Wikipedia:Arbitration/Requests/Case/Rich Farmbrough/Workshop

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

Case clerk: TBD Drafting arbitrator: TBD

The purpose of the workshop is for the parties to the case, other interested members of the community, and members of the Arbitration Committee to post proposed components of the final decisions for review and comment. Proposals may include proposed general principles, findings of fact, remedies, and enforcement provisions, which are the four types of proposals that can be included in the final decision. The workshop also includes a section (at the page-bottom) for analysis of the /Evidence, and for general discussion of the case.

Any user may edit this workshop page. Please sign all suggestions and comments. Arbitrators will place proposed items they believe should be part of the final decision on the /Proposed decision page, which only Arbitrators and clerks may edit, for voting, clarification as well as implementation purposes.

Motions and requests by the parties

Increase evidence length

1) Because Rich Farmbrough (talk · contribs) has made over 1,000,000 edits himself and with his bots, the evidence length in words and diffs should be increased because of the number of edits that need to be gone through and scrutinized.

Comment by Arbitrators:

As the lead drafter in this case, I think the parties should be given reasonable leeway to present their evidence. As I've stated on-wiki more than once (most recently on the proposed decision page in the last case we've voted on), I've never favored strict enforcement of rigid limits, which can have the unwanted effect of cutting the Committee off from relevant evidence. On the other hand, we can't draw much value from lengthy monologs or unwieldy collections of undifferentiated diffs, either. I would ask for input from the parties and other case participants on what length they think would be reasonable. Newyorkbrad (talk) 21:55, 5 April 2012 (UTC)

Comment by parties:

Comment One problem about the number of diffs has struck me forcefully - and it applies not just to this case. Where a contention is made about a pattern of behaviour, assuming, for example 5 diffs are cited to support that, assuming they actually support the contention, then a much larger number of diffs is required to establish that this is not a pattern. Rich Farmbrough, 22:49, 6 April 2012 (UTC).

Oppose scaling the whole thing up just scales the drama and results in more TLDR. I would rather scale it down in absolute terms. Rich Farmbrough, 22:49, 6 April 2012 (UTC).

Honestly, I don't really care much one way or the other. I think I can cover most of the issues I wanted to bring up in 500/50, and others I note have already been raised by others. OTOH, adding more wouldn't hurt, since Rich does have a rather long list of edits to review. Hersfold ^(t/a/c) 22:10, 8 April 2012 (UTC)

Comment by others:

Support as proposer. Whenaxis (contribs) ^{DR goes to Wikimania!} 00:50, 5 April 2012 (UTC)

Template

2)

Comment by Arbitrators:

Comment by parties:

Comment by others:

Template

3)

Comment by Arbitrators:

Comment by parties:

Comment by others:

Template

4)

Comment by Arbitrators:

Comment by parties:

Comment by others:

Proposed temporary injunctions

All bots indefinitely blocked

1) All bots owned (whether active or not) by Rich Farmbrough are to be indefinitely blocked to ensure no further harm is done to the community until a final decision is made by the Arbitration Committee. These bots include: SmackBot (talk · contribs), Smackbot (talk · contribs), Mirror Bot (talk · contribs), Chron Bot (talk · contribs), Translate Bot (talk · contribs) and Femto Bot (talk · contribs).

Comment by Arbitrators:

I agree with Hersfold. If the case is still pending around the time the block is to expire, this can be addressed, preferably informally without needing a temporary injunction. Two other points: (1) if there is a bot that undisputedly is currently operating without any problems (I haven't checked), it doesn't need to be blocked for the sake of blocking it; and (2) I could see an argument for extending Rich Farmbrough's unblock permission to allow him to edit actual article content, as opposed to edits involving any form of "gnoming" or automation whatsoever (although I don't believe he has asked for this as yet and although I'm speaking only for myself on this point). Newyorkbrad (talk) 22:05, 5 April 2012 (UTC)

Comment by parties:

His bots aren't supposed to be editing until the time his block would have expired anyway, another three weeks at least. This is redundant. Hersfold ^(t/a/c) 03:50, 5 April 2012 (UTC)

Comment by others:

Support as proposer. It's either or for these two options, not to be used in combination with each other. Whenaxis (contribs) ^{DR goes to Wikimania!} 01:10, 5 April 2012 (UTC)

This would seem to be redundant and punitive besides. These bots aren't editing and Courcelle's comment has already established that they shouldn't be editing. Until and unless they edit, good faith should be assumed and they should remain unblocked. — madman 02:20, 5 April 2012 (UTC)

I wasn't aware of these already imposed blocks against Rich and his bots so I proposed these. Whenaxis (contribs) ^{DR goes to Wikimania!} 22:11, 5 April 2012 (UTC)

That's what I figured; just letting you know. — madman 03:45, 6 April 2012 (UTC)

• Clerk Comment Femto Bot (talk · contribs) keeps on editing --Guerillero | My Talk 05:03, 5 April 2012 (UTC)
• BAG note FemtoBot has never been problematic and several Wikiprojects depend on it. Please unblock that one / allow for FemtoBot to be exempt from temporary injunctions. Headbomb {talk / contribs / physics / books} 12:48, 5 April 2012 (UTC)
  • Oppose BAG should not approve closed source bots absent compelling reason, especially if the bot is doing something important and there is no succession plan in place in case the main operator has to stop running it for whatever reason. This is the consequence, BAG shares responsibility. 67.119.15.149 (talk) 05:14, 6 April 2012 (UTC)
  • I intend to unblock the bot; I've let Rich know I'd be more comfortable doing so if he would suspend its task 0 ("Creating needed monthly clean up categories"). I think it may fall within the scope of the concerns that instigated the case but none of the others do. As for the BAG being responsible for forcing open-source licenses upon developers and ensuring the high-availability of all bots on the English Wikipedia: Nonsense. — madman 13:46, 6 April 2012 (UTC)
    • BAG can't force any sort of licenses on developers any more than Wikipedia can force Creative Commons licenses on writers. People who don't want to use those licenses are free to not submit stuff. Considering what happens if a bot or its operator becomes unavailable should be part of any sensible approval process, and BAG is incompetent if it doesn't think about that. BAG should not approve bots (regardless of license) whose subsequent unavailability will cause disruption unless it's got a way to keep them running if something happens to the main operator. Wikipedia also should not depend on closed bots doing anything important, since those bots in a sense become part of the server, and letting them be closed interferes with users' rights to fork wikipedia. If you want to submit server patches, they have to be GPL, and bots that users rely on should be the same.

      I'm not too concerned about Rich going nuts with the bot, so unblocking it in this particular instance won't be an immediate disaster. However, generically, letting a blocked editor operate a bot is a bizarre concept. The only legitimate way do it is turn control of the bot over to someone else, and transfer the approval. 67.119.3.15 (talk) 18:14, 6 April 2012 (UTC)

      • I agree that if an editor is blocked then the bots should also be blocked. Whether WikiProjects or something else relies on the edits done by said bot are important in this discussion but are not relevant to whether they should be running. If the community doesn't trust the editor and don't want them editing then that is a consideration that must be kept in mind when blocking them. You can't get the milk for free if you give away the cow so to speak. 71.163.243.232 (talk) 01:22, 10 April 2012 (UTC)

BAG requests on hiatus

2) All requests that are not yet closed at the Bots Approval Group (BAG) are to be placed on hiatus until a final decision is given by the Arbitration Committee. These requests include: Helpful Pixie Bot Request #47, Helpful Pixie Bot Request #48 and Helpful Pixie Bot Request #50.

Comment by Arbitrators:

Per comments below, It looks like the BAG has already addressed this issue. Newyorkbrad (talk) 22:28, 5 April 2012 (UTC)

Comment by parties:

Again, Rich can't participate in the BRFAs for another three weeks or so anyway. Hersfold ^(t/a/c) 03:50, 5 April 2012 (UTC)

Comment by others:

Support as proposer. It's either or for these two options, not to be used in combination with each other. Whenaxis (contribs) ^{DR goes to Wikimania!} 01:10, 5 April 2012 (UTC)

This has already been done; the requests have been removed from Wikipedia:Bots/Requests for approval with no presumption of what action will be taken upon closure of this case. — madman 02:23, 5 April 2012 (UTC)

I wasn't aware of these already imposed blocks against Rich and his bots so I proposed these. Whenaxis (contribs) ^{DR goes to Wikimania!} 22:11, 5 April 2012 (UTC)

Allow to work on own userspace

3) User:Rich Farmbrough to be explicitly allowed to work on pages in his own accounts user-space, despite nominal block. In particular to allow these to be used to gather evidence. Rich Farmbrough, 22:44, 6 April 2012 (UTC).

Comment by Arbitrators:

Rich, why not use your talk page or a word processor? I am not inclined to allow this request, though I am inactive on this case and would not vote either way. AGK [•] 23:49, 6 April 2012 (UTC)

People get pissed off when I do stuff using my talk page. And a word processor can't link to WP diffs, or render wikitext properly. Rich Farmbrough, 02:39, 8 April 2012 (UTC).

Comment by parties:

OK it looks like expecting anything to be passed here before the last day of the evidence phase was a triumph of hope over experience. Rich Farmbrough, 23:56, 17 April 2012 (UTC).

