At 19:36, 21 June 2017 (UTC), less than two hours after Magioladitis started, the first objection was raised by Justlettersandnumbers on Magioladitis's talk page. At 07:33, 24 June 2017 (UTC), Materialscientist also raised a similar concern on Magioladitis's talk page. Magioladitis' last ISBN edit (in this series) was at 07:37, 24 June 2017 (UTC).[reply]

In the discussions which ensued with Justlettersandnumbers and Materialscientist, Magioladitis stated that he was aware two bots where already making these changes. He also stated that the reason he was making these edits, and not leaving them for the bots, was that he wanted to make general fixes as well (which would be prohibited as cosmetic on their own).

Comment by Arbitrators:
Comment by parties:
...and then Magioladitis started again a week later, triggering more complaints. ~ Rob13Talk 06:14, 22 August 2017 (UTC)[reply]
Comment by others:
The concern is indeed the similar: it is focussing on 'flooding watchlists'. AFAIK no other concerns were mentioned. --Dirk Beetstra T C 11:20, 21 August 2017 (UTC)[reply]
Also noting that there is about 5 minutes worth of discussion with Materialscientist until Magioladitis indicated that he stopped. —Dirk Beetstra T C 03:54, 22 August 2017 (UTC)[reply]
@BU Rob13: No, he started again 9 days later (see below). One of the people mentioned here returned with exactly the same concern - flooding watchlists. Then the AN/I was started, and when that gained traction Magioladitis stopped (he finished I believe) and has not used AWB since. (not that this has anything to do with this FoF, that he restarted and attracted more concerns is another FoF). --Dirk Beetstra T C 07:05, 22 August 2017 (UTC)[reply]
@Callanecc: can you please insert links to the exact timings into this - Justlettersandnumbers complained expressed her concern on 19:36, 21 June 2017, Materialscientist added the expressed a similar concern on 07:33, 24 June 2017 (which is not about 24 hours later but 60 hours later - note that he was running the task for 60 hours before getting a second complaintconcern was expressed and stopped soon after). I think it is then also fair to show the timing of the first diff of the first run. --Dirk Beetstra T C 07:11, 22 August 2017 (UTC) (adapted language --Dirk Beetstra T C 07:13, 22 August 2017 (UTC))[reply]
I've added specific times. Callanecc (talkcontribslogs) 09:24, 22 August 2017 (UTC)[reply]
@Callanecc: this is the diff with the remark from Materialscientist, which is definitely not at 21:35, 22 June 2017, and hence definitely not less than 2 hours earlier, it was at 10:33, 24 June 2017, 60 hours later. The first post in that thread is by Magioladitis himself, which is indeed at 21:35 on the 22nd. Magioladitis responded within minutes (diff at 10:34, 24 June 2017, and indicated that he stopped at 10:42, 24 June 2017, less than 10 minutes after the second concern. You still may want to include what the exact timing was when he stopped .. he may have (temporarily) stopped before 10:42, 24 June 2017.
I still remain, he edited for more than 60 hours, one single concern, stating 'you are flooding my watchlist' (which he also addressed before continuing, diff in relevant contrib list). Also, the request does not specifically request to stop, it asks '... Is there any reason why you can't leave this task to the bot? You'd be doing people a favour ..."
I still ask the committee to address the concerns expressed - was Magioladitis violating policy/not following guidelines or common practice in said concern. Ask yourself, what do you do when a cop is stopping you on a highway in the Netherlands (following current Dutch law) with the concern 'Sir, I am stopping you because you are driving in a red car' (hint: it is not forbidden to drive in a red car in the Netherlands, speeding and strange lightning are, but nothing according to the colour of the car ONLY). I hope the committee will address the expressed concerns on their validity and in relation to common practice. --Dirk Beetstra T C 10:57, 22 August 2017 (UTC)[reply]
Fixed. Thanks, I was looking at Magio's edit. I've also added a sentence about his last ISBN edit in this series. Callanecc (talkcontribslogs) 11:11, 22 August 2017 (UTC)[reply]
Thanks, please also see below. --Dirk Beetstra T C 11:36, 22 August 2017 (UTC)[reply]
@Callanecc: is what Justlettersandnumbers is stating in the initiation of the linked thread really an 'objection' .. or is it a concern, question or request. --Dirk Beetstra T C 11:40, 22 August 2017 (UTC)[reply]
Magioladitis continued making ISBN edits
edit

7) At 07:46, 30 June 2017 (UTC), Magioladitis continued making ISBN edits, which he had previously agreed to stop making. Magioladitis indicated that he was working on ISBN links not found by Magic links bot (talk · contribs) and that he believed this was an opportunity to make general fixes at the same time.

However, he also indicated that he did not believe he should have to stop making semi-automated edits when only one editor (even though he had been asked by both Justlettersandnumbers and Materialscientist) had requested that he stop.

Comment by Arbitrators:
Comment by parties:
I think it's worth explicitly noting that (a) Magioladitis was running the same code as his pending Yobot bot task, and (b) Magioladitis had been asked to stop by multiple editors, not one. ~ Rob13Talk 16:37, 21 August 2017 (UTC)[reply]
@Beetstra: For b, see the proposed finding of fact 6, which shows two already. For a, this is by Magioalditis' own admission that he was running his bot task in "manual mode" (meaning semi-auto, presumably, since you can't run code in fully manual): [31]. ~ Rob13Talk 03:42, 22 August 2017 (UTC)[reply]
I've done a in #5 and b in #7. Callanecc (talkcontribslogs) 09:24, 22 August 2017 (UTC)[reply]
Comment by others:
After the first concern, raised here and resuming there are roughly 9 days. The concerns remain the same. --Dirk Beetstra T C 11:23, 21 August 2017 (UTC)[reply]
@BU Rob13: can you please providethe diffs from the evidence? —Dirk Beetstra T C 03:19, 22 August 2017 (UTC)[reply]
OK. So he was asked by two editors , which is part of the earlier FoF, he stopped for 9 days, and then one of those two same editors returned with exactly the same concern. I just wanted to define your ‘multiple’ here. No new diffs then. —Dirk Beetstra T C 03:49, 22 August 2017 (UTC)[reply]
@Callanecc: Also here, he was only asked by Justlettersandnumbers. The request to stop from Materialscientist came 60 hours later, and he stopped there within 10 minutes after that (he may have finished already at that point, and may have actually stopped earlier). The whole run was with one concern 'you are flooding my watchlist' from one editor. May I ask the committee to properly check timelines and diffs in the evidence, and not just go by the word of mouth that 'multiple editors were complaining', while in fact it is only 2 (granted, 2 is multiple). --Dirk Beetstra T C 11:03, 22 August 2017 (UTC)[reply]
@Callanecc: Did Justlettersandnumbers ask Magioladitis to stop, or did they only show concerns, asked questions, gave suggestions. I see from e.g. Justlettersandnumbers first post in the first run that they would not be able to use their watchlist until Magioladitis would stop, but not a request 'can you please stop, you are flooding my watchlist'. --Dirk Beetstra T C 11:43, 22 August 2017 (UTC)[reply]
Community sanction: AWB
edit

8) In response to Magioladitis's ISBN edits and not stopping when asked, a discussion at the Administrators' noticeboard for incidents was started in July 2017. There was consensus that:

The proposal to ban Magioladitis from using AWB passes, and he is banned from using AWB or any other semi-automated or automated editing tools on his main account for a period of 2 months ... Magioladitis can appeal to the community for his AWB ban to be removed on or after 7 September, so giving a firm 2 month ban and leaving the community with the option to lift the ban or continue it as they see fit after that time. If no consensus is reached then, I would suggest the ban automatically be lifted at that point (i.e definite consensus needed to extend the ban). I would also ask/suggest that any breach of this ban be reported to the Arbitration Committee rather than raising the issue at ANI. ... (includes) any new or existing additional accounts without a bot flag beginning to use AWB or any other semi-automated or automated tools.

Comment by Arbitrators:
Comment by parties:
Comment by others:
@Callanecc: I still fail to see that he was ‘asked to stop’ in those threads. There were concerns (you are flooding mywatchlist unless you stop’), but no clear ‘stop it’, ‘please stop’. —Dirk Beetstra T C 17:35, 22 August 2017 (UTC)[reply]
edit

9) Magioladitis has demonstrated a failure to understand concerns raised by community members by continuing to make semi-automated edits after being asked not to by community members. These edits are disruptive and border on being tendentious.

The ISBN edits also effectively disregarded the spirit of the reminder issued by the Arbitration Committee which only specified that he should not make edits which Yobot has been "stopped or blocked" from making (not edits which are still pending approval).

Comment by Arbitrators:
Comment by parties:
I don't think it's quite accurate to say one can game a reminder, since a reminder isn't a restriction that one has to follow. Perhaps "effectively disregarded the spirit of the reminder" would be better? Not sure how to phrase this best. ~ Rob13Talk 16:39, 21 August 2017 (UTC)[reply]
Yep, thank you. Callanecc (talkcontribslogs) 08:46, 22 August 2017 (UTC)[reply]
Comment by others:
The concern raised, consistently, is that semi-automated edits on a main account are flooding watchlists.
The second sentence reads both ways. --Dirk Beetstra T C 11:31, 21 August 2017 (UTC)[reply]
'the spirit of the reminder', really, as a 'Finding of Fact'? --Dirk Beetstra T C 11:35, 22 August 2017 (UTC)[reply]

Magioladitis and cosmetic editing

edit
Arbitration Committee remedy: Cosmetic edits
edit

10) As a result of the first arbitration case (January to March 2017), Magioladitis was restricted from:

Making any semi-automated edits which do not affect the rendered visual output of a page. This restriction does not apply to edits which address issues related to accessibility guidelines. Further, Magioladitis may seek consensus to perform a specific type of semi-automated edit that would normally fall under this restriction at the administrators' noticeboard. Any uninvolved administrator may close such a discussion with consensus to perform a specific type of semi-automated edit. All discussions should be logged on the case page, regardless of outcome.