Comment by others:

Consider this an observation from the outside both on the request and the response to an Arb:

• The wiki-mark-up can be keyed in something as simple as Notepad with only the ability to "preview" is not present.
• The text can be prepared in a word processor to allow an editor to check spelling and/or grammar with the software. This can be copied to Notepad or the like to strip it down to minimal characters and the wiki-mark-up - templates, brackets, etc - can be added at that point.
• Link information, either a URL or "page#section", can be copied into Notepad.
• The result in Notepad is plain text and can be copied into a Wikipedia edit window.
• A "Show Preview" button exists and can be used to verify that mark-up typos or errors are not present.

With that, it seems unneeded for a special dispensation is needed to in the limitations of the block lift. - J Greb (talk) 04:22, 8 April 2012 (UTC)

If someone is unblocked for the purposes of participating in an ArbCom case it seems only reasonable to allow them to assemble evidence in their userspace. I don't see any benefit to denying this request. It seems silly to make him jump through hoops to properly format his evidence. 28bytes (talk) 04:25, 8 April 2012 (UTC)

Not only that, but there's literally nothing to be gained, nor any disruption to be prevented, by disallowing personal user space edits. In fact, I'd go even farther and argue that the only real edits that really need to be curbed are bot and bot-like edits. Headbomb {talk / contribs / physics / books} 06:40, 8 April 2012 (UTC)

I agree with Headbomb and 28bytes, I see no reason why Rich should not be allowed to edit in his own namespace. 71.163.243.232 (talk) 01:39, 10 April 2012 (UTC) This user self-identifies as user:Kumioko, editing in evasion of the block on his account. A fluffernutter is a sandwich! (talk) 18:47, 10 April 2012 (UTC) Matter resolved on my talk; Kumioko is allowed to edit as an IP, so this isn't block evasion. A fluffernutter is a sandwich! (talk) 19:41, 10 April 2012 (UTC)

Just to clarify I am not socking, The IP 71.163.243.232 is me (Kumioko) and IP's 138.162.8.57 and 138.162.8.58 might be if I edit from work depending. The last 2 are used by thousands of people so its not always me. I used to put my name in parens but was told its not appropriate. Fluffernutter is just being dramatic adn this is off topic so I'll leave a message on their page. 138.162.8.57 (talk) 19:20, 10 April 2012 (UTC)

I'm fine with it too (I posted the earlier comment about keeping the bot blocked). The block is to prevent disruption because Rich has trouble telling bot edits from regular edits, but restricting him from his user space is unnecessarily annoying. Alternatively, he could be permitted to make some sub-pages within the arb case and work in those, so it's clearly part of the proceedings. 67.117.131.84 (talk) 05:42, 10 April 2012 (UTC)

Questions to the parties

Questions from Kirill Lokshin

1. With regard to Rich Farmbrough's use of bots and other automation tools, are his automated edits generally (a) both substantively correct and authorized by the appropriate community process, (b) substantively correct but not authorized by the appropriate community process, (c) not substantively correct but authorized by the appropriate community process, or (d) neither substantively correct nor authorized by the appropriate community process? Kirill ^{[talk] [prof]} 03:38, 18 April 2012 (UTC)

2. With regard to Rich Farmbrough's conduct when his use of bots and other automation tools is questioned by other editors, are his responses generally (a) both technically correct and civil, (b) technically correct but uncivil, (c) not technically correct but civil, or (d) neither technically correct nor civil? Kirill ^{[talk] [prof]} 03:38, 18 April 2012 (UTC)

3. With regard to the existing restrictions on Rich Farmbrough, have these restrictions been effective in preventing recurrences of the conduct they were intended to prevent? If not, was it because (a) the restrictions were poorly written, (b) the restrictions were inadequately enforced, (c) the conduct continued despite enforcement, or (d) some other reason? Kirill ^{[talk] [prof]} 03:38, 18 April 2012 (UTC)

Hersfold's replies

1. Due to the large number of edits he's made, I'd have to say that in general, his edits do meet the criteria in A. However, he frequently makes edits which fall under B as part of these otherwise acceptable tasks (occasionally as part of a separate task, as with the recent category creations); edits that in full or part violate his editing restrictions as he hasn't sought out any sort of approval to make them. In the cases where it is a fully separate task, the number of D-type edits is not insignificant.
2. In my experience, D. While his technical knowledge (with respect to programming) is unquestionably strong, his interpretations of what is and isn't appropriate with respect to Wikipedia policies and his restrictions often differs from the community's consensus. As for civility, it seems as though he believes anyone raising such concerns is directly attacking him and not acting in good faith, and so he responds in kind. When blocked, his behavior is not unlike a temper tantrum.

Funnily enough, Elen had said she would block me for a day, a week a month and a year, in that order, should she do so. She blocked me over the technically incorrect claim of "defamatory" categories - for a month. My "temper tantrum" consisted of a humorous response, pointing out the flaws in the argument. (Note: she never responded.) I'd say that was pretty calm, for a prima facie incorrect block of an excessive length, contrary to previous statements of the blocking admin. Rich Farmbrough, 03:57, 25 April 2012 (UTC).

1. Evidently not, and I'd say due to both B and C. There were a number of violations where Rich was not blocked; at the same time, the high number of violations indicates that he isn't heeding the restrictions. Hersfold ^(t/a/c) 04:01, 18 April 2012 (UTC)

Rich Farmbrough's replies

1. a. Bot edits: I would have to say that they are substantively correct and authorised. There may be minor glitches, such as edit conflicts, or bugs with new code.
1. b. Manual edits made with AWB. These tend to be shorter runs, they are therefore more prone error as a percentage. Nonetheless the error rate overall is low, if we are talking about actual errors that mean something is not working after the edit that was working before.
2. It is certainly true that I may have become irritable when certain editors just keep turning up and do not engage in proper discussion. There is more to civility than please and thank-you. It is perhaps noteworthy that I asked for a mutual interaction ban with User:Fram several times. I suspect I can handle him now, but it does not mean that it has been a pleasant experience- and indeed, the more successful I am in dealing with him, the more he takes stuff to AN, AN/I and makes unwanted comments on BRFAs.
3. No they are not effective. The reason is that I stopped doing the things they were meant to prevent anyway. What we have now are just a couple of editors chasing up the ERs for the sake of it. If there was a valid concern then we could do something about it, but all there is is an attempt to rules-bully. If there were a valid concern, I would have hundreds of editors all over my talk page, instead of getting complaints only when they are initiated by something posted by Fram or CBM. (I don't mean here real bug reports, I get those too.)

Fram's replies

1. His bot edits (Helpful Pixie Bot mainly) generally fall under a), both authorized and correct. His automated edits on his own account are more mixed and come more often under d), not authorized and not correct, and c) authorized (implicitly, not explicitly) but not correct. Often it is the case that in a task of e.g. 1,000 edits of minor significance (like changing U.S. to US), he makes a small number of significant errors (e.g. breaking links), with the result that the number of correct edits is higher than the number of incorrect ones, but the result of these errors is much more problematic than the very minor improvement made by these edits.
2. His replies are usually civil when people arrive for the first time, but rapidly deteriorate when people return with more problems or when people question him (e.g. at BAG requests). Note e.g. [1], where he uncloses his own BAG request, and claims that "most of the comments seem to have no relation to the actual request, but are just moronic soap-boxing." Wikipedia:Bots/Requests for approval/Helpful Pixie Bot 46/ad hominem also contains examples of civility like "VeblenBot doesn't have a crazed admin chasing it around blocking it for ridiculous reasons.", "Yes it is possible to have pathological code to deal with pathological editors." Note that Wikipedia:Bots/Requests for approval/Helpful Pixie Bot 46 was eventually denied "Per editing restriction, problems with scope, and a lack of civility in response to feedback that is utterly unacceptable for a bot operator." Similarly, on his talk page, you can find things like User talk:Rich Farmbrough/Archive/2012Mar#Whitespace around section titles, where he eventually agreed to stop his unapproved (and guideline violating) cosmetic changes, but where the first reactions were less than satisfactory, e.g. when asked what approved task covered these edits, he just replies "Only on Wikipedia ...", and then in the next reply offers two BRFAs that don't include the objected whitespace removal in their description...
   • Here we see my prediction coming true, in a rather odd way. Though the close has been cited in this case against me the sensible perception of Anomie that one would have to be a "bit paranoid" to think this stuff would be trotted out in future, is, sadly proved doubly wrong. Not only has the close been cited four or five times, the attempt to prevent it being abused has been cited, twice I think. I forget the illustrious editor who said that if something is stated sufficiently often on Wikipedia, people will believe it, regardless of evidence, but I would hope it is not true of ArbCom. Rich Farmbrough, 16:44, 18 April 2012 (UTC).
     