Comment by Arbitrators:
@Magioladitis, BU Rob13, Headbomb, and Beetstra: Question for you all, would it be correct to say that Magioladitis violated this restriction as the ISBN edits were approved for bots to make not for him to make (nor discussed at AN per the restriction)? Would they be covered by the accessibility guidelines? Callanecc (talkcontribslogs) 09:25, 22 August 2017 (UTC)[reply]
Comment by parties:
Per my comment for #2, these can be condensed into one finding of fact pointing out the previous case. ~ Rob13Talk 16:40, 21 August 2017 (UTC)[reply]
I originally had it like that but wanted to separate them into the two 'narratives/themes' I was going with. But I think I'll combine all three remedies from the previous case in one (as FoF 2). Callanecc (talkcontribslogs) 09:24, 22 August 2017 (UTC)[reply]
Responding to your question, no, I wouldn't call that a violation. The edits affected visual output. Magioladitis' edits were fine until editors complained about his method of doing this, at which point failing to seek consensus specifically for doing them without a bot flag becomes an issue. ~ Rob13Talk 13:25, 22 August 2017 (UTC)[reply]
@Callanecc::
a) the edits were changing the rendered visual output. There was no restriction for these edits. Rob wrote: "actually, it does introduce a new ISBN link, so this is fine".
b) the RfC concluded that the replacement has consensus in general. It also concluded that a bot is needed. But not exclusivelly. -- Magioladitis (talk) 11:56, 22 August 2017 (UTC)[reply]
Thanks. Callanecc (talkcontribslogs) 00:08, 23 August 2017 (UTC)[reply]
Comment by others:
Sure, which I think that Magioladitis quite strictly followed. No discussions logged on the case page. --Dirk Beetstra T C 11:32, 21 August 2017 (UTC)[reply]
@Callanecc: My response to this is extremely nuanced.
Is the replacement of ISBNs magic links by ISBN templates technically cosmetic? No, it is not. There is a change in the rendered output of the page. It's a fucking subtle one, but it is there. So on whether or not Magioladitis violated the restriction, the answer is No, he did not. Nor did he violate WP:COSMETICBOT with those edits, if you read that policy like a lawyer.
HOWEVER, in WP:COSMETICBOT, the words are typically considered... appear. The emphasis on typically is there on purpose because the term 'cosmetic' traditionally been used to cover all edits that didn't meet a certain threshold of usefulness. Where exactly the threshold was/is has never been defined, because it cannot be defined. Ambiguous cases are resolved by establishing WP:CONSENSUS somewhere (typically one of the village pumps). In WP:COSMETICBOT, this concept of a threshold of usefulness is in the preamble in the form of edits that are not worth the time spent reviewing them.
We give examples, and set an extremely low bar of usefulness, under which edits can be presumed to be unwanted by the community by default. Clearing this bar does not mean your edits have met the threshold of usefulness, but failing to clear the bar almost always means they haven't met the threshold. This is why why we additionally have Minor edits are not usually considered cosmetic but still need consensus to be done by bots.
Bots (and WP:MEATBOTs) need to cross the threshold of usefulness. This is the spirit of the WP:COSMETICBOT section (and of WP:BOTREQUIRE #2/#4). And this, Magioladitis has violated.
The first time, before he was asked to stop by multiple editors, the ISBN replacements are at worse an WP:AGF violation of the spirit of the policy, which can happen to anyone. I think it was obvious to everyone but Magioladitis that such a series of edits were terrible idea to do manually (see also WP:AWBRULES #4), but I [like BU Rob 13] also don't think that's a punishable offence at this point. The WP:COSMETICBOT bar was cleared, even if the threshold of usefulness wasn't. Up until that point, nothing super problematic happened, and we're in WP:AGF territory.
The problem happens when Magioladitis resumed. There, with years of complaints about making useless/near-useless edits, an ARBCOM case, an RFC on COSMETICBOT, multiple dead-end discussions all resulting in people telling Magioladitis that whatever he was trying to achieve wasn't going to happen, a topic ban on such discussion, which was subsequently gamed and violated, ... there's literally no excuse, which is why he got slapped with a second topic ban, this time covered all semi-automated editting. Headbomb {t · c · p · b} 10:50, 22 August 2017 (UTC)[reply]
@Callanecc: in reply to your question: The way I read the ArbCom remedy, Magioladitis was not allowed to perform semi-automated edits that do not change the visible output of a page unless he has specific permission to make semi-automated edits that do not change the visible output of a page. I therefore do not think that it is correct to say that he violated this .. changing 'ISBN 123456789x' to '{{ISBN|123456789x}}' does change the visible output of a page (it gives different functionality). --Dirk Beetstra T C 11:55, 22 August 2017 (UTC)[reply]
Changes to the COSMETICBOT policy
edit

11) Following the Committee's recommendation to clarify the WP:COSMETICBOT policy in the previous case, a request for comment was held from March to May 2017. The RfC was advertised at Wikipedia:Centralized discussion and consensus was achieved to adopt a new wording to the policy.

Comment by Arbitrators:
Comment by parties:
Comment by others:
COSMETICBOT discussions
edit

12) From 30 Dec to 28 Feb 2017, Magioladitis started two discussions related to the COSMETICBOT policy and general fixes (30 Dec, also 28 Feb). From 12 Jun to 20 Jun 2017, Magioladitis started five discussion related to the COSMETICBOT policy and general fixes (12 Jun (see this also), 15 Jun, also 15 Jun, 20 Jun)

During these discussions, Magioladitis was requested to stop starting discussions about the COSMETICBOT policy as his proposals were not gaining traction.

Comment by Arbitrators:
Comment by parties:
Comment by others:
Community sanction
edit

13) In June 2017, a discussion at the Administrators' noticeboard was started following Magioladitis having started a number of similar discussions about the COSMETICBOT policy. In this discussion, Magioladitis was accused of disrupting Wikipedia by failing to drop the stick and bludgeoned the process by creating a number of similar discussions in a short time-span. The discussion resolved that:

Magioladitis is topic banned from initiating or participating in discussions concerning the amendment, removal, or replacement of WP:COSMETICBOT, or the in discussions concerning the impact of WP:COSMETICBOT on other bot operators (such as whether or not bot operators are allowed, or should be required, to perform WP:GENFIXES with their own bots, or theoretical bots which may be developed in the future). As an exception to this ban, he may make a single !vote with a short (<300 words) rationale if the discussion calls for !voting, and give single short replies (<300 words) to other editors when directly asked a question (1 reply per direct question).

Magioladitis may, in good faith, seek specific clarifications on how to interpret COSMETICBOT for his own bots and projects he is involved with (such as WP:CHECKWIKI), but may only do so on his bots' talk page, at the relevant project's talk page (such as WT:CHECKWIKI or an appropriate subpage), or at WP:BOTN. In other words, asking questions like "When is CW Error #02 considered a cosmetic edit?" or "Should CW Error #02's priority be lowered to 'Medium' or 'Low' important?" is fine, asking/arguing "WP:COSMETICBOT hampers our ability to do WP:CHECKWIKI fixes, how can we ammend/fix/ignore WP:COSMETICBOT?" is not.

As an additional exception, Magioladitis may continue to participate in any active discussion he has started from before the topic ban until its natural conclusion. He may not start new threads, subthreads, or unarchive old threads, and a discussion is considered concluded when the thread is archived, either by bot or by user.

Comment by Arbitrators:
Comment by parties:
Comment by others:
Magioladitis was blocked for breaching community sanctions
edit

14) On 12 July 2017, Magioladitis was warned by Kingpin13 (talk · contribs) for making edits in violation of his community-imposed topic ban relating to participating in discussions about the COSMETICBOT policy. He made another edit in violation of the topic ban and was further warned. After making two further edits, of the type he was warned to stop ([32]), he was blocked for two days.

Comment by Arbitrators:
Comment by parties:
Comment by others:
Gaming COSMETICBOT topic ban
edit

15) Also on 12 July 2017, Magioladitis created an "information page" at WP:COSMETICEDIT. While this was not technically a violation of the topic ban, it does appear to game the intent of the topic ban (which was to remove Magioladitis from discussions concerning COSMETICBOT, with limited exceptions) and was disruptive.

Comment by Arbitrators:
Comment by parties:
Comment by others:

Proposed remedies

edit

Note: All remedies that refer to a period of time, for example to a ban of X months or a revert parole of Y months, are to run concurrently unless otherwise stated.

Template

edit

1) {text of proposed remedy}

Comment by Arbitrators:
Comment by parties:
Comment by others:

Template

edit

2) {text of proposed remedy}

Comment by Arbitrators:
Comment by parties:
Comment by others:

Proposed enforcement

edit

Standard enforcement provisions will apply.

Analysis of evidence

edit

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.


Place here items of evidence (with diffs) and detailed analysis

Response to "Bots will make out of scope edits" from Beetstra

edit

As the BAG member who approved Wikipedia:Bots/Requests for approval/PrimeBOT 13, every edit you cited is well within scope. The community endorsed replacing magic links with templates in all namespaces. This is comparable to how templates are replaced/substituted in all namespaces before being deleted after a TfD, which often involves editing talk page archives. This is often misunderstood by community members who are unfamiliar with these processes, but when a template (or magic link) goes away, the rendered output changes unless we make these edits. ~ Rob13Talk 17:27, 23 July 2017 (UTC)[reply]

Comment by Arbitrators:
Could someone please explain this dispute over the ISBN links to those of us with no background on the issue? Thanks. Newyorkbrad (talk) 03:07, 27 July 2017 (UTC)[reply]
Comment by parties:
@Newyorkbrad: The ISBN links/magic links changes are what led to the second topic ban. I tried not to get overly specific in describing the exact edits in my evidence, since the whole situation gets a bit complex. Basically, some Mediawiki devs got together and decided to get together and do away with magic links. Magic links are what causes ISBN 123456789X to appear as a link (note that there is no special wikitext there - just normal text). In response, the community decided that we need to replace all current magic links with an equivalent template via bot. These were the edits in question. The rest of the situation is described in my "Mass changes from main account – second topic ban" section. Please let me know if you have further questions after reading that. ~ Rob13Talk 03:43, 27 July 2017 (UTC)[reply]
@Newyorkbrad:In short: COSMETICBOT changed to now allow edits that do not change the visual output. In this discussion many bot owners participated and decided that the claims that "all edits that do not change the visual outcome ar cosmetic" is nonsense. This means that now COSMETICBOT allows more edits without any futher discussion to be needed. The problem is that the community was/is not 100% ready for this change. Slight changes to the visual output like the ISBN fixes are still considered by some people as "cosmetic" tasks. This, we still have complains and we should give time to the community that now many of the wikignome edits me and ohers do are not considered cosmetic anymore. Rob at some point claimed that the ISBN changes were cosmetic even if he was the one that approved a bot to perform the task. So, yes the ISBN story is connected to the COSMETICBOT. If it was clear to the community that these changes are useful because we want to help the Mediawiki programmers and make their life easier, there will be less complains. Same holds for other changes that I have in mind, like adding/removing non-breaking spaces, non-visible characters, etc. The discussions about COSMETiCBOT should become wider and show that the community endorses edits that are not cosmetic otherwise there will always be complains. Beetstra is right that the dicussions so far have been dictated by bot operators. This is the reasonI wanted to start an RfC. Just chaning the policy is not enough. The community has to decdie whether to endorse non-cosmetic changes in large scale. -- Magioladitis (talk) 00:19, 28 July 2017 (UTC)[reply]
@Headbomb: The previous COSMETICBOT has a reference to a pywiki script and many people claimed that "all edit that do not change the visual outcome". After my effort to show examples COSMETICBOT chnaged to included examples of edits that are not considered cosmetic even if they do not chang the cisual output. I still wnt to make an RfC about which edits should be done and in which way. This is still pending. No policy that edits that cna be done by bots, can't be done by editors. (ISBN fixes part 1). ISBN fixes part 2 contained edit not done by the bots. -- Magioladitis (talk) 00:37, 28 July 2017 (UTC)[reply]
@Headbomb: I believe that the community will do the obvious when asked: All changes described in the Manual of Style are allowed to be done by single editors. Otherwise, we will have WP:OWNERSHIP problems with people claiming that some pages do not have to follow the MoS. COSMETICBOT's new wording helps in the wanted direction. -- Magioladitis (talk) 00:41, 28 July 2017 (UTC)[reply]
Recent discussions in my latest BRFAs show that judging general fixes case by case is nice but it can turn to unneccessary bureaucracy and that we still lack a good guideline of how to work with secondrary tasks. Till now I have not seen (m)any people sayinh that minor tasks should not be done. How we will actually do them is still a question. I have tried different approaches but an RfC would help in forming a bettr strategy based not only in the opiinion og bot operators. -- Magioladitis (talk) 00:47, 28 July 2017 (UTC)[reply]
Comment by others:
@BU Rob13: sure, but if I say: First you type an I, then an S, then a B, then an N, and then a space, followed by the numbers 123456789, and finally an X, you will get 'ISBN 123456789X', then that is different from First you type an I, then an S, then a B, then an N, and then a space, followed by the numbers 123456789 and finally an X, you will get 'ISBN 123456789X'. If I wanted to instruct someone to write the latter, I would say First you type two curly brackets, then an I, then an S, then a B, then an N, and then a piping symbol, followed by the numbers 123456789, an X, and two closing curly brackets, you will get 'ISBN 123456789X'. I know that everything is rendered the same, it is not what I instructed to type. It changes the quote. --Dirk Beetstra T C 17:45, 23 July 2017 (UTC)[reply]
@Beetstra: When magic links go away, the former will render as ISBN 123456789X. That is the problem. It changes not just the wiki text but the actual rendered outcome. You lose a link. It's similar to if Wikipedia suddenly became Wikipedia. A decade of consensus at TfD tells us that, in such cases, we change wikitext to preserve rendered text because we prioritize the latter over the former. ~ Rob13Talk 17:51, 23 July 2017 (UTC)[reply]
@BU Rob13: sure, but this is not about the rendered text. But I will make it more clear: The bots will replace 'ISBN 123456789X' with '{{ISBN|123456789X}}'. That is different in the rendered text. In some cases, editors do not mean the rendered text, in other cases editors mean the non-rendered text. The point is, whatever a bot is programmed to do, some things break. --Dirk Beetstra T C 18:26, 23 July 2017 (UTC)[reply]
The bots in question do not edit within nowiki tags and haven't for some time. That was an early bug which was quickly worked out in trials. ~ Rob13Talk 23:11, 23 July 2017 (UTC)[reply]
@Newyorkbrad: Here's the breakdown, as far as I am aware:
As far as I can tell, the issue here is with the fact of the topic ban and the issues that led to it being imposed. All of the rest of the bullets are background to the situation, no wrongdoing is alleged in them. Some of the other statements in this case are confusing the issue with various irrelevancies, including:
  • Whether these edits are cosmetic as defined by WP:COSMETICBOT.
  • Whether WP:COSMETICBOT is binding on human editors.
  • Whether the bots themselves made or make mistakes.
  • Whether the bots would have made the changes that Magioladitis made from his non-bot account.
  • Whether the approval of the three bots was "properly coordinated".
HTH. Anomie 19:33, 27 July 2017 (UTC)[reply]
@Magioladitis: "COSMETICBOT changed to now allow edits that do not change the visual output."
COSMETIC did not change in that regard. Cosmetic edits were always allowed provided they had consensus, as were minor edits. This was true before the RFC just as it is true now. The issue is that you did not have such consensus to do those edits, regardless of whether they were cosmetic or not. You tried to do them by bot initially, without consensus. Consensus was then established, and your bot was approved accordingly. When you did them manually, a significant amount of people complained, telling you to instead do this by bots rather than cluttering their watchlists. But rather than simply accept that this is a task best left for bots, you kept doing those edits and arguing that you were allowed to do them.
Headbomb {t · c · p · b} 00:27, 28 July 2017 (UTC)[reply]
@Magioladitis: "The community has to endorse non-cosmetic changes in large scale."
The community most certainly doesn't have to do that, and will likely most likely never approve anything that writes a blank check for any edit as long as it is non-cosmetic. This is why WP:COSMETICBOT also mentions "Minor edits are not usually considered cosmetic but still need consensus to be done by bots." and why WP:MEATBOT mentions "it is irrelevant whether high-speed or large-scale edits ... are actually being performed by [a bot or human. No matter the method, the disruptive editing must stop or the user may end up blocked."
Headbomb {t · c · p · b} 00:34, 28 July 2017 (UTC)[reply]
If the ISBN links on Wikipedia are rewritten by a bot, I'd want them to be unhyphenated for reasons given in another post. Hyphenating them was a good faith mistake that should be undone if we're going to mess with all those pages anyway. Among other things, hyphenation messes up the Wiki search function: if you want to find all the pages referring to a given ISBN, you have to look for both the hyphenated and unhyphenated versions, and figuring out the hyphenated one is complicated. So we should standardize on unhyphenated ISBN's just like every other site I've been able to find has done. I don't understand why we're getting rid of the magic links and won't address that here. 173.228.123.121 (talk) 23:11, 30 July 2017 (UTC)[reply]
Hyphenation is a different matter. You can start a discussion at WP:VPT if you want all ISBNs to be hyphenated or not (I'd rather have them be all hyphenated personally). If there's a consensus for it (one way or another), templates make it easy to strip or add the hyphens. Headbomb {t · c · p · b} 23:44, 30 July 2017 (UTC)[reply]
Hmm, there would have to be a user preference or special css to turn the hyphenation rendering on and off; the search engine would still have to deal with both hyphenated and unhyphenated ISBN's (maybe the template could store both somehow); and while a Lua template or server extension could do the hyphenation, I think it would be hard with traditional templates. Maybe Rich can tell us how he got the idea of hyphenating the ISBNs in the first place. I can't speak for him but (having done similar things myself more times than I can remember) I can empathize with the idea that a just-slightly-tricky problem like ISBN hyphenation presents the same irresistible attraction to programmers that laser pointers do to cats. IMHO it's best to do that type of thing on your own computer rather than on one used by millions of other people. 173.228.123.121 (talk) 07:24, 31 July 2017 (UTC)[reply]
"Template" as a placeholder for "the thing that will be invoked when you place {{ISBN|9780132456789}} in an article. The template would likely invoke a Lua module, much like the CS1/2 templates invoke Lua modules. (CS1/2 templates would also have the hyphenation/de-hyphenation functionality in them if the proposal passes). I like the idea of hyphens being displayed or not by preferences, but the point is the details should be hammered out elsewhere than an ARBCOM page. Headbomb {t · c · p · b} 14:25, 31 July 2017 (UTC)[reply]
Ok, point taken about where to discuss the technical ISBN stuff. I hope Arbcom will still ask itself how we got in the ISBN hyphenation business in the first place, and the non-trivial software project it will turn into if we stick with it. My own answer: the implementers decided we should do it because we can, even though it's of negative value to the project (at least in its current incarnation); and this microcosmically reflects many other bot projects, including drama-free ones.

Joseph Weizenbaum's 1976 book Computer Power and Human Reason has a section on "compulsive programmers" (excerpts) that profiles an extreme version of what I'd call a forerunner to today's obsessive Wikipedia bot programmers. Some parts are outdated but some parts never change.

N.b.: Weizenbaum states that chapter is actually autobiographical, describing an early phase of his own programming career. Most of us here who program have probably experienced the same thing at some level. 173.228.123.121 (talk) 04:04, 1 August 2017 (UTC)[reply]

Self analysis of "The community did not significantly participate nor endorse the requested review on the policy on cosmetic edits" from Beetstra

edit
counts (so people can check / correct)

2) The community is encouraged to hold an RfC to clarify the nature of "cosmetic" edits and to reevaluate community consensus about the utility and scope of restrictions on such edits. Technical feedback may be provided at phab:T11790 or phab:T127173. The committee notes that an RfC on this topic is currently under development.

Passed 14 to 0 at 23:37, 22 March 2017 (UTC)

An RfC titled "WP:COSMETICBOT update" was initiated by Headbomb (bot operator/BAG member), and closed by User:BD2412 (bot operator).



Edit counts:


Oppose/Support

Schonken suggested an alternative that was opposed by 3 bot ops (Anomie, HellKnowz, Headbomb) and 1 non bot op (Snuge Purveyor).


To be complete (even though I am not asserting that this did not happen):

Comment by Arbitrators:
Comment by parties:
To my knowledge, this is the most widely attended bot policy related RfC since I joined the project over two years ago. This needs to be viewed in the context of participation at similarly technical discussions. The discussion was open to all and advertised in many very public locations: AWB talk page, CHECKWIKI talk page, Village Pump (policy). The total number of editors on Wikipedia who care about both technical matters and policy discussions is very low. ~ Rob13Talk 19:38, 24 July 2017 (UTC)[reply]
Sure. But then don't assert that it has a broad community consensus. It had the consensus of the bot operators. --Dirk Beetstra T C 19:40, 24 July 2017 (UTC)[reply]
There is consensus among the entire community who bothered to show up to a well-advertised discussion. We can't drag them kicking and screaming to the discussion. WP:SILENCE applies if you're trying to invoke the masses of editors who saw the Village Pump notice and declined to participate. In any event, it's worth keeping in perspective what that discussion on COSMETICBOT was. It was a simple reword to make the previous section clearer. I wouldn't expect rewords for clarity to draw 100 editors. Nothing of substance changed in how the policy was enforced. There's a separate philosophical question here, by the way. In the American South, do we care about consensus on global warming among scientists or the general population? I would say consensus among all bot operators on a matter that is best understood by bot operators is plenty. ~ Rob13Talk 19:51, 24 July 2017 (UTC)[reply]
But then again, there are zero COSMETICBOT violations being alleged since the close of the previous case, so my comment below applies here as well. ~ Rob13Talk 19:53, 24 July 2017 (UTC)[reply]
Comment by others:
This was also posted at WP:CENT. Headbomb {t · c · p · b} 18:00, 25 July 2017 (UTC)[reply]
@Headbomb and BU Rob13: I hope it is clear that I am not asserting in any form that the community did not have a chance to participate. I show that the community did not - as encouraged by the committee and as asked by User:Mkdw in her questions. You are free to provide evidence to show that despite not having participated, that the community does take the policy into account (i.e. show that it is an exception that members of the community who use automation do not perform purely cosmetic edits, and that members that do so nonetheless do get strongly admonished by the community, etc. etc.). --Dirk Beetstra T C 11:25, 26 July 2017 (UTC)[reply]
"However, the discussion was heavily dominated by bot operators (16 out of 20 edits by bot operators, remaining 4), and did not solicit broader community input (e.g. 9 edits by OP, 7 by project members, remaining 4)." Emphasis mine. The bold part claims exactly that, and your recent tweak to that passage only achieved a non-sensical contradiction. Community input was requested, but was not solicited? I fail to see the difference between a request for input, and a solicitation of input. Headbomb {t · c · p · b} 13:56, 26 July 2017 (UTC)[reply]

Self analysis of "The community did not significantly participate.2C nor endorse the requested review on common fixes" from Beetstra

edit
counts (so people can check / correct)

Discussion initiated by User:Headbomb (not project member).