     • Please indicate where I have used the original close against you? Please tell me where I abused it? I said that you reverted the close of your own BAG request, claiming that it should just be closed as "ultra vires", while the actual request had a lot of actual opposes besides the procedural ones. Yoru attempt to hide this by undoing a close against your own request instead of e.g. simply requesting that the close would be changed. And Anomie's comment just means that future requests to do mass creation shouldn't be prejudived by that BAG, not that what happened in it shouldn't be used in dispute resolution. Fram (talk) 07:09, 19 April 2012 (UTC)
       • You are using it here. And I didn't say it was used four or five times by you. I don't really care who used it, it doesn't matter. What matters is that
         1. It is normal practice on Wikipedia to rake muck, and use stuff out of context. It shouldn't be.
         2. You bring up matters irrelevant to the question "eventually agreed to stop his unapproved (and guideline violating) cosmetic changes" thus propagandising you point of view. That is typical of your approach, you throw in muck from one argument into another, without properly supporting it in the first. Eventually it is written in so many places that people think it must be true. Rich Farmbrough, 02:16, 20 April 2012 (UTC).
            
3. C. Their have been fewer incidents, but time and again he restarts with some task that technically violates his restriction, and often such a task causes problems. When confronted with this, he usually claims that he doesn't recognise the validity of the restrictions, which had according to Rich no consensus behind them and were placed by a now retired admin (the relevance of the last argument escapes me). User talk:Rich Farmbrough/Archive/2012Mar#Editing restrictions and Pixie Bot contains some of these arguments, and a healthy dose of incivility as well. At least if the restrictions would be affirmed at this ArbCom as being genuine community decided, valid restrictions, he would have less arguments to wikilawyer around them. Fram (talk) 07:14, 18 April 2012 (UTC)

The relevance of the last argument is that I was able to have a sane discussion with RD232, despite your unfortunate interjections to our conversations. And since I have already said that several times in the past, it is strange that it should now escape you. Rich Farmbrough, 07:59, 18 April 2012 (UTC).

A) diffs? B) Whether the admin who placed the restrictions is retired or not has no bearing on whether they are valid and should be heeded or not. And why, if you were able to have sane discussions with him or her, didn't you discuss the validity of the restrictions. Why did this sane admin place apparently unsupported restrictions in the first place, and why didn't he remove them after discussion with you? Fram (talk) 08:28, 18 April 2012 (UTC)

Because it was a modus vivendi. And as to why he placed unsupported restrictions and didn't remove them for the first point I suspect he didn't read the discussion closely enough (I comment there that it would take hours to understand the situation), which is supported by the fact that he simply cut and pasted someone else's words. I had thought at the time that he had put proper consideration into the matter and drafted a considered compromise, but I over estimated in that respect, and under-read. Rich Farmbrough, 08:48, 18 April 2012 (UTC).

So, no diffs or links. I'll reply again if you start using those, it's more useful to reply to actual facts than to your currently unsupported version of them. Fram (talk) 09:05, 18 April 2012 (UTC)

I'd much rather you didn't reply again. so we are done here. Rich Farmbrough, 02:16, 20 April 2012 (UTC).

Headbomb's replies

1. It depends on how you count things. If you look at what happens globally (several hundred thousands of edits), the vast majority of them are problem-free. However, due to the nature of the edits, problems happen in 'chunck'. For example, let's some approved task like ISBN hyphenation occurs, and the list of pages is built through scanning the database. The scanning the database for articles to fix means that the only time where RF/bots would have a chance of making a useless cosmetic edit would be if someone hyphenated the ISBN between the time of the dump and the bot's run. So you have this period of time where everything going fine. But then some different logic could be used (like fetching the pages from a category rather than a database dump), and then several pages not be in need of hyphenation fixes would nonetheless get edited or trivial reasons because the filtering is not good enough. And since so many edits get made, even something like a 1-2% fail-rate means several thousand purely cosmetic edits per month. And then there's also the issue of doing RF's personal cosmetic fixes (which even without the editing restrictions would not be OK to do, even as part of approved bot tasks), and other ER violations (mass page creation, e.g.).

   HPB will not make an ISBN hyphenation edit that does not hyphenate an ISBN. Rich Farmbrough, 08:43, 18 April 2012 (UTC).
   

2. In my experience D). Anyone raising concerns about the pertinence of the edits is dismissed as irrelevant or part of an anti-RF crusade. It's not that RF gets the technical details wrong, issues of civility/communication means the discussion just never reaches a point where technical details can be discussed.

   Erm, I'm sorry you feel that way. But it's simply not the case, if someone comes and asks a reasonably polite question - or even "Hey looks like you screwed up these edits!" they get a polite response. If they come and make themselves unpleasant, threaten blocks and sanctions, make reports to AN without discussing with me first, never admit when they are wrong and are otherwise boorish, they can expect somewhat more brusque treatment. I don't think that's unreasonable. Rich Farmbrough, 08:43, 18 April 2012 (UTC).

   This and this certainly are not examples of expected bot op behaviour. Headbomb {talk / contribs / physics / books} 16:58, 18 April 2012 (UTC)

   I agree, perhaps you could ask CBM to stop. Rich Farmbrough, 02:25, 20 April 2012 (UTC).
   

   More seriously, I apologise to those who had to read anything that was less than civil. Nonetheless there is no doubt that CBM posted to those BRFAs not to raise technical issues, or other issues to be resolved, but with the purpose of preventing the BRFAs from passing. Nor I think is there more than the smallest window of doubt that he posted because they were my BRFAs. Therefore this constitutes a direct attack on me as persona non grata. To my mind this is unacceptable behaviour, CBM can raise technical issues, policy issues, consensus issues (even if they are being raised to be obstructive, though he shouldn't) provided that they are valid. We - the rest of the community - will deal with those issues, either by modifying the proposal, getting extra input or withdrawing/denying the BRFA. A second opportunity arises during trials (and should be limited to the trial edits, not to other edits out-with the trial). And when the bot is running, any problems can be brought up on the bot operators page, or at BON. Finally even after a task is complete, any problems can be brought to the bot operator, or the noticeboard to be retroactively fixed. Numerous examples exist of bot operators fixing up something after a bot run, where needed, both for their own bots, and for others'. Attacking BRFAs solely on the basis of their proposer is not appropriate behaviour, and not something that I would have expected from CBM, despite his problematic behaviour in other spheres. Rich Farmbrough, 15:25, 20 April 2012 (UTC).
   

   The point of the BRFA is to vet both the technical aspects of the task and the bot operator who proposed doing it. Thus there is nothing inappropriate about raising concerns about the bot operator at a BRFA provided those concerns are germane to task itself or the suitability of the operator to conduct the task. Each thread I stated on the BFRA, visible now at [2], was focused on the task itself or the task description; you changed the topic each time to focus on yourself instead of the task. My goal was to simply comment on problems with the task (like lack of community consensus) or the task description (vague specifications), because a good task description that has community consensus is somewhat operator-proof. — Carl (CBM · talk) 16:22, 20 April 2012 (UTC)

   You may well claim that, however you raised no such objections at VeblenBot 4. Therefore they are targeted not at the task, but at the editor, as indeed you confirm later in the threads, and have said elsewhere. Rich Farmbrough, 19:00, 23 April 2012 (UTC).
   

Let me get this straight, your evidence that CBM is hounding you is CBM's lack of objection for his own bot's BRFA ? Headbomb {talk / contribs / physics / books} 19:38, 23 April 2012 (UTC)

No you have it backwards. It is one of the many things that CBM thinks are fine and dandy for him to do, but are evil for us lesser mortals. Similarly he now suggests he take on the {{Infobox language}} task that he created so many objections to. Similarly he reverted a bunch of stub-updating edits and then ran the full task on his own bot. And so on... Rich Farmbrough, 17:50, 24 April 2012 (UTC).

I think you are talking about the following: someone renamed some stub templates out-of-process, and I reverted that because it was premature and a violation of policy. Later, after the process was followed, and the outcome was to rename the template, my bot handled the implementation of the closure of those discussions. There's nothing improper about my edits nor the bot's edits, and they do not conflict because there was a deletion discussion closed in between. — Carl (CBM · talk) 19:49, 24 April 2012 (UTC)

there was no out of process these were listed at WP:AWB/TR. This type of argument is too process bound for Wikipedia. Basically you were edit warring, and you choose to justify it with process. The fact that on this occasion you "saw the light" and implemented the rest of the changes can be considered a good thing, or it can be considered otherwise. The failure, in my opinion, is to be battling against other editors over this stuff, whether in this case approved AWB edits, or some other minor fix that you take exception to, regardless of whether they have explicit policy or consensus backing them. That's why I urge you to cease from reverting or complaining about "cosmetic" changes by other editors. Rich Farmbrough, 03:48, 25 April 2012 (UTC).

There were about 3,000 of these "rail stub" articles - clearly enough that a bot needed to make the change. The AWB list has templates that may be changed, but if there was a consensus that they must be changed they would already be changed. In particular WP:BRD still applies to user edits made with AWB. In general, AWB requires substantial operator discretion, because it can be easily misused to perform large-scale tasks that do not have consensus. Fortunately, most AWB operators stay inside the lines and never have any complaints. — Carl (CBM · talk) 10:33, 25 April 2012 (UTC)

You keep making this assertion that a bot was needed, but there was no urgency. Edits could have been saved. Your comment is full of other quatsch, this panglossian non sequitur "if there was a consensus that they must be changed they would already be changed" - if consensus was instantly, magically, implemented Wikipedia would be much better (but we would probably al be dead). Of course BRD applies to AWB edits. Common sense and good faith apply to manual edits too. Also I see you have picked up the art of clumsy innuendo, best to unlearn that as soon as possible. Rich Farmbrough, 00:40, 26 April 2012 (UTC).

1. Well we're here, aren't we? The ERs, had they been followed, would have solved all problems. Edits must be tied to specific BRFAs + no non-standard cosmetic fixes. I don't really know if this is a lack of proper enforcement, but it's not exactly as if attempts to cull this behaviour was not made before. The main problem seems to stem from RF's refusal to recognize the current ERs as valid / applicable / desired by the community. Whether it means that making the current ERs official ARBCOM restrictions would solve the issue, I don't know. But it's IMO something to at least consider as an alternative to blocks or some kind of blanket restrictions on any bot-activity. Headbomb {talk / contribs / physics / books} 07:41, 18 April 2012 (UTC)

   No we are here partly because I asked Arbcom to accept the case. Even Fram gets a civil reception if there is a real problem. User_talk:Rich_Farmbrough/Archive/2012Mar#Helpful_Pixie_bot_mistakes_and_minor_problems Rich Farmbrough, 08:05, 18 April 2012 (UTC).
   

   We're here because you don't follow your editing restrictions and previous attempts to stop the problem failed to do anything about it. I'm pretty sure ARBCOM would have accepted the case regardless of whether you asked them to or not. Headbomb {talk / contribs / physics / books} 16:58, 18 April 2012 (UTC)

   +1. ArbCom has declined cases in the past where all named parties were willing to participate because there wasn't an issue meriting their attention. They've also accepted cases where a party was unwilling because there was such an issue. The individual inclinations of a party to participate have little weight when it comes to the decision to accept a case. Hersfold non-admin^(t/a/c) 21:00, 18 April 2012 (UTC)

   Read! Several Arbs specifically mentioned my request in their votes to accept. Rich Farmbrough, 02:24, 20 April 2012 (UTC).
   

   We are here for a number of reasons. I don't think prejudging the case to establish the reason we are here is very helpful. Maybe you think differently. Rich Farmbrough, 02:24, 20 April 2012 (UTC).
   

   See also User_talk:Rich_Farmbrough/Archive/2012Mar#BOT_problem

   See also User_talk:Rich_Farmbrough/Archive/2012Mar#Helpful_Pixie_Bot_.26_ISBN

   I could go on... Rich Farmbrough, 08:14, 18 April 2012 (UTC).
   

Questions by Newyorkbrad

My thanks to the parties for their anticipated responses to these questions. Some of them may overlap issues addressed in the evidence or elsewhere, so cross-references can be provided where appropriate. Thank you. Newyorkbrad (talk) 17:59, 21 April 2012 (UTC)

1. Is the error rate of Rich Farmbrough's automated edits substantially greater than, about the same as, or substantially less than what one would typically find or what one might generally regard as acceptable? Has the error rate improved, remained about the same, or worsened over time?

2. Are there particular types of automated edits that Rich Farmbrough makes that are particularly more or less prone to errors or to raising the types of concerns that have given rise to this case? For example, are there particular bots or bot tasks that have operated error-free for a long time, or others that are particularly buggy?

3. If Rich Farmbrough were hypothetically restricted from making all or some portion of the automated edits he has made over the years, would likely lead to (a) the edits being made by others with reduced problems, (b) the edits being made by others without reducing the probems, (c) the edits not being made at all, to the detriment of the encyclopedia, or (d) the edits not being made at all, without detriment to the encyclopedia?

4. Are there any significant concerns regarding (a) Rich Farmbrough's non-automated mainspace edits, and/or (b) Rich Farmbrough's administrator actions unrelated to his automated edits?

5. Rich Farmbrough has asserted, in substance, that the other parties to this case have pursued complaints against him selectively or vindictively. Other than the history of the complaints themselves, is there any evidence that this is the case?

Responses by Hersfold

1. Actual errors (that is, edits that would be inappropriate regardless of editing restrictions) aren't too terribly high, not all that different from your average bot. Inappropriate edits (changes that shouldn't be made in light of Rich's restrictions) are made much more frequently; it also seems that while he will eventually stop those changes when confronted, a few weeks/months later it's happening again.
2. It seems the general maintenance sort of tasks (tweaking references, templates, etc.) are the sort that have errors most often (so, in general, just about anything done by HelpfulPixieBot). AFAIK, FemtoBot has been chugging along without incident for some time now.
3. Assuming he abides by the restriction, I'm sure that these tasks would quickly be taken up by others with an lower rate of problems. They are mostly necessary and useful tasks, and a handful of editors with AWB can prevent backlogs from getting too long while a proper bot is designed and approved.
4. I honestly haven't looked much into either, focusing more on the other concerns. I'm not really aware of any concerns in those areas.
5. No. At no time has anyone raising these concerns acted in a manner that would indicate their motives are less than appropriate. These assertions, in my opinion, are attempts to deflect attention away from Rich's violations.

   You have, presumably, reviewed the last two years interactions between the parties? Did it not occur to you that it was in order to shed some light on these matters that I encouraged the committee to accept the case? Rich Farmbrough, 05:39, 22 April 2012 (UTC).
   

   You might like to explain how this is appropriate

"As far as I recall, I made reports to AN without prior discussion with you twice, the one that lead to your one-month block and this ArbCom case, and this one in January, where you eventually blocked yourself for 8 hours. "

Here Fram implies that I self blocked as some kind of mea culpa - and, incidentally shows that he has a detailed knowledge of my actions that makes his other claims to ignorance hard to believe - whereas my block summary was something like "Block self, drahma will still be there in the morning". If anyone can explain to me how this can be seen as other than another step in insidious blackening of my name, let them do so. They will either have to be incredibly naive or incredibly ingenious. Rich Farmbrough, 07:10, 22 April 2012 (UTC).

1. My "claims to ignorance"? You mean when you make statements without any link to show what they are actually about? People can check my replies to you to find the number of errors in your statements I have been able to find already; the other ones are beyond correcting because I haven't gotten a clue what you are basing them on. As for "knowledge of your actions"; it isn't hard to find the result of a certain discussion, and knowing your block log isn't so incredible during an ArbCom about you. What you read in my comments and the motives you ascribe to them are entirely up to you of course. I just present the facts, I don't try to pretend to know why you do or don't do things. Fram (talk) 09:44, 22 April 2012 (UTC)

" I don't try to pretend to know why you do or don't do things." So how come you said "Only caring for his reputation" ? Do you think we have the memory span alleged for goldfish? Rich Farmbrough, 01:10, 25 April 2012 (UTC).

Responses by Elen of the Roads

1. I don't think the error rate is spectacular. As with similar cases, it is the sheer volume of edits that generates the number of errors. Although having said that, creating SPI categories for every sockmeister since Noah was an error to start with.
2. A lot of the complaints were about making cosmetic changes with AWB, rather than errors. However the community has low tolerance for breaking things while making what are seen as 'useless' edits.
3. I'm sure others could do the necessary edits. I'm not really sure why Rich can't - it's a mindset problem, not a coding skill problem. Someone with less skill but a view of themselves as producing a product for the community is preferable to someone with greater skill who regards themselves as the arbiter of what the community wants.
4. Not that I'm aware of.
5. There were a lot of complaints about technically violating his editing restrictions by making cosmetic changes through souped up AWB which seemed niggly to me, which is why I let things go after initially threatening to block Rich. I thought Kumioko talked a great deal of sense to him about the need to ensure community consensus for large scale automated editing. But then he ups and does completely daft things.

Elen these answers are well meant, and very kind, thank you, but there are some real issues here.

• "I don't think the error rate is spectacular" Actually you have no idea what the error rate is, except indirectly that no one is kicking up a stink - please see my answer to this part.

• "However the community has low tolerance for breaking things while making what are seen as 'useless' edits." Do you claim that there are more than a vanishingly small number of "useless" edits? Do you further claim that they a "breaking things"?

• You say (or imply) that I "regard myself as an arbiter of what the community wants" - what do you mean by that?

• "I thought Kumioko talked a great deal of sense to him about the need to ensure community consensus for large scale automated editing."

I have no idea where this comes from - perhaps you can point me to it?

• "But then he ups and does completely daft things."

Well we all do completely daft things form time to time. I presume, though, you mean the suspected sock-puppet categories - which is odd, since we have disposed of that myth on day 1. If you think there is a problem with these please talk to me about it on my talk page, there may well be, but unless someone discusses it I will be assuming its down to this crazy "crawl-able webpages" idea that was debunked as soon as it was raised. If you mean something else (apart perhaps from putting too much time and effort into this case) then the same applies. Unless I know what it is I can't do anything about it. Rich Farmbrough, 06:31, 22 April 2012 (UTC).
06:31, 22 April 2012 (UTC)

• • Rich, I'm not sure you're helping your case by arguing with everyone who responds to NYB. However...The statement 'the community has a low tolerance for breaking things while making useless edits' is a general statement, which I also made of Beta. I don't know what your error rate is. It's lower than Betacommand's was, because it gets fewer complaints. People do notice your edits and freqently remark (to me - the small section of Wikipedia that says such things to me) that they appear useless - you could compare that for example with the number of complaints User:Fastily used to get for deletions in error - because he got a complaint every single time he got it wrong, it looked like he had huge numbers, but the number of correct deletions was massively higher.
    • On the other hand, when , in the past I have let inaccuracies go unchallenged, they then get repeated as true. And CBM claims that simply not challenging something gives it force. Rich Farmbrough, 05:01, 26 April 2012 (UTC).
      