In terms of project members:

In terms of edit counts (20 edits total):

In terms of bot operators

Note, discussion was announced in other venues, e.g.:

To be complete (even though I do not assert that the community did not have a chance to comment), this has also be announced in:

(which are sent to people, and mentioned in other venues).

(unless I have missed further discussions - which I will then analyse accordingly).

Comment by Arbitrators:
Comment by parties:
Note that I do not believe anyone is contending that Magioladitis has made CHECKWIKI edits that were cosmetic in nature since the previous case closed. I certainly am not. ~ Rob13Talk 19:28, 24 July 2017 (UTC)[reply]
@Beetstra: My point is that this analysis really doesn't impact this case. I'm not criticizing your analysis at all. I fully understand what you're responding to, and it answers the question that's been asked. I just think we may have the wrong question. At first glance, a lack of clarity or community buy-in looks like a mitigating factor. I'm making the point that it's not a mitigating factor to behavioral issues if those behavioral issues aren't impacted by the policy in any way. To use an analogy, I'm worried about a situation where an editor is unblocked for blatant juvenile vandalism because they didn't understand what WP:CONTRAST means. One is not related to the other. I'm sure this is obvious to you, but it may not be to non-technically-inclined arbs who lump everything bot-like in one pile. ~ Rob13Talk 19:46, 24 July 2017 (UTC)[reply]
Comment by others:
@BU Rob13: what has that to do with this analysis? I am not talking about his bot edits, I am talking about the discussions and whether they have significant community input (as per request from User:Mkdw, above). --Dirk Beetstra T C 19:33, 24 July 2017 (UTC)[reply]

Self analysis of "The outcome of the first topic ban was heavily influenced by bot operators" from Beetstra

edit
counts (so people can check / correct)

Discussion: Wikipedia:Administrators' noticeboard/Archive290#Proposal to topic ban Magioladitis from COSMETICBOT-related discussions

  • Topic ban proposal by Headbomb (bot operator)
  • Topic ban discussion closed by Cyberpower678 (bot operator)
  • I have removed the many comments by User:Magioladitis (bot operator) as well as comments (not !votes) from a couple other users.

Support (chronological order of supporting - 16 editors, 7 bot operators, 9 non bot operators): Headbomb (bot operator), Anomie (bot operator), BU Rob13 (bot operator), Hellknowz (bot operator), Xeno (bot operator), Dennis Brown, Hchc2009, Ealdgyth, FACE WITH TEARS OF JOY, Beyond My Ken, Lugnuts, Only in death, Hasteur (bot operator), Fram, Green Cardamom (bot operator), JohnBlackburne

Oppose: (chronological order of opposing; 4 editors, 1 bot operator, 3 non bot operators): Rich Farmbrough (bot operator), Agathoclea, TParis, KoshVorlon

Comment by Arbitrators:
Comment by parties:
Among all bot operators, support for the proposed topic ban was 7/8 (87.5%). Among all non bot operators, support for the proposed topic ban was 9/12 (75%). These are similar percentages – certainly no statistically significant difference. Even if one removed all bot operators, 75% in support of a ban with 12 editors participating in the discussion would almost certainly be sufficient to enact a topic ban. I would appreciate if Cyberpower678 would comment on how they would have closed this discussion if only the non bot operators had participated. Would you still find consensus? ~ Rob13Talk 19:57, 24 July 2017 (UTC)[reply]
@Beetstra: The big difference is that neither topic ban had anything to do with technical matters. One was about WP:BLUDGEON and the other was about WP:CONSENSUS (editing against, specifically) and WP:DE. Those are things the community can fully handle with ease (and they did). I will again reiterate that it has not been alleged anywhere that Magioladitis improperly operated his bot or violated the bot policy at any time after the first case. I do want to be fair to him, so it's important not to confuse that issue and assume he's being accused of replicating the exact behavioral issues of the last case. He has not; these are new behavioral issues. The only pattern is that of poor judgement, in the sense of WP:ADMINCOND. Also, Cyberpower678 was not BAG when they made that close. ~ Rob13Talk 21:13, 24 July 2017 (UTC)[reply]
@Beetstra: This all begs a question - non-technical editors almost never contribute to anything related to technical things, because that's not where their interests lie. Such a bar would be so high that we could not meet it; getting more than 10 non-technical editors to even think about a technical matter isn't going to happen based on past experience. If the Committee were to find that we can't have technical-related policy without affirmative non-silent consensus among non-technical editors and broad participation from them, do we lose the ability to have any changes to technical policy or the creation of new technical policies? How does this affect bot policy and other technical policies long-term? ~ Rob13Talk 17:38, 26 July 2017 (UTC)[reply]
Comment by others:
@BU Rob13: Of course, we shouldn't forget that User:Cyberpower678 is a bot operator himself as well (currently BAG, not sure if they were at the time of closing this discussion - if so I will update above tallies). Of course the return question is: how would a non bot-operator assess this consensus when the !voting was solely by bot operators on a fellow bot operator.
Or maybe I should see this in the light of your above comment consensus among all bot operators on a matter that is best understood by bot operators is plenty. WP:CIR rears its ugly head again - maybe we should only consider the consensus as brought forward by the !votes of the bot operators, after all, they best understood what Magioladitis did wrong.
Talking about statistics: the funny thing is, in both the support and oppose !votes the community seems to follow the bot operators. --Dirk Beetstra T C 20:17, 24 July 2017 (UTC)[reply]
  • To be perfectly clear, my close was only in an admin capacity, of assessing consensus, which in my view shouldn't matter if the participants are bot ops or not. They are human editors and each are entitled to their views. I have little interest in punishing or imposing sanctions against Magioladitis, and refuse to get involved in this ArbCom case. My interactions and participation in events concerning Magioladitis will continue to remain unbiased and neutral.—CYBERPOWER (Around) 20:57, 24 July 2017 (UTC)[reply]
@Beetstra: With respect, I disagree with your theseis that bot operators voting en-masse was some form of impermissible canvassing. As bot operators who operate under the same rules of the road we desire not to open an even bigger can of worms by having one member of our "project" bring excessive scrutiny by their actions. While we are not a bureracracy or hierarchy, there is some authority to speak of things is conferred based on being an editor who does similar things, but stays (for the most part) out of the black book of sanctions. Hasteur (talk) 16:23, 25 July 2017 (UTC)[reply]
@Hasteur: Now you have drawn conclusions from my evidence that I have not asserted. I do not assert whether the sole !voting by bot operators would be something impermissible. User:Mkdw above asks "by the community enacted through broad enough consensus (Relatively to ANI)" (my bolding). I question that, a majority of bot operators set the rules (WP:COSMETICBOT RfC), and a significant number of bot operators !votes to uphold those rules (application of sanction). I will leave it to the Arbitrators to decide whether that is following applicable conduct.
I know that you do not want to open a bigger can of worms (me neither, I operate under the same rules) .. the question that I am trying to address is: does the community understand under which rules bot operators operate (and they themselves as well), and does the community think that they should follow those rules. I have yet to see evidence that the community agrees with it beyond WP:SILENCE. --Dirk Beetstra T C 12:16, 26 July 2017 (UTC)[reply]
If you don't want to open the can of worms tear down this wall and strike the entire point/analysis. Your calling out that a non-trivial portion of people who wanted the restriction are already involved in BotOps is something that should be lauded, not aspersions cast at. To pull this out of the context of this problem (and use the veil of ignorance test), how would you feel if editors who had no experience in deletion and notability arguments started !voting and closing XfD arguments? While we can read their contributions, they won't have the same amount of weight as an editor who argues from a reasoned position using our operating policies and is familiar with how they're typically applied. Hasteur (talk) 12:30, 26 July 2017 (UTC)[reply]
I don't think I opened a can of worms. That can of worms already cracked open years ago .. And I am not afraid of any consequences (which, by the way, I am sure there are not going to be). (note, I have not been involved in the first case where these worms may have surfaced). --Dirk Beetstra T C 14:05, 26 July 2017 (UTC)[reply]

Response to "Re: The community did not significantly participate, nor endorse the requested review on common fixes" by Headbomb

edit
Comment by Arbitrators:
Comment by parties:
Comment by others:
@Headbomb: I have not asserted why the community did not participate, or whether that is a problem. In the first case about Magioladitis, the Arbitration Committee encouraged the community to look into the issues when examining Magioladitis (a non-binding remedy). I don't know the reason why the community did not participate, I just assert that what the committee encouraged the community to do did not happen. I am however unsure what evidence you provide there that shows that it is indeed not important that the community did not follow the encouragement of the committee (it looks more like an analysis of my evidence). The committee apparently found it important in the first case that the community should look into the subject. --Dirk Beetstra T C 16:51, 26 July 2017 (UTC)[reply]

Response to "Re: The community did not significantly participate nor endorse the requested review on the policy on cosmetic edits" by Headbomb

edit
Comment by Arbitrators:
Comment by parties:
Comment by others:
@Headbomb: I have not asserted why the community did not participate, or whether that is a problem. In the first case about Magioladitis, the Arbitration Committee encouraged the community to look into the issues when examining Magioladitis (a non-binding remedy). I don't know the reason why the community did not participate, I just assert that what the committee encouraged the community to do did not happen (and User:Mkdw asks that in their questions: ".. and with broad enough support"). I am however unsure what evidence you provide there that shows that it is indeed not important that the community did not follow the encouragement of the committee (it looks more like an analysis of my evidence). The committee apparently found it important in the first case that the community should look into the subject. --Dirk Beetstra T C 16:51, 26 July 2017 (UTC)[reply]

Response to "The outcome of the first topic ban was heavily influenced by bot operators" by Headbomb

edit
Comment by Arbitrators:
Comment by parties:
The average topic ban outcome at ANI has far fewer than 22 editors participating, so the "no broad consensus" point just doesn't make much sense to me. If you wanted to actually show it doesn't have broad consensus relative to ANI, go to WP:RESTRICT, take the last N topic bans issued for sufficiently large N, and see what percentile the Magioladitis topic ban falls within in terms of participation. I expect it's in the 70th percentile or higher. ~ Rob13Talk 17:07, 26 July 2017 (UTC)[reply]
@Beetstra: Fine, see my analysis below. ~ Rob13Talk 18:28, 26 July 2017 (UTC)[reply]
@Beetstra: Jumping in to answer your question to Headbomb, this is the point I've been trying to make the entire case. Those edits were not cosmetic-only because they introduced a new wikilink to ISBN. They did not violate any policy or guideline initially. The edits were fine. See my section on the second topic ban of evidence. The moment they became not fine was the moment that multiple community members asked Magioladitis to stop, breaking the silence of consensus in favor of the task, and he chose to continue doing them against consensus. WP:CONSENSUS and WP:DE are very relevant. The AWB Rules of Use, which requires that you don't use the software for controversial edits without seeking consensus, is relevant. WP:COSMETICBOT is not. ~ Rob13Talk 18:35, 26 July 2017 (UTC)[reply]
Given that the concerns raised were "Stop, you're spamming our watchlists with edits that could be done with a bot flag", no. The complaints you fielded were that you shouldn't do those edits en masse without a bot flag. ~ Rob13Talk 18:46, 26 July 2017 (UTC)[reply]
This is the space for parties to comment, but I'll avoid threaded comments as a courtesy. The bot operators in question (e.g. Primefac) have disputed that these edits weren't done by their bots or couldn't be done by their bots, from my understanding. In any event, "I was right" is not a reason to ignore complaints and refuse to start a discussion/get consensus. That's why the topic ban was implemented. Magioladitis repeatedly brought up the pending Yobot task in the topic ban discussion, presumably noting that the edits he did could have been done by the flagged bot if approved. ~ Rob13Talk 19:01, 26 July 2017 (UTC)[reply]
Comment by others:
@Headbomb: User:BU Rob13 states in his opening statement: "The community cannot make a decision to desysop under current policy." That means, in my opinion, that BU Rob13's conclusion is that the community wants Magioladitis desysopped, that the community has sufficient problems with Magioladitis' actions to warrant such a remedy. However, the applicable rules that Magioladitis' 'violated' are rules that the community did not heavily participate in, and the bans that were enacted without what seems a broad consensus by the community (User:Mkdw also asks "Were the restrictions placed against Magioladitis by the community enacted through broad enough consensus? (Relatively to ANI)").
Also here, I fail to see what evidence you provide as to why is it not important that there is what is seemingly not a broad consensus by the community. --Dirk Beetstra T C 16:51, 26 July 2017 (UTC)[reply]
"However, the applicable rules that Magioladitis' 'violated' are rules that the community did not heavily participate in, and the bans that were enacted without what seems a broad consensus by the community". No, what is under investigation is if Magioladitis violated WP:ADMINCOND, by repeatedly exhibiting WP:IDIDNTHEARTHAT behaviour, by being WP:TE/WP:DE enough to earn himself no one but two topic bans following an ARBCOM case, and so on, and if those violations are egregious enough to de-sysop him.
And, despite your assertions, the bans do have broad consensus within the community. Headbomb {t · c · p · b} 17:02, 26 July 2017 (UTC)[reply]
@Headbomb: That is strange, Headbomb .. he was here in the first case because of.. problems with COSMETICBOT (there is a remedy for Magioladitis for that), after the case he got a community ban on .. COSMETICBOT. Again, WP:COSMETICBOT was mainly decided by bot operators, and he was banned from discussing cosmeticbot. So who did he not listen to, to the community, or to the bot operators.
@Headbomb: Also the second topic ban, though technically for editing from main account, was instigated due to Magioladitis performing COSMETIC edits (ISBN 123456789x to {{ISBN}}) on their main account. Why do you say that the case is focussed on IDIDNTHEARTHAT and ADMINCOND if everything that led to sanctions is about COSMETICBOT. --Dirk Beetstra T C 18:28, 26 July 2017 (UTC)[reply]
Because something can be related to a thing without being about that thing. The scope of the case whether or not Magioladitis is still fit to be an admin or not. What lead to this point is Magioladitis' WP:IDIDNTHEARTHAT behaviour, failure to stop beating WP:DEADHORSEs. Per WP:BOTPOL/WP:MEATBOT, when edits are contentious, it doesn't matter if those edits are made from a bot account or from a main account, they must stop, and they should have stopped long before topic bans were required, regardless of whether or not such edits fell under WP:COSMETICBOT. Headbomb {t · c · p · b} 19:22, 26 July 2017 (UTC)[reply]
@BU Rob13: I say that there is not a large input of the community. If you think it is, I invite you to show so. --Dirk Beetstra T C 18:06, 26 July 2017 (UTC)[reply]
@BU Rob13: Also, my evidence states that the decision was heavily influenced by bot operators .. I did not assert there that the overall (or even, if you do not take bot operators in account) was weak or strong (relatively to ANI). --Dirk Beetstra T C 18:14, 26 July 2017 (UTC)[reply]
Re "Everything started with the COSMETICBOT..." You have repeated many times, without any sort of evidence, that "people" (This is about User:bgwhite) left because they were "targeted". This has never been demonstrated or even hinted at, and even if somehow bgwhite felt that way, it's certainly not true that they were, in fact, targeted. Likewise, it is most definitely not true that "the discussion never made it to CHECKWIKI to remove some of the tasks that [are/were] considered cosmetic", see this discussion in particular, which I have pointed you to countless of times, and in which you yourself participate. Headbomb {t · c · p · b} 20:11, 26 July 2017 (UTC)[reply]

Analysis of recent restrictions at ANI for context

edit

Let's look at every community restriction implemented at ANI since May, sans the Magioladitis topic bans. I will list the total number of editors who participated under the topic ban heading, or if no heading, in support or opposition of the topic ban. I will not count the person banned, obviously. They're in reverse order.

Comment by Arbitrators:
Comment by parties:
These counts aren't perfect by any means, and I didn't put a huge amount of time into meticulously counting. Expect things to be off by 1-2 either way. Still, the general trend is that 22 editors commenting on a topic ban is exceedingly normal, perhaps a bit on the high side. There are 9 discussions with less editors commenting and only 2 with more since May 2017. That would put the first Magioladitis topic ban above the 75th percentile, about where I thought it would be. ~ Rob13Talk 18:27, 26 July 2017 (UTC)[reply]
Comment by others:

Relevance of COSMETICBOT

edit

Magioladitis (in "Bans are unrelated to cosmetic changes editing" section) and BU Rob13 (in "Spamming discussions – first topic ban" and "Mass changes from main account – second topic ban" sections) agree that Magioladitis' behavior has nothing to do with cosmetic edits. Since both parties agree the substance of COSMETICBOT is not relevant to "Magioladitis' conduct since the previous case was closed" (which is the scope of the case), it appears to be rather clearly out of scope. ~ Rob13Talk 03:04, 27 July 2017 (UTC)[reply]

Comment by Arbitrators:
Comment by parties:
It is good to finally agree on something. ~ Rob13Talk 03:04, 27 July 2017 (UTC)[reply]
The below argument breaks down pretty fast given that I approved those exact edits for PrimeBOT and expressed support multiple times for the similar Yobot task [33] [34] (there are others). Those ISBN edits have consensus to be done by bot, so COSMETICBOT doesn't apply. ~ Rob13Talk 04:33, 27 July 2017 (UTC)[reply]
@Beetstra: I've tried to make this point a few times, and Magioladitis has made it as well. Re your new "The second topic ban precipitated due to high speed edits that did not change the rendered visual output of a page" section: The edit introduces a new link to ISBN. Click on "ISBN" for both links. It changes where you go, so it alters the rendered output of the page. ~ Rob13Talk 17:43, 27 July 2017 (UTC)[reply]
@Beetstra: Don't I know it! [35] ~ Rob13Talk 18:07, 27 July 2017 (UTC)[reply]
@Beetstra: I definitely agree many editors initally didn't understand the change being made, but the complaints at Magioladitis' talk page asking him to stop weren't related to whether the changes should be made. They were specifically related to whether the account doing them should have a bot flag. The problem, from those editors' point-of-view, was the watchlist spam that they couldn't easily hide without delving into installing scripts, etc. (As an aside, yeah, that's technically true - note COSMETICBOT provides alternative ways to be "substantive" other than visually changing the page, such as changing where a link goes to.) ~ Rob13Talk 18:43, 27 July 2017 (UTC)[reply]
This is the problem here. The edits are not related but the COSMETICBOT policy does. Not everyone understood that the problem was that the policy can be used in order to disallow many edits that the community in fact endorses. Rob and others are trying to prevent even a discussion about how to make edits the community endorses. The problem is deeper. For example Rob tried to throw the ISBN fixes under the COSMETICBOT policy to prevend them. -- Magioladitis (talk) 03:51, 27 July 2017 (UTC)[reply]
Comment by others:
@BU Rob13 and Magioladitis: I have removed that point of evidence. That is friggin' subtle. To the naked eye these things look exactly the same. --Dirk Beetstra T C 18:02, 27 July 2017 (UTC)[reply]
@BU Rob13: <pedantic>Visually, the pages are the same .. just the functionality changed.</pedantic> .. This must have thrown a lot of editors of and a lot of editors must have gotten upset because they did not see the functionality change (I guess the last paragraph of WP:SURPRISE somewhat applies). --Dirk Beetstra T C 18:19, 27 July 2017 (UTC)[reply]
@BU Rob13: Sure, for the last ban that is indeed the case (I just thought, wrongly, that also there there was a COSMETICBOT factor). --Dirk Beetstra T C 19:28, 27 July 2017 (UTC)[reply]

Self analysis of "The community has not defined 'high speed editing', nor whether or when this is disruptive" from Beetstra

edit
examples of edit speeds
The following discussion has been closed. Please do not modify it.

(I have, arbitrarily, chosen speeds higher than 6 per minute for editors, as the bot policy uses '1 edit every 10 seconds' (i.e. 6 edits per minute) for unimportant, slow tasks. Those speeds would certainly bring the edits in bot-like-speed territory).

Comment by Arbitrators:
Comment by parties:
There was a recent discussion about bots editing at higher speeds, and the consensus was rather clearly that it's ok as long as maxlag is respected. Based on that discussion, I reduced the throttle on my bots. This should probably be changed in the policy given the unanimous consensus. Pinging Headbomb, since I think myself directly editing the bot policy during this case is probably not wise. Note that the speeds at which bots edit are spectacularly irrelevant to whether Magioladitis' editing meets the definition of disruptive editing under WP:DISRUPTSIGNS. WP:MEATBOT is a side show. ~ Rob13Talk 18:42, 29 July 2017 (UTC)[reply]
@Beetstra: I do hope you realize that everyone is agreeing with you, which makes me question why you're arguing this. Every single argument about Magioladitis' editing has been centered around WP:DE, and yet you keep focusing on WP:MEATBOT. If the edits are disruptive, then the fact they're made at high volumes worsens the disruption. You're instead arguing that edits can be high volume/rate without being disruptive, which is a totally unrelated argument. See straw man. ~ Rob13Talk 18:46, 29 July 2017 (UTC)[reply]
@Beetstra: First, that's all that's needed; WP:DISRUPTSIGNS states editors who edit over the objections of others without seeking consensus is disruptive. By definition, that identifies his actions as disruptive. As far as what makes these specific edits worth objecting to, I would say it's a subjective combination of the non-time sensitive nature of the edits, the fact that bots were handling this task without the watchlist problem, and the fact that the edits were extremely minor. Yes, subjective - subjective in the sense of what spawned the complaint. But not subjective in the sense that once you get the complaint, you seek consensus. That consensus will determine whether the complaints are valid or not. Who knows, if he sought consensus, maybe he'd be doing the edits right now.