• • However, generating all those sockmeister categories was just plumb bloody stupid. Point me at were anyone ever had requested that this be done for all sockers. Point me to where you did the tests that would have highlighted the problems. "I'm sorry I blew up the Chemistry lab, but it can't be that bad because the school haven't rebuilt it yet" really won't wash as an answer.
    • You can say "stupid stupid stupid" as much as you like, but the reasons giving for deprecating the creations have been shown to be false. Pilot runs back on the 23rd of March created no concerns. Sure it was stupid in the sense that I should, perhaps, have predicted that there would be a Fram attack, but on the other hand the previous AN he had just been told the edits I was making were perfectly good. (And of course I had stopped the runs as he requested on my talkpage, although that never came up for discussion, and he assumed that I hadn't.) And maybe I should have expected that when you said you would block for "a day, a week, a month and then a year" you actually intended to start with a month, and break [3] this project. — Preceding unsigned comment added by Rich Farmbrough (talk • contribs)
      • "Pilot runs back on the 23rd of March created no concerns." Was anyone aware that you were making these "pilot runs"? Did anyone comment on them? Or is it just a case of "no one noticed at the time, so I made a lot more soon afterwards? "the previous AN he had just been told the edits I was making were perfectly good." was about a totally unrelated matter (ISBN editing), so how this can be used as an argument that these creations were supposedly acceptable is not really clear. Anyway, as far as I can see, I don't believe you have explained anywhere how you thought these categories would in any way be an improvement to Wikipedia. What is the actual benefit of from now on having a category that links e.g two short-lived accounts from 2006, or two IP adresses that may have been used by the same person for a short while in 2010? Fram (talk) 06:52, 26 April 2012 (UTC)

• • The significant chain of comments featuring Kumioko (who has also posted on the talkpage here, rather than posting in evidence, but I think it should definitely be read) is here.Elen of the Roads (talk) 23:40, 25 April 2012 (UTC)
    • Yes I went and re-read the entire thread before asking for a pointer. The only thing I can see is

I don't think its a lack of consensus so much as not taking the time to document it. In most cases (I admit not all) the edits are helpful and generally reflect the attitude of the mob, but instead of starting a long boring discussion about something that 98% of editors either agree with or don't care about, he just did it. This was about a week after I had stopped contributing to the thread, and wasn't addressed to me. So having just spent another night of wasted effort (in the sense that people could, if they wished, use their heads and work cooperatively, rather than combatively) I will ask you again to actually point to what you meant rather than several screens of text. If you misremembered or if I am just to tired at 6:01 AM to see it, that's fine. But if it's there please let me know. Rich Farmbrough, 05:01, 26 April 2012 (UTC).

Put it this way Rich. Kumioko and I both thought we were having a discussion with you. The fact that you had decided to slip out the back will be particularly disappointing to Kumioko, who is just about the one editor in the whole of Wikipedia who has gone in to bat for you. Which must say something about something. Elen of the Roads (talk) 23:15, 27 April 2012 (UTC)

You are imputing Kumioko's thoughts to him. And certainly I read most of that thread, but we have learned - not only Kumioko and myself, but those of us subject to these unwanted attentions generally - have learned that sometimes less is more. And don't think I am not grateful for the support, here, on my talk page, and by email, from everyone who has offered it, and for more-or-less neutral comments from our mystery IP. What is disappointing is the overwhelming, blinkered negativity. We really do need some decent facilitators and mediators. Rich Farmbrough, 03:20, 28 April 2012 (UTC).

In fact it's fairly clear that Kumioko is primarily addressing you (Elen) I have asked Rich to not [make] anymore changes [from] Portal box to Portal until the discussions conclude. Obviously you might be relying on a fleeting memory of what you read at the time, but I find it troubling that you are making such fundamental errors in understanding. The technical errors I can forgive (though I question very strongly whether a self-proclaimed technical ignoramus should be blocking editors on technical grounds), but to have you fundamentally misread this sort of thing is very worrying. Rich Farmbrough, 11:21, 28 April 2012 (UTC).

Just to clarify, my comments were in general and not directed at any one editor in particular, even though I named a couple specifically. I, as much as anyone, am guilty of trying to do too much and have suffered the wrath of the community for it. As with my case I think that there are a lot of heated tempers here that are not making things any better. I think Rich made some mistakes, some of which were clearly obvious and should have been caught like using Category:Living people when building articles for people who had been dead for 100+ years. I also think that some editors like CBM have become overzealous in their quest to protect the Wiki from Rich's actions. I think that mistakes aside Rich does far more good than harm to Wikipedia and losing him over this stupidity would be a great loss and that's why I have voiced my support of Rich. Everyone makes mistakes and Rich's volume of edits is part of what magnifies the errors that he makes and draw so much attention to them. I also think that the nature of these Arbcom cases also tends to crush the editor being debated. Its human nature that we tend to speak up more in complaint than in praise and Wikipedia is no exception. Those who would support Rich are likely busy doing other things or don't even know of this case leaving those that want him gone free rein to say what they want without direct opposition. Kumioko (talk) 01:20, 29 April 2012 (UTC)

Response by Rich Farmbough

1. Is the error rate of Rich Farmbrough's automated edits substantially greater than, about the same as, or substantially less than what one would typically find or what one might generally regard as acceptable? Has the error rate improved, remained about the same, or worsened over time?

No one knows the answer to these questions, because no-one has analysed the data. While I did analyse a large number of SmackBot edits in 2010 this is now very out of date, as not only has the code changed but the operating conditions.

I can however give my impressions:

• For Helpful Pixie Bot's main task of dating maintenance tags the number of broken articles is probably about 1 in 50,000 or less. The number of articles with an incorrect correction of a date is probably about 1 in 5000 to 1 in 10000 or less, and will improve. The number of articles not dated at all is probably around 1 in 1000.

• For Helpful Pixie Bot's ISBN task there are a number of known but minor bugs. The most serious may have resulted in a very small number of valid ISBN's being tagged as invalid. The other bug simply resulted in some ISBNs not being picked up at all.

• For ad-hoc tasks, it is impossible to give a figure. The usual approach is to run a significant number of test edits manually to iron out issues. Generally the main run is then done in conjunction with the person or group requesting the task, and adjustments and re-work are done as required.

The error rate overall is probably reducing as the established codebase grows, and as experience is gained.

2. Are there particular types of automated edits that Rich Farmbrough makes that are particularly more or less prone to errors or to raising the types of concerns that have given rise to this case? For example, are there particular bots or bot tasks that have operated error-free for a long time, or others that are particularly buggy?

All the long-running tasks are virtually error free. Newer tasks are more likely to have problems, these can, however usually be resolved quickly.

The most error prone tasks are manual AWB tasks, the majority of errors are mistaken saves that are immediately reverted. Sometimes an edit with no errors but which does not change the rendered page is saved. In this case, I often consider reverting, but invariably decide that that would be just foolishness.

3. If Rich Farmbrough were hypothetically restricted from making all or some portion of the automated edits he has made over the years, would likely lead to (a) the edits being made by others with reduced problems, (b) the edits being made by others without reducing the problems, (c) the edits not being made at all, to the detriment of the encyclopedia, or (d) the edits not being made at all, without detriment to the encyclopedia?

There would be a mixture, some of the edits would be taken over, some would fall into abeyance to the detriment of the encyclopedia. There would be new problems and some of them would look like problems that have been fixed. This has already happened. So the answer is B and C.

4. Are there any significant concerns regarding (a) Rich Farmbrough's non-automated mainspace edits, and/or (b) Rich Farmbrough's administrator actions unrelated to his automated edits?

I'm sure someone has some... lets see.

5. Rich Farmbrough has asserted, in substance, that the other parties to this case have pursued complaints against him selectively or vindictively. Other than the history of the complaints themselves, is there any evidence that this is the case?

Lets take this question apart. Because while the tenor of the statement is not inaccurate, it is not precisely what I have asserted, and probably what I have asserted is not precisely what I mean.

Certainly I have never asserted any such thing about either Elen or Hersfold. And I do make assertions of a somewhat similar nature about Fram and CBM. And I will lay out now exactly what the case is as I see it - something which I had hoped to avoid when the case started, but which now seems to have all been brought into the open. The only relevant information I can think of that I am keeping back is personal communications from other editors.

I assert that CBM has opposed certain editing practices, some of which are subject to explicit consensus supporting them, and some of which are not. I assert that he blocked SmackBot numerous times over one of these, and insisted that I rewrite the code in AWB to remove this change. I assert that he also has a problem (as drawn out in the workshop) with the community consensus over the "Otheruses" templates (among others) and reverts, blocks, threatens and otherwise uses his Wiki-fu to attempt to prevent this. I assert that this behaviour has spilled over into "enforcing" wherever he can what he perceives as me "breaking rules". I assert that this has moved from, as he used to claim, simply reading his watchlist, to looking at my contributions page. I assert that it has gone as far as leaving long threads on Elen's talk page and attempting to sabotage BRFAs.