Our guidelines will never fully address every reason why one might want to complain about an edit or why an edit may be undesirable. There's no specific guideline that says I can't make one million edits in my user sandbox at bot-like speeds from my main account. And yet, if I did it, people would tell me to stop because it's clearly not something I should be doing. And I would have to listen or seek consensus, because that's how the project works per WP:DISRUPTSIGNS, WP:CONSENSUS, and just plain common sense. That's why we have those policies and guidelines which instruct us that it's disruptive not to be collaborative and to ignore others. ~ Rob13Talk 19:39, 29 July 2017 (UTC)[reply]

Comment by others:
The community doesn't need to define these things in policy or elsewhere. Defining what amounts to "high-speed editing" can't be done without running into the Sorites paradox. For when this is disruptive, we have WP:DE. Headbomb {t · c · p · b} 15:05, 29 July 2017 (UTC)[reply]
(Previously posted in a wrong section) As my bot's editing has been listed in this "disruptive" nature, I take particular offense. HasteurBot sends user talk page notices to users whose pages are within at maximum 1 month of being nominated for CSD:G13 that their page is in danger of being deleted, or to perform the process of nominating the page for G13 if the page is still eligible 1 month after it reminded the editor. During the CSD nomination phase the bot nominates up to 50 pages (I think) for deletion at one time so as to not disruptively flood the CSD nominations page. Now as to if the reminder process is disruptive, then the MassMessage bot is also disruptive and should be shut off. Wait, you like that one? Then I say to you, you're cherry picking an arbitrary list to invalidate the argument that the combined problem of Magioladitis' questionably authorized actions in addition to the rate in which these actions were being made (in line with Wikipedia:Fait accompli) constributed to the community enacting a restriction. Hasteur (talk) 18:06, 29 July 2017 (UTC)[reply]
@Hasteur: (moved from section below) I am not asserting that the bots are disruptive, the contrary actually. --Dirk Beetstra T C 17:49, 29 July 2017 (UTC)[reply]
And with that, I completely, utterly, fully agree with you: editing at very high speeds is NOT disruptive. However, that is what the whole initiation of that thread is about - it is even called 'high speed editing' .. it has NOTHING to do with high speed editing. --Dirk Beetstra T C 17:52, 29 July 2017 (UTC)[reply]
@Hasteur: You here exactly make my point, actually. --Dirk Beetstra T C 18:27, 29 July 2017 (UTC)[reply]
@BU Rob13: Because that editing speed / flooding the watchlist is the only disruption alleged. Which is totally unsubstantiated. You have not defined what is disruptive about his edits. You allege that he is disruptive because he did not listen to the complaints from editors about him flooding their watchlist .. --Dirk Beetstra T C 18:49, 29 July 2017 (UTC)[reply]
I can't speak of watchlists since I don't have one, but I can tell you that the massive amount of "robot poop" in the edit history of almost every article is a big annoyance in trying to examine the history to see how the article content evolved. 173.228.123.121 (talk) 04:07, 1 August 2017 (UTC)[reply]

Self analysis of "For the second topic ban, Magioladitis was not editing at an high edit rate" by Beetstra

edit
overall counts and comparisons to other editors
The following discussion has been closed. Please do not modify it.
  • In the thread, an edit speed of 13 was asserted,
  • On that day, 30 June 2017, he performed 1677 ISBN replacements
  • Highest speed in one minute: 19 edits at 11:25
  • Average editing speed while replacing: 10.5 edits per minute.
  • the 1677 edits were performed between 07:46 to 12:04 (258 minutes), overall 6.5 edits per minute.

In comparison (same edit sprees as mentioned in evidence, where cherry picked single minutes were shown over 12 edits per minute):

Note that none of these sprees go over the edit speed that is deemed 'fast' for bots performing crucial tasks ('once every 5 seconds', or 12 edits per minute).