Now as to this being selective, I think that in the latter stages it might have become so, but not uniquely. CBM is also reverting other users, and interfering with other bots in a similar way. Is it vindictive? I don't think so. CBM is annoying, but he is straightforward, except for the slight confusion about the meaning of ad hominem he is usually clear about what he wants to do, and why he wants to do it. However when consensus is against him he has continued to insist that he is right, that "otheruses" is "semantically correct" for example, nor does he show much flexibility in the interpretation and application of rules. These traits may be admirable in certain disciplines, but they are not appropriate for a complex human environment such as Wikipedia. Nonetheless I believe in time CBM and I will resolve our differences, and there is absolutely nothing to stop me working collegially with him now except in fora where he arrives with negative agenda.

As to Fram, I am constantly trying to interpret his actions in a positive light, and I am sure that when all is said and done, he really does believe on some level in what he is saying. He is certainly selectively targeting me, although I am not the only person he has targeted. I do not think he is vindictive in the sense that I think he really believes that there is an underlying problem that justifies his actions - i.e. his attempts to get me indef blocked, for example were "for the greater good". What I am unable to cope with is two aspects, not so much of what he does but as the way he does it. These are firstly adopting a superior and unfriendly attitude, telling others what to do, not listening to careful responses to explain technical (and other) matters and continuing to maintain that which has been shown to be false. Secondly - and I keep trying to excuse this on the basis of English not being his mother tongue, but failing - all the relatively low level abuse, such as when I fixed a template to remove red-linked categories from a number of pages he says I am "hiding a problem" implying that I am concealing some wrongdoing on my part, whereas I am fixing something created by a user who left the project long ago. This behaviour I find unacceptable, and it has been going on for so long that the chance that I will ever consider Fram a colleague is vanishingly small.

Let it just be said that in the pages of AN/I referred to by other parties (which I have been re-reading - and they are pretty grim) there are also people who behave worse than Fram, as well as some, with various views on the topics in question, who calmly stand for what they think is the right thing. I take my hat off to them.

The statements about CBM's editing habits and interlocutions are supported by evidence given in the workshop and by CBM himself, and by the BRFAs that have been cited.

The statements about Fram are supported in the linked to AN/I pages, in the AN report which lead to this case and in even in examples of his statements in this case.

Addendum

Here is an previous analysis of about 5000 of SmackBot's edits.

Responses by Fram

1. His error rate in his automated editing? With his bots, the rate has decreased clearly after October 2010, when SmackBot got scaled down to only a handful of tasks (mainly tag dating), and the application of other (mainly cosmetic) changes at the same time were largely abandoned. It increases again with new tasks, but the bots are IMO not the main remaining problem. His error rate in his automated (or semi-automated, script-assisted,...) editing on his main account is a more serious problem. With ambitious tasks, like the creation of articles based on the DNB, the error rate is very high. With smaller tasks, it varies, with a largely unproblematic replacement of Encyclopedia by Encyclopaedia on the one hand, and e.g. a much more problematic replacement of U.S. with US on the other hand, with the latter changing e.g. bluelinks into redlinks and so on. I don't feel that the error rate has increased greatly in his editing as RF (not as a bot), and while the massive amount of edits makes it hard to put figures on the overall acceptability of the error rate, at least in a fair number of tasks it was unacceptably high, certainly when compared to the benefit the task brought (like with the replacement of U.S. c.s. to US). 2. Femtobot has made some errors, but these were soon corrected. E.g. User talk:Rich Farmbrough/Archive/2012Mar#Template blanking or User talk:Rich Farmbrough/Archive/2011Nov#Femto Bot error. 3. A, the edits would be taken over by others, with less errors. 4. Yes, he e.g. regularly edits protected templates without prior discussion, sometimes with very poor results. He has also deleted an article that was kept (a "merge") at an AfD and refused to undo his deletion. My evidence in this RfA, and my case statement in the rejected previous ArbCom case, contain the necessary links. He hasn't on the other hand made a lot of normal admin-tool edits, so it's not as if removing his admin bit would increase the workload for other admins enormously. 5. No. Selectively, i.e. posting when there are problems, but not when things are going allright? This may be largely true. Selectively in the sense of ignoring problems by other editors but not by him? No, when I become aware that some editor makes a lot or problematic edits, I keep an eye on him. Editors who don't have that problem or where I am not aware of that problem don't get that treatment of course. This only seems logical to me. Vindictively? Revenge for what? I don't believe I have any reason to take "revenge" on Rich Farmbrough.

Responses by Headbomb

1. Regarding HPB, tasks are done very cleanly, others not so much. If we're talking about RF himself, his error-rate is much higher than what would be allowed for a bot.
2. Pretty much any manual task will be error-prone because the edits are not properly reviewed, or made against consensus. For bots, there are so many tasks it's often hard to tell what exactly the edit was meant to do, or some minor extension is implement without BAG oversight, leading to less-than-ideal edits. For example tagging invalid ISBN are done via {{Please check ISBN}} appended as |ISBN=XXXX {{Please check ISBN}} in citation templates. This is not the best implementation possible: The template, if placed there cannot display a message because of how {{citation}} works. It would have been much better to append it to the citation (aka |ISBN=XXXX {{Please check ISBN}}, so the template could display a message similar to .
3. a/d) mostly. c) for some bot tasks. Headbomb {talk / contribs / physics / books} 17:27, 23 April 2012 (UTC)
4. No other problem with his edits that I know of. But then again, I never seen RF made any non-automated/semiautomated edits.
5. Bot operators sometimes get hassled with trivial complaints. However, as far as I can tell, the complaints RF receive are legit, especially with regards to his lack of compliance with WP:COSMETICBOT (for example in Wikipedia:Bots/Requests for approval/Helpful Pixie Bot 48 (collapsed section). Maybe CBM complains more than others, but in this case, policy is on the side of CBM. Headbomb {talk / contribs / physics / books} 17:27, 23 April 2012 (UTC)

Excheese me? In several cases in that discussion you point out that CBM is basically wrong. And anyone reading that discussion will see that there is clear obstructionism going on, whether intended or not. The person requesting the run says

I figured this would be so trivial that it wouldn't need discussion, ...'

And yet a fairly straightforward BRFA becomes a 45k discussion. Rich Farmbrough, 19:17, 23 April 2012 (UTC).

Final questions by Newyorkbrad

Questions to Rich Farmbrough

Q. How do you select what tasks you will perform on an automated and semi-automated basis. In advance of mass-performing a task, do you make an effort to consult with other editors or the community regarding whether a given task is likely to be useful, and/or how it can best be implemented so as to effectuate the improvement desired without causing any undesirable side-effects?

A. It depends. I have often picked up tasks from Wikipedia:Bot requests. Things like reg-ex typo fixing are established tasks, done by many editors, historically, though it seems a lot of them have given up. I have submitted BRFAs (over 50 of them) and sometimes that system has worked well. Sometimes it slows things down to the extent that the task has been finished manually, either by me, or by someone else, before we get anywhere. Some things don't need additional consensus building, like specific typo corrections, or MoS fixes.

Q. If not restricted as a result of this case, are there any changes that you intend to make in your use of automation or semiautomation in your editing going forward, including with regard to how you choose what tasks to perform and how to perform them?

Yes six things spring to mind:

1. Working on the core Pixie functionality to improve it per WP:COSMETICBOT.
2. Providing a fine grained on-wiki control of bot tasks, enabling individual tasks to be turned on and off.
3. Submit a lot more tasks to BRFA, and improve the control system to handle the complexity of keeping track of the tasks.
4. Create a log of minor manual work and planned work, open to contribution.
5. Tie manual runs into the log by using the edit summary.
6. Create a bug logging system, so that users will not loose faith, and will see the both the progress and effort on bugs. (Largely  Done)

Rich Farmbrough, 23:42, 25 April 2012 (UTC).

Questions to other parties

Q. Can you suggest any methods by which Rich Farmbrough could improve his task selection and implementation, thereby capturing the benefits of his dedication to the project and his willingness to make large numbers of repeated edits, while minimizing the types of issues that have given rise to this case?