Comment by Arbitrators:
Comment by parties:
@Beetstra: WP:BOTREQUIRE states most bots should edit once every 10 seconds, for 6 edits per minute. Urgent bots should edit once every 5 seconds, for 12 edits per minute. 13 edits per minute exceeds both. ~ Rob13Talk 15:12, 29 July 2017 (UTC)[reply]
WP:DISRUPTSIGNS states that one example of a disruptive editor is when one "continues editing an article or group of articles in pursuit of a certain point for an extended time despite opposition from other editors." Magioladitis did that to a tee on a massive scale. ~ Rob13Talk 15:15, 29 July 2017 (UTC)[reply]
@Beetstra: The bot policy states only that fast disruptive editing is against the bot policy. You have to meet the criteria at WP:DISRUPTSIGNS to be disruptive. As mentioned above, Magioladitis does. High-speed editing alone is not a problem. As soon as quality is compromised by speed or you get objections and edit over them, now you're disruptive and editing against bot policy. ~ Rob13Talk 15:28, 29 July 2017 (UTC)[reply]
That's one way to violate the bot policy, Beetstra. It also says you can't edit disruptively at high speeds. Either way, we can just ignore the bot policy – as I've wanted to do this whole case – and focus on WP:DISRUPTSIGNS. Disruptive editing is bad whether or not it violates the bot policy. ~ Rob13Talk 15:45, 29 July 2017 (UTC)[reply]
@Beetstra: High-speed editing alone is not disruptive. However, you can be disruptive while also happening to edit high-speed. If you are disruptive while also happening to edit high-speed, then WP:MEATBOT applies, but the high-speed business does not influence whether or not your edits are disruptive. By way of analogy, some dogs are brown, but being brown does not make something a dog. In other words, you can ignore the bot policy when determining whether an edit is disruptive. Then we come to the fact that you seem to be ignoring WP:DISRUPTSIGNS, which I've now linked four times on this page. I believe there's a reason for that, since it exactly describes Magioladitis' editing pattern in the first bullet. I pose the question to you: Does the first bullet of WP:DISRUPTSIGNS describe what Magioladitis did? ~ Rob13Talk 16:11, 29 July 2017 (UTC)[reply]
@Beetstra: Again, you're missing nuance. The objection was spamming watchlists in very high volume (not just high rate, but over a long period of time - not since I was a new editor with no appreciation for these issues have I done a semi-auto run of 1000+ pages as Magioladitis does regularly) when bots were at the very same time making these same changes with none of those problems. You may not think that is a problem. Community members who asked Magioladitis to stop did. Community members who then topic banned Magioladitis from continuing to do that also did. You're entitled to your opinion, but it is a sharp minority. ~ Rob13Talk 17:47, 29 July 2017 (UTC)[reply]
You're welcome to "hunt down" when I edit quickly and ignored requests to stop. That would be a null set, but you can certainly look. You're also welcome to contest any fast edits I make, at which point I'll go to an appropriate noticeboard and get consensus. Keep in mind that editors may be upset by pointless edits that spam watchlists but not be upset by very productive edits that spam watchlists. (Magioladitis' edits had a point, but they were pointless in the sense that bots could do the same thing without the watchlist spam.) I agree with you that it would be subjective to declare fast edits disruptive along those lines, which is why it's great news we don't do that! WP:DE exists to make things very "bright line" for us. When people start complaining, you stop and seek consensus if you don't already have demonstrated consensus, or you're acting disruptively. I have never been asked to stop and trampled over people anyway. I've stopped and sought consensus. WP:DE is a very well-established guideline. It's extremely clear. Note that my evidence section is very clear that Magioladitis did absolutely nothing wrong until people complained, he promised to stop making the edits, and then he went back on his word and resumed anyway. ~ Rob13Talk 18:37, 29 July 2017 (UTC)[reply]
@Beetstra: Per WP:MEATBOT, no I don't. Editing at high speeds is not inherently disruptive (ironically, something you argued elsewhere on this page). If you object to a specific edit I'm doing or the fact that a specific edit is flooding your watchlist, address that with me on my talk page and I'll seek consensus to ignore you. I very rarely edit semi-auto in mainspace these days; the comma business was a brief semi-auto test for a bot which was later discarded. But then again, that's totally out of scope. ~ Rob13Talk 18:48, 29 July 2017 (UTC)[reply]
I don't know what else to say to you, Beetstra. You've done an excellent job of rebutting an argument that no-one is making. The guideline is WP:DE, which involves making edits while ignoring the good-faith objections of others. You're talking about speeds at which those edits occurred, which isn't relevant to DE. ~ Rob13Talk 21:24, 29 July 2017 (UTC)[reply]
What is unreasonable is determined by consensus on Wikipedia. Clearly the community does not find the complaints unreasonable given that they topic banned Magioladitis for ignoring them. Unsurprisingly, you do not get to dismiss the community by fiat because you find them unreasonable. This will be my last comment on this topic. ~ Rob13Talk 21:48, 29 July 2017 (UTC)[reply]
@Beetstra: If an editor objected to me using Huggle to quickly revert vandalism, I would send them links to numerous discussions and evidence of consensus that supports using Huggle at high speeds to combat vandalism. Funny how consensus works. ~ Rob13Talk 22:16, 29 July 2017 (UTC)[reply]
Comment by others:
Editing at bot-like speeds is most definitely high speed editing in the sense of WP:MEATBOT. And while it's true that high-speed editing is not by itself disruptive, it certainly can be. All the drama surrounding this case should be more than enough evidence for this. Headbomb {t · c · p · b} 15:04, 29 July 2017 (UTC)[reply]
@Headbomb: two things:
  1. none of these speeds are bot-like, they are all slower than what is defined as suitable bot speed in the bot policy, and way slower than what many bots run at.
  2. Reread this sentence: "And while it's true that high-speed editing is not by itself disruptive, it certainly can be" .. you are a bot operator, programmer, so let's make it binary: "And while it's true that one is not zero, it certainly can be" .. That is totally arbitrary. --Dirk Beetstra T C 15:09, 29 July 2017 (UTC)[reply]
none of these speeds are bot-like, as member of BAG, I will tell you that those are absolutely bot-like speeds by any reasonable interpretation of "bot-like speed". If you want to bring numbers in, WP:BOTPOL says "Bots doing non-urgent tasks may edit approximately once every ten seconds, while bots doing more urgent tasks may edit approximately once every five seconds." Once every 10 seconds is 6 edits per minutes, once every 5 seconds is 12 edits per minutes.
And if you want to bring formal logic into this, read the following "While drinking lots of fluids in one sitting is not by itself hazardous to your health, it certainly can be." Headbomb {t · c · p · b} 15:21, 29 July 2017 (UTC)[reply]
You are talking to a chemist .. drinking lots of fluids in one sitting is by itself hazardous to your health.. --Dirk Beetstra T C 15:28, 29 July 2017 (UTC)[reply]
Again, depending on what "lots" mean. The analogy was chosen for a reason, and the vagueness of what 'lots' mean is one of those. Two glasses in one sitting is fine. A bathtub of is isn't. Headbomb {t · c · p · b} 15:38, 29 July 2017 (UTC)[reply]
No, that is not depending on what "lots" mean .. either high speed editing (whatever that is) is disruptive, or high speed editing is not disruptive. Or the statement has to be, "Two glasses in one sitting is fine. A bathtub of is isn't." or in other words: "an editing speed for humans of 5 edits per minute is fine, but an editing speed of 10 isn't" .. You and Magioladitis have editing sprees with 13 edits per minute mentioned. Neither of you have any errors in that specific editing spree. Were you disruptive, Headbomb? --Dirk Beetstra T C 15:45, 29 July 2017 (UTC)[reply]
If Magioladitis changes two ISBNs per day, no one would have batted an eye. But when you bring it to hundreds per day, you are flooding watchlists/histories/recent changes with edits, and you've entered WP:MEATBOT territory. Was I disruptive? Given none of my edits ever have attracted a complaint, I'd argue that no (WP:SILENCE). Are Magioladitis' edit disruptive? Evidently yes, given people keep complaining that the amount of improvement they bring they aren't worth the cost of reviewing them. Speed alone is not the factor, is what you do and how fast you do it that can cause issue. Creating stubs is fine, but when you want to create a lot of them in a short period of time, we ask that you to go through WP:MASSCREATION to ensure the community isn't all-of-a-sudden swamped with an massive and unnecessary reviewing burden. Headbomb {t · c · p · b} 21:49, 29 July 2017 (UTC)[reply]
@Headbomb: With this and this you come to hundreds of 'identifier linking' edits in a day as well. So yes, you were there flooding watchlists/histories/recent changes with edits (and at the moment, someone is flooding recent changes with AWB edits). And Magioladitis (and your) edit summary is honest, I don't need to review all of your edits (note: you have quite a couple of cosmetic edits in them (this was cosmetic, right when you do 10 edits a minute), and since we are in MEATBOT territory ..). People would have been full right to complain about them: you were doing a massive number of edits using AWB at sometimes high speeds (8 edits a minute, well above 1 edit every 10 seconds for a slow bot). And I don't have to argue to you that can do that very well with a bot. --Dirk Beetstra T C 22:06, 29 July 2017 (UTC)[reply]
I'm fully aware of the edits I've made. While they are similar to templatifying ISBNs, they are also giving links where none existed before, adding substantial functionally and accessibility to articles. Templatifying ISBNs does not provide much if any additional functionality, the links are there before using templates, and are there after using templates. It's done because we're transitioning to a new system and deprecating the old one. This is a completely non-urgent task that needs to be done "at some point" before June 2018.
Additionally, I'm currently developing the regex for such a bot. That involves a substantial amount of testing, seeing which regex is safe, which isn't, and so on, and that can only be done semi-automatically. There are a bunch of corner cases, for a bunch of different identifiers. Offloading a significant chunck of those edits is the whole point of CitationCleanerBot 2. Headbomb {t · c · p · b} 22:30, 29 July 2017 (UTC)[reply]
@Headbomb: You don't need to defend these edits to me. I am fully aware of what you are doing, and admire what you are trying to achieve. And I would not consider to ask you to stop those edits. Especially not for 'flooding my watchlist' (which these edits obviously do), as I think that that is a completely irrelevant, unreasonable, and subjective objection. And what would I achieve with that? (q, is 'templatifying' a word?) --Dirk Beetstra T C 23:03, 29 July 2017 (UTC)[reply]
@BU Rob13: .. Yes, so are we now defining that one editor who is editing at >12 edits per minute is disruptive, or are we defining that all editors who are editing at >12 edits per minute are disruptive. This point is completely arbitrary. Wikipedia:Don't call the kettle black. --Dirk Beetstra T C 15:20, 29 July 2017 (UTC)[reply]
@BU Rob13: "As soon as quality is compromised by speed" .. there was no comprimise in the quality. The only assertion that is in the second topic ban discussion is that the high speed editing is flooding someone's watchlist. --Dirk Beetstra T C 15:30, 29 July 2017 (UTC)[reply]
@BU Rob13: Good, let's try. How is high speed editing (without errors) disruptive, BU Rob13? I am going to ask you the same question as Headbomb: you and Magioladitis have editing sprees with 13 edits per minute. Neither of you have errorrs in that specific editing spree. Where you disruptive, BU Rob13? --Dirk Beetstra T C 15:52, 29 July 2017 (UTC) (note: BU Rob13, your actual spree there was 19, Magioladitis 13 .. --Dirk Beetstra T C 15:55, 29 July 2017 (UTC))[reply]
@BU Rob13: No, because the only thing I see people complain about in Wikipedia:Administrators'_noticeboard/IncidentArchive958#User:Magioladitis high speed editing is that he is editing at high speed (which is not disruptive), and that he is flooding watchlists (and if he is flooding watchlists at 13 edits per minute, so were you and Headbomb). It would be tendentious, per WP:DISRUPTSIGNS, if there was opposition against the edits .. and there is none. We already have established that they are not COSMETIC, they are a substantial change. So I am sure that you can show me editors who show opposition against the content of each edit. --Dirk Beetstra T C 16:59, 29 July 2017 (UTC)[reply]
As my bot's editing has been listed in this "disruptive" nature, I take particular offense. HasteurBot sends user talk page notices to users whose pages are within at maximum 1 month of being nominated for CSD:G13 that their page is in danger of being deleted, or to perform the process of nominating the page for G13 if the page is still eligible 1 month after it reminded the editor. During the CSD nomination phase the bot nominates up to 50 pages (I think) for deletion at one time so as to not disruptively flood the CSD nominations page. Now as to if the reminder process is disruptive, then the MassMessage bot is also disruptive and should be shut off. Wait, you like that one? Then I say to you, you're cherry picking an arbitrary list to invalidate the argument that the combined problem of Magioladitis' questionably authorized actions in addition to the rate in which these actions were being made (in line with Wikipedia:Fait accompli) constributed to the community enacting a restriction. Hasteur (talk) 17:44, 29 July 2017 (UTC)[reply]
@Hasteur: I am not asserting that the bots are disruptive, the contrary actually. --Dirk Beetstra T C 17:49, 29 July 2017 (UTC)[reply]
And with that, I completely, utterly, fully agree with you: editing at very high speeds is NOT disruptive. However, that is what the whole initiation of that thread is about - it is even called 'high speed editing' .. it has NOTHING to do with high speed editing. --Dirk Beetstra T C 17:52, 29 July 2017 (UTC)[reply]
This was for the wrong section Hasteur (talk) 18:06, 29 July 2017 (UTC)[reply]
Also my reply to you then. --Dirk Beetstra T C 18:09, 29 July 2017 (UTC)[reply]
@BU Rob13: Your 70 edits I mention here in the analysis were part of a set of at least 1000 edits, probably close to 2000. And you misunderstand me as well - I am annoyed by these edits, I am annoyed by the edits from the bots, and I was furious at the editing sprees of the bots that removed the interwikis when WikiData started up. And then that bloody bot that delivers the Signpost to all my friends .. there goes my watchlist.
But there is nothing to complain about - as we have not set a standard. Or I have to be consistent. Any bot editing over 12 edits per minute at the moment is in violation of the Bot policy (except if there is explicitly a defined higher rate agreed in the BRFA), and any (non-bot) editor that shows up with a significant number of their AWB edits in my watchlist should require a complaint from us (especially when they actually edit over 12 edits per minute, and likely even over 6 edits per minute). It has been Pot and Kettle from the beginning. And do mind, I have edited with AWB over the speed - I am just as black.
So, I totally agree that we should have a consensus of what editing speed can be deemed 'disruptive', even when it is without errors - but we don't, and at that time that complaint is without merit (if it were not a massive WP:POINT violation, I would hunt you down for every time you pass 6 edits per minute in mainspace). --Dirk Beetstra T C 18:20, 29 July 2017 (UTC)[reply]
@BU Rob13: you are editing in mainspace at said speeds, or faster, and whether or not there is consensus for the edits behind it does not matter. At over 6 edits per minute you edit at bot-like speeds ánd are flooding my watchlist, and better file a BRFA to have a bot account do it. --Dirk Beetstra T C 18:44, 29 July 2017 (UTC)[reply]
@BU Rob13: I did not mean that to your person per sé, I am sure that at this very moment someone is running AWB at an edit speed of 7 edits per minute or higher. It would be equally WP:POINTy of me to post to their talkpage and tell them that they are being disruptive because they are flooding my watchlist (if I would take 10, I am sure a couple would plainly tell me to f*ck off). However, to most of us, including me, an edit rate of 13 edits per minute (even 20) is reasonable (you allude to that in the second topic ban discussion) - I run AWB, if you know what it should do one look at a diff is more than enough, 1-2 seconds an edit, close to 30 a minute (I had 13 per minute when AWB was skipping 75% of the pages in my list because it was not clear enough whether they were simple enough to determine to fall within my criteria! I mainly sat there waiting for the next edit to approve. I think I stepped back a couple of times - 'wait, did I see that right - no, revert self'). User:Hasteur takes offense in my comments/evidence that the edit rate of their bot is disruptive (their interpretation of my evidence). I read that they means that it is unreasonable that I say that their edit rate of 39 is 'too fast', that their edit rate of 39 is disruptive. And I would fully agree with them: it is unreasonable to say that. I do not think it is a legitimate concern, especially because there is no benchmark. It is a perception, it is arbitrary. But I will presume that it is up to the ArbCom to determine that. --Dirk Beetstra T C 20:49, 29 July 2017 (UTC)[reply]
I clicked 'recent changes', got a message, saw an AWB edit as 5th edit, clicked contribs on the editor, and see that in the last minute(11:50) they made 11 AWB edits. --Dirk Beetstra T C 20:55, 29 July 2017 (UTC)[reply]
Another one through recent changes, 8 edits (a bot operator!). --Dirk Beetstra T C 21:05, 29 July 2017 (UTC)[reply]
@BU Rob13: The point is, "while ignoring the good-faith objections of others" .. it is a good-faith objection, but it is also an unreasonable objection, it is a subjective objection. If you revert vandalism, and an editor comes to you to say: 'I object, you reverted that within 12 seconds' .. what do you then say? 'Oh, that is reasonable, I'll revert again and we can discuss it first. Tea or coffee?'?
By the way, I spotted #3 already. Keeping tabs. --Dirk Beetstra T C 21:33, 29 July 2017 (UTC)[reply]
@BU Rob13: You evaded the question. --Dirk Beetstra T C 22:08, 29 July 2017 (UTC)[reply]
@BU Rob13: Indeed, it is funny how WP:CONSENSUS works. --Dirk Beetstra T C 22:23, 29 July 2017 (UTC)[reply]

Comments regarding "General remarks (2)" by Headbomb

edit

You correctly correct me that the actions are even mostly by WP:BAG members. Going through the individual elections (where members were elected, some predate that process), most of the !votes are by bot operators and fellow WP:BAG members. It may be well publicized, nonetheless also in the BAG elections there is not a lot of participation by community. In your own election, I see quickly 7 bot operators/BAG members.