A. The problem with some of the tasks is that there is no discussion. I do think the community as a whole needs a little discussion on automated (non-bot) editing, as it really cannot be done on a BRD basis. Discussion is essential to uncover potential flaws (eg changing U.S > US breaking links, or creating sockmeister categories linking named accounts to IP addresses). The second is testing. No coder ever can fully UAT, and Rich seems to have an unfortunate tendency to ignore his own mistakes sometimes. Again, that might be something for a wider discussion.--Elen of the Roads (talk) 23:55, 25 April 2012 (UTC)

A. The key point is that task selection for bots and other large-scale jobs should be a community endeavor, using BRFA or other public discussion boards, not something decided by individual editors like me or Rich who happen to be able to write bot code. As I suggested in my proposed remedy, I think that moving all large-scale tasks (including AWB and page creation) to the bot account, so that BRFAs are required, would greatly reduce the problems. At the same time, BAG needs to insist on meaningful and complete task descriptions so that any editor can objectively tell whether particular bot edits are authorized. In particular, any "extra" cosmetic edits that the bot will carry out need to be described in the bot description and approved by BAG. Then any complaints would be for BAG rather than for Rich. — Carl (CBM · talk) 00:02, 26 April 2012 (UTC)

Here [4] is a summary of the situation with the notability template in 2011 that I wrote after taking the time to figure out what exactly had happened. There are a couple other explanations by Fram and Errant near my comment in the full thread [5]. That incident is typical of the problems created by lack of prior discussion for changes that involve large numbers of articles. — Carl (CBM · talk) 15:41, 27 April 2012 (UTC)

A. I can't add much more to this than has already been said. BRD does not work for large-scale tasks, as while the B part is easy, R gets difficult when things are spread over a large number of pages, and that makes the D part hell. All large-scale tasks need some sort of discussion before they take off, so that both the person doing the task and the community are clear on what exactly is to be done. What's also critical is ensuring that once a discussion has been held, you keep going with that plan. Any changes other than minor bug fixes should be discussed as well; particularly when replacing something previously removed as part of a discussion. Hersfold ^(t/a/c) 03:55, 26 April 2012 (UTC)

A. (Not a party) My longish comment about BRD and bots is here--short version: it says approx the same things as Hersfold. Besides discussing large-scale tasks in advance, I wish Rich could have a change of heart about editing in general. He seems at times to be out for glory, in terms of edit count, article creations (even with no useful content), etc; so he does a lot of things that don't need doing. I wish he'd let go of that. And he is obsessed with eradicating minor stylistic inconsistencies across the project. I'd ask that he contemplate WP:ENGVAR for a while: most of the community not only tolerates stylistic inconsistency, but in some cases, for good reason, actually mandates it. So he should stop dealing with cosmetic and stylistic trivia and instead do stuff that matters. Readers never even notice the US/UK spelling inconsistencies or other such issues as far as I've encountered. They care instead what the articles actually say. 67.119.3.53 (talk) 08:07, 26 April 2012 (UTC) Added: another thing: all these scripts should be thoroughly tested before wide-scale deployment. This is generally required in BRFA's (bot is approved for a small test run, then maybe a larger test run, before a full run can start). Rich doesn't seem to bother with that, for his unilateral operations (example from January 2011). This applies to template programming too, not just bots. Here Rich made some buggy ill-advised changes to the Notability template, which is in 1000's of articles, causing an ANI. Complex changes to high-visibility templates should be tested in a sandbox or test server in addition to being discussed beforehand. 67.117.130.107 (talk) 21:27, 26 April 2012 (UTC)

Sorry, but the notability thing is just wrong. The people commenting on it by and large didn't understand it (and why should they?), I get criticised in that discussion for "filling up" a technical tracking category I had created, as if that was a problem. Rich Farmbrough, 15:28, 27 April 2012 (UTC).

OK, I changed "buggy" to "ill-advised" as my memory of that incident is apparently faulty. My recollection (maybe conflated with something else) was that I found my way to that thread because the TDMCA template was breaking something in a page I was working on, that wasn't transcluding it, and it took me quite a while to figure out that it was getting into the page through the notability template, and it was still quite unclear after that what the new template actually was for. The comments from 71.141.88.54 in that thread were by me. Independently of this incident (and of Rich), I fear that we're heading into an era of insane amounts of disruption from buggy templates[6] and we're going to have to develop and enforce processes about testing them on a staging server before putting them on the live server. 64.160.39.25 (talk) 21:17, 28 April 2012 (UTC)

A. I could care less. Editors choose tasks as they like, and no-one cares. People use bot-functions of Twinkle, or operate AWB as a bot on a near continuous way. Yes, it would be better if there is discussion, but for many tasks it is not needed. The problem is, that if someone sees a bot-task being performed, the editor behind the bot-task is being targetted. E.g., the mass creation of sockpuppet categories was targetted as a set of bad edits. It turns out that there was something else broken on Wikipedia, but, intrinsically there is nothing wrong with the task that R.F. was performing. However, it is turned into a massive WikiDrama, because it is R.F. I do not think that R.F. should adapt his thinking very much, editors should consider whether what is going on is really a problem with R.F., or a problem that is either broken somewhere else, or whether the error is something that needs reporting and wikidrama or can just be fixed and ignored, and/or actually, if it is a problem that needs to be solved. The problem is, if s.o. is doing 100 good edits, and 1 bad edit, then that 1 bad edit is turned into a wikidrama (YOU made the mistake, solve it NOW - without any of the above considerations) in stead of it being solved and we move on. --Dirk Beetstra ^{T C} 05:20, 28 April 2012 (UTC)

Q. Put very directly, do you believe Rich Farmbrough's dedication and skills can be rechanneled through remedies in this case so that they can be saved for the overall benefit of the project, or have matters reached the point where complete or substantial restrictions on automated, semiautomated, or repetitive editing by him are required to eliminate disruption? Please explain.

A. I hope this can be resolved. We're not talking about someone who is massively disruptive, incompetent, acting in bad faith etc. Just someone who hasn't adjusted to Wikipedia becoming a more controlled environment, with less tolerance for even accidental mistakes. I think he's found being blocked for a month very trying, and maybe if he thinks the consequences will be serious in the future, he will engage in discussions as to how to avoid problems in the future. --Elen of the Roads (talk) 23:55, 25 April 2012 (UTC)

A. As above, I think the problems would be remedied by allowing repetitive edits only through his bots, which he would still be able to give as much effort as he likes. He has proposed making the bots more transparent, which would complement this. Most of the problems seem to come from repetitive tasks carried out on his main account, or unapproved "extra" edits by his bots. — Carl (CBM · talk) 00:02, 26 April 2012 (UTC)

A. My sole concern here is whether he will respect the remedies. One doesn't get to a month-long block by abiding by one's editing restrictions. On the other hand, his proposed changes above do sound promising. I also echo CBM's comments that restricting his automated editing to his bots will help significantly. Unfortunately, the only way to remove access to AWB for an admin is to remove the tools. Hersfold ^(t/a/c) 03:55, 26 April 2012 (UTC)

A. I agree with CBM, I don't have a problem with Rich running approved bots. Rich's idea that a 20,000-edit AWB task doesn't need approval is wrong, IMHO. I don't think withdrawing AWB will make any difference, for reasons discussed elsewhere. He has enough technical ability to not be slowed down by the withdrawal. That he even has the desire to undertake these big edits is itself an area of concern. Betacommand was in a purely behavioral dispute: he wanted to run bots for any reason he could, didn't care much about what they did, and used areas like NFCC enforcement as vehicles to run his bots on "important" tasks. Rich seems to really be pursuing an editorial goal, that isn't shared by his critics. So this is in a sense a content dispute with Rich using bots to gain advantage over opponents. Rich offered further down to enter mediation; maybe that could help uncover what the issues really are.

I see high-volume "semi-automated" editing as a recurring method of gaming BOTPOL (we've also seen it extensively from Betacommand and others) and I think arbcom and the DR community should be quicker to reject it when it happens. I don't see any workable ways to restrict it other than suggestions like Carl's (50 page editing limit per day, e.g.) that don't distinguish between automatic and other edits. Maybe it's reasonable to start with that restriction, with the possibility of relaxing it if some agreements can be reached in mediation. (I'd appreciate if the other parties could say what they think of the mediation idea).

I asked Rich further up (07:59, 19 April 2012) about his intentions after the case. You can find his response there. I note the possibility that Rich could continue contributing his programming skills to the project through toolserver or server-side programming, even if he has to completely stop running bots. I don't get the impression he is interested in that, however. 67.119.3.53 (talk) 08:07, 26 April 2012 (UTC)

Toolserver is a great way to deliver services to WMF projects. The only problem is that the signup process is rather clunky, I jumped through all the hoops about a year ago, ran into a problem and then needed to revisit. You need to make about 4 accounts, a PGP key pair and other miscellaneous actions. Maybe I should write a bot to do that. Rich Farmbrough, 19:42, 29 April 2012 (UTC).

A. I am not sure if it is something that is a problem with R.F., or a general problem with the community, or even, if this needs to be channeled. Tell people to move on. --Dirk Beetstra ^{T C} 05:22, 28 April 2012 (UTC)

Questions by Rich Farmbrough

1. Do you think that it reasonable to expect error free editing?
   1. If you answered "no" to 1, do you think that correcting errors in a reasonable time frame is a better expectation?
      1. If you answered "yes" to 1.1 do you think that this requires codification or monitoring? If so how?
   2. If you answered "no" to 1, do you think that there is a maximum reasonable error rate that is acceptable?
      1. Given that you had the means to identify errors, and compute statistics about them, over what period of time would you consider this metric relevant? Alternatively would you consider some other subset, such as the last 10% of an editor's contributions, or the last 1000 edits?
      2. How would you classify the severity of errors, and would this affect your answer to the previous question?
      3. Would the acceptable error rate depend on the responsiveness of the editor in fixing their errors, either pro-actively, or as a result of notification?
2. Do you think that the edits made over suspected IP sock categories cause a problem of defamation of named users?
   1. If you answered yes please explain how, given that these categories are no_indexed, and the red-linked categories are still nagvigable.
   2. Do you think it is a good thing or a bad thing that a tag marking an IP as a suspected sock (when the CU had failed), was removed?
      1. If you think it a good thing, why did you not remove it?
3. Do you think there are known errors on the English Wikipedia created by me, that I have refused to fix?
   1. If yes, please list them, together with refusal to fix.
4. Do you think there are known errors on the English Wikipedia created by me, that I have not yet fixed?
   1. If so please list them for future reference, with approximate dates.