All in all, it still does not show that the community outside of the bot operators has so many problems with Magioladitis. --Dirk Beetstra T C 17:47, 3 August 2017 (UTC)[reply]

Comment by Arbitrators:
Comment by parties:
Comment by others:
Your repeated attempts to delegitimize process and policy are both misplaced and futile. The initial objections were made by non-bot ops, only then did the bot community stepped in to police their own. But even if only the bot community had objections, we are still part of the Wikipedian community. Plenty of non-bot ops have made complaints about Magioladitis' edits (including a couple of non-registered users), and plenty of non-bot ops (including IPs) have participated in every step of the process that appointed people to BAG, or made policy what it is today. Headbomb {t · c · p · b} 18:39, 3 August 2017 (UTC)[reply]

Comments regarding “Re: Magioladitis was not restricted from automated editing from their main account” by Headbomb

edit

Headbomb states: if the claim here is that Magioladitis was not restricted from automated editing, rather than semi-automated editing, that's obviously untrue. Everyone is restricting from running automated processes from their own account. See WP:BOTACC/WP:BOTBLOCK in particular. (My bolding)

  • the claim is regarding semi-automated, I’ll update that. I must already have done so.
  • WP:BOTACC & WP:BOTBLOCK do not restrict editors from running fully automated tasks from their main account. I invite the committee to review the faster editing accounts in my evidence regarding Magioladitis, editing speed, where editors were editing at speeds that too fast for proper review (note that the minutes are cherry picked, they are for some edits part of much larger sets of edits at similar speeds; e.g. 400 edits in 15 minutes, avg. 26 per minute, obviously fast clicking without reviewing and hence indiscernable from fully automated, and hence obviously a task better performed by bot). Editors are regularly running tasks that are indiscernable from fully automated, tasks that can just as well be done by bot from their main account. If everyone is restricted like that, then why are we here with Magioladitis?
Comment by Arbitrators:
Comment by parties:
Dirk, many AWB edits require editor eyes on them to skip bad edits. They can be done at high speeds only while editors are skipping many of the edits. This review can often be fast, which is the whole reason we don't sanction purely for speed. The issue is editing over the complaints of editors who have holistically reviewed the circumstances surrounding these edits (e.g. the fact they can be done by bot and the fact they're at extraordinarily high volumes) and determined that the bad outweighed the good. That's where WP:CONSENSUS and WP:DE comes in. It seriously bothers me that an admin appears not to understand that it's inappropriate for an editor to edit over the vocal disagreement of many other editors without any demonstrated consensus for doing so. ~ Rob13Talk 07:53, 20 August 2017 (UTC)[reply]
Dirk, I'm not saying 50 edits have 100 skips. I'm saying maybe 50 edits have 2-5 skips. That's definitely doable via semi-auto, but an error rate of 4-10% is unacceptable for bot edits. I've done it myself in the past for certain tasks. Further, even if your premise were correct (it is not), "other people violate policy therefore I can violate policy" is perhaps the worst argument I've ever seen from an administrator. ~ Rob13Talk 17:42, 20 August 2017 (UTC)[reply]
@Beetstra: Please find me the diff where I say Magioladitis was editing fully automated or redact your inaccurate statement. (Hint: It doesn't exist.) Further, please find me the diff where I say Magioladitis' behavior violates policy because it is flooding in isolation of the complaints or redact your inaccurate statement. (Second hint: It doesn't exist.) The strawman arguments are getting tiresome. Lastly, please find me a policy, guideline, or RfC that states you can ignore WP:CONSENSUS and WP:DE if the complaints have to do with watchlist flooding or redact your inaccurate claim. (Third hint: Still doesn't exist.) ~ Rob13Talk 20:27, 20 August 2017 (UTC)[reply]
First comment: This page is used like a talk page. This is serious abuse of the procedure followed here. Second comment: I have demonstrated that the second run was ot done by bots. So revisiting this concept is nonsense. Third comment: I have provided an example to Headbomb of a seris of 5,000 edits done in a few hours bu a single editor, using AWB. Thiss task is already assigned to bots and can easily done by bots. So, the argument that since bots do something, editors should not do it is void. -- Magioladitis (talk) 08:03, 20 August 2017 (UTC)[reply]
I think that we now have to understand each other. Headbomb says that "the community may decide..." but till now the community lack this mechanism. This procedure is done so far by editors running tasks from main accounts and then at some point someone proposes a bot to do it, there are some talks in talk pages, etc. In my run to fix ISBNs, people like Headbomb should be aware of what I was doing and that the bots were not fixing that part. So, they could have helped explain to the non-experts that the ISBN conversion is desirable and that the edits I was doing needed some limited editor attention. There was also the alternative to have a better ISBN coordinance whch never happened as I proved to the Evidence page. Instead of that, the experts joined the complaining side which is disappointing.
For the Wikiproject tagging, when a bot dos it I have helped in forming some very tight rules to help the bot procedure which involes Wikiproject notification etc. Still, ofcourse, many people do tagging using AWB in hgh speeds from their main account. -- Magioladitis (talk) 12:34, 20 August 2017 (UTC)[reply]
The administrator role, and especially the role of a BAG member, should be to explain to those complaining the neccessity of the given edits. Because, mass editing always makes people uncorfortable and not everybody is aware of the decisions taken in various places and not everybody is aware fo the difficulty to create a flawless bot task. My editing should have been protected and drama should have been avoided. I various cases admins and BAG members read complains in talk pages and help to reduce drama. This should have happened in my ISBN fixes case too. -- Magioladitis (talk) 20:32, 20 August 2017 (UTC)[reply]
Comment by others:
Concerning restrictions on automated editing on non-bot accounts, I'll directly quote from policy:
  • "Contributors should create a separate account in order to operate a bot."
  • "A block may also be issued if a bot [...] is logged in to an account other than its own" and "Administrators blocking a user account suspected of operating an unapproved bot [...] should block indefinitely."
As for why we are here, I'll point you to WP:MEATBOT, WP:DEADHORSE, WP:IDIDNTHEARTHAT and WP:CONSENSUS. The difference between others and Magioladitis, is that others either 1) have consensus for the tasks they do, 2) stop when challenged to discuss the issue, and 3) don't try to wikilawyer their way around WP:MEATBOT and instead address the challenge by directly addressing the complaint by rectifying their behaviour, or getting consensus before resuming the task. Headbomb {t · c · p · b} 18:08, 17 August 2017 (UTC)[reply]
So, you concur that no-one is following said policy - people are running AWB/JWB continuously at speeds that do not show human interaction and hence can easily be done by a bot. You still fail to explain why you single out Magioladitis, and why we are here - if he was running an unapproved bot, he should have been blocked indefinitely, and there is no need for the dramah of an ArbCom. --Dirk Beetstra T C 06:07, 20 August 2017 (UTC)[reply]
I most certainly do not. Headbomb {t · c · p · b} 12:02, 20 August 2017 (UTC)[reply]

@BU Rob13: if you are accepting 50 edits a minute, then your argument that the editor is rejecting another 100 is not making much sense (while sustaining that speed for hours). Your argument is that such editors may very well properly review every edit they make does not make sense. Moreover, that also means that Magioladitis was reviewing every single edit, because he was editing at said speed (and may very well have needed it). My examples include a handful of editors who perform fail-safe edits (as in, adding a template to the top of a page, tagging images), which is very easy to do blind. There is NO need to review, there is nothing that can go wrong there (believe me, I know my regex), that is work that can easily be done by a bot. And that is exactly what you guys are accusing Magioladitis of, not listening to the concerns. No-one is following the bot policy in these, AWB is on a regular basis used for edits at high speed ánd with high volume for edits that can be done, and even are already done, by bot. It is a concern, but not a legitimate one. --Dirk Beetstra T C 10:53, 20 August 2017 (UTC)[reply]

@BU Rob13: But you are absolutely sure that Magioladitis was doing those 13 fully automated .. not skipping 1 every 2 minutes. Your knife cuts both ways, BU Rob13. Your assumption that user:Koavf is skipping 2-3 a minute also goes for Magioladitis. No, BU Rob13, if Magioladitis is flooding then Koafv is flooding, you are flooding, and I am flooding. Either all of those edits have to be done manually on a bot account, or they are all fine. ‘You are flooding my watchlist’ is not a legitimate concern, and hanging Magioladitis for it is plainly wrong.

Forgot to sign, and not finish. My argument is not a WP:OTHERCRAPEXISTS argument, the whole point of the argument is that there is no community consensus.. the community consistently ignores the bot policy. —Dirk Beetstra T C 20:11, 20 August 2017 (UTC)[reply]

Do you draw complaints with your edits? If yes stop, discuss, get consensus before resuming. If no, you can assume consensus. Is that so hard to get? Show me where Koavf 1) got complaints and 2) continued to edit without addressing them. THEN you'll have an actual argument which goes beyond 'but what about X'. Everyone else you've mentionned got no complaints for their edits. Magioladitis got them. This is why Magioladitis needed to stop, and that Koavf was in the clear. Headbomb {t · c · p · b} 21:10, 20 August 2017 (UTC)[reply]
@BU Rob13: Exactly, BU Rob13, there is no evidence that these edits were better done on a bot account, there is no Policy or RfC that talks about flooding watchlists .. there is nothing. But there is WP:CONSENSUS. —Dirk Beetstra T C
@Headbomb: Yes, and I continued. See WP:CONSENSUS. —Dirk Beetstra T C 03:16, 21 August 2017 (UTC)[reply]
And to be clear, the concerns addressed at me were policy based concerns, up to AN/I, ‘and by multiple editors. —Dirk Beetstra T C 03:24, 21 August 2017 (UTC)[reply]
I, as opposed to Magioladitis in the concerned AWB run, even made mistakes in the edits I did save. —Dirk Beetstra T C 03:32, 21 August 2017 (UTC)[reply]
I expect you'll resign your bit then. Headbomb {t · c · p · b} 05:11, 21 August 2017 (UTC)[reply]
No, no need. The point is that this happens more often then you expect. Your argument is that no-one complains on the hundreds and hundreds of high speed (as defined by the bot policy) and high volume edits that are performed by non-bot accounts, except on this one of Magioladitis. Complaints are not necessarily a reason to stop, complaints do not necessarily trump consensus. Moreover, what you (pl.) advocate is a bureaucratic reading of policies, and a failure to adhere to WP:IAR. --Dirk Beetstra T C 08:52, 21 August 2017 (UTC)[reply]

@Magioladitis: - the Arbitration Committee has implemented here a section for analysis of evidence. That is what we here do. We have evidence from editors, and we are discussing said evidence. That is hence not an abuse of this page/procedure. --Dirk Beetstra T C 10:53, 20 August 2017 (UTC)[reply]

On the series of 5000 AWB edits. Yes this is a task that some bots are approved to do, and something that can indeed be done by bots. However, depending on how the list was compiled, human-review is often needed to determine whether or not an article should be tagged by a certain WikiProject's template, or if it should be double-tagged by two projects. In addition, and this is the important thing, that tagging run drew no complaints, made high-impact substantive changes to the page, and (I presume) was done at the request of the Wikiproject in question, or by a WikiProject member.
To be clearer, the community may decide that A) some edits should be done by humans rather than by bots (e.g. WP:CONTEXTBOT). It may also decide that B) some edits can both be done by bots and humans. And it may decide that some edits C) should be done by bots rather than by WP:MEATBOTS. Headbomb {t · c · p · b} 11:46, 20 August 2017 (UTC)[reply]
The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.