Replies by Beetstra

1) No, not even for an editor at 1 edit per day. 2) No, that concern is absolute <redacted>, 3) Maybe, but are you talking about errors that other editors know about and did report to you (if so, the answer is probably 'no'), or errors that other editors know about and did not report to you (if so, it is yes), 4) Of course there are, but this is a collaborative project, or they will be reported back to you (I hope). --Dirk Beetstra ^{T C} 05:27, 28 April 2012 (UTC)

Replies by Hersfold

1. .
   1. .
      1. .
   2. .
      1. .
      2. .
      3. .
2. .
   1. .
      1. .
3. .
   1. .
4. .
   1. .

Replies by Elen of the Roads

1. .
   1. .
      1. .
   2. .
      1. .
      2. .
      3. .
2. .
   1. .
      1. .
3. .
   1. .
4. .
   1. .

Proposed final decision

Proposals by User:Wnt

Proposed principles

Good editing deserves consideration

This is revised from a proposal at WP:Arbitration/Requests/Case/Cirt and Jayen466/Workshop that was not accepted.

A prolific editor does the work of several average editors, and if he can be replaced, his mistakes will be replaced by those of several average editors. Thus, rare but significant violations of policies, guidelines, or other restrictions should be evaluated as a fraction of the work the editor has done, rather than as an absolute number. Otherwise, every editor, if he is not perfect and edits long enough, will eventually be sanctioned. Specifically, when a large number of actionable edits exist, they should be considered as a fraction of the total number of useful edits made in a similar context, with the threshold being approximately 1 per 1000.

Comment by Arbitrators:

Disagree. Would prefer something along the lines of "In deciding what sanctions to impose against an administrator or other editor, the Arbitration Committee will consider the editor's overall record of participation, behavioral history, and other relevant circumstances. An editor's positive and valuable contributions in one aspect of his or her participation on Wikipedia do not excuse misbehavior or questionable judgment in another aspect of participation, but may be considered in determining the sanction to be imposed." PhilKnight (talk) 10:51, 12 April 2012 (UTC)

I understand what is being driven at here though I wouldn't put it in exactly these words. Some of my questions to the parties (above) attempt to put some specifics on the ideas implicit here. Newyorkbrad (talk) 18:17, 21 April 2012 (UTC)

Comment by parties:

I agree with the principles stated here, but not with the hard number proposed. Rich Farmbrough, 23:55, 11 April 2012 (UTC).

I've never been terribly supportive of principles such as this - to some extent, they (and this one especially) encourage editcountitis, and imply that you can get away with even the worst of disruption provided you make an equivalent number of useful edits. This particular principle also makes it impractical to levy any sort of sanction against Rich, even the existing ones, unless everyone participating in the case are able to provide a total of 1,000 separate problematic edits; a near impossibility given we're limited to 50 diffs each.

All that said, I did not request this case because I believe Rich is "the spawn of the devil," regardless of how he characterizes my motivations here. Rich does fine as an editor, and many of his bot tasks are in fact useful. The problem is that he seems incapable of working within policy and his restrictions, and I don't see that as being compatible with adminship. Hersfold ^(t/a/c) 02:08, 13 April 2012 (UTC)

I think that properly applied they need not encourage editcountitis. For example an editor who writes 100 long articles in one edit each and makes typos in each of them would still be a prolific editor with 100 typos. A typo fixer that fixed (say) 100,000 spelling errors, and at the same time made 100 errors, would would also be a prolific editor with 100 typos. That's why oppose the simplistic numerical idea, but support the principle wholeheartedly. Rich Farmbrough, 23:16, 24 April 2012 (UTC).

Comment by others:

The problem is not that mistakes happen, the problem is that the same preventable mistakes happen, despite repeated requests to fix them. Feedback is dismissed, civility thrown out the window, and WP:BOTPOL ignored (particularly WP:COSMETICBOT). As a side-note, the acceptable fail-rate is also something that is usually left up to BAG's discretion, and putting hard numbers on it would create an ugly precedent. Some tasks have a 0-mistake tolerance. Other tasks can have error rates significantly higher than 1 in 1,000. Headbomb {talk / contribs / physics / books} 20:36, 11 April 2012 (UTC)

Actually, you are wrong here in many respects. Following WP:COSMETICBOT for bot edits is a non-issue. Would that I were given that much latitude. Secondly using this proposal to divert into restating (wrongly) calumnies such as "Feedback is dismissed, civility thrown out the window" is bad form as if you consider these are issues you should submit evidence, not weasel-word them into subsidiary pages. Rich Farmbrough, 23:50, 11 April 2012 (UTC).

The number was proposed in the discussion at the Cirt case; in that case as I recall he would have been on the safe side by a modest ratio. The decision of what actually counts against it, though, is crucial; I don't think that every instance of a chatty comment on a talk page, or a bot edit that takes a space out, should really count against it, but I don't think the edits should have to be at the level of a 3RR violation either; from the Cirt case, more or less, "what's worth citing in the evidence section". If someone can think of a better criterion or threshold by all means propose it; with anything near a consensus I'm more than willing to strike out and revise to get some kind of standard proposed. Wnt (talk) 02:42, 12 April 2012 (UTC)

It's not really something that is amenable to numerical weights. Maybe if one were inclined one could invent a scoring system, there are indeed some academic papers that work along these lines, however evaluating edits is hard, especially when there are people systematically reverting good edits - reverts is one of the key methods the academics use to measure goodness (or badness if you prefer). Rich Farmbrough, 00:05, 18 April 2012 (UTC).

It sounds like the numerical part isn't going over well (though I think there is edit-counting done when they're bad edits). I hope maybe the first two sentences by themselves might find agreement? Anyway, thanks for thinking about this. Wnt (talk) 18:42, 24 April 2012 (UTC)

Proposals by User:Hersfold

Proposed principles

Community sanctions

1) The community has the authority to impose sanctions (such as editing restrictions or bans) on any user whose edits are a detriment to the encyclopedia.

Comment by Arbitrators:

Straightforward. Jclemens (talk) 05:45, 23 April 2012 (UTC)

Comment by parties:

From Betacommand 3. Rich disputes the fact that his editing restrictions apply to him. Hersfold ^(t/a/c) 03:47, 13 April 2012 (UTC)

I have no problem with this principle. However I do have a problem with citing other cases. This seems an attempt to create case law, which is something that ARBcom was explicitly meant not to do when it was formed. Rich Farmbrough, 02:59, 18 April 2012 (UTC).

There is a second problem here. Citing the principle begs the question implicitly, and moves the locus to community sanctions. Resolving firstly whether the complained of edits were causing the problems that were claimed (patently they weren't) seems like the first step. Secondly the other assertions made at the evidence page should be evaluated. Rich Farmbrough, 02:59, 18 April 2012 (UTC).

Comment by others:

Presumably he can both agree with this principle and still dispute that the restrictions apply to him, since he almost certainly denies that his edits are a detriment to the encyclopedia. It would be much more useful(and therefore unlikely to happen) to clarify who gets to make that decision e.g. "The community has the authority to impose sanctions (such as editing restrictions or bans) on any user whose edits the community deem to be a detriment to the encyclopedia" or "The community has the authority to impose sanctions (such as editing restrictions or bans) on any user whose edits Abrcom later deem to be a detriment to the encyclopedia", or amend to specify whoever it is who actually gets to make that decision. 87.254.68.92 (talk) 19:15, 15 April 2012 (UTC)

That's very supportive. I would however agree to a stronger principle, de facto "The community has the authority to impose sanctions". The community may or may not be correct in imposing sanctions (we have seen some classic cases, such as Mantemoreland) but nonetheless "code is law". The English Wikipedia community has chosen not to codify the methodology much, which is a respectable decision (or lack of decision if you prefer) and is generally good on following consensus. It may be that the phrasing you suggest would be a good way to introduce some methodology to sanctions, but that would have to be proposed new procedure for the general community. The good side is that - as with all such procedures - a more equitable situation would obtain. Conversly more complex procedures cause problems and delay of their own. Rich Farmbrough, 15:35, 20 April 2012 (UTC).

If code is law, your repeated deployments of new code without prior consensus amount to trying to pass laws unilaterally. That would seem to make you into a kind of wannabe-dictator. Is that really what you mean?? :-( 67.117.130.107 (talk) 22:50, 26 April 2012 (UTC)

No. Am I that hard to understand? The MediaWiki code gives the community, through the agency of admins, the power to impose sanctions based primarily on the ability to block. Barring 'crat or dev involvement that is the bottom line. This would be better as a finding of fact than a principle. I am of Rich Farmbrough, 00:41, 27 April 2012 (UTC).

Oh, ok. I don't think that's what Lessig meant by "code is law", but whatever. 67.117.130.107 (talk) 04:01, 27 April 2012 (UTC)

As I understand it the original concept relates to the fabric of cyberspace, which in this case is the software supporting the wiki. The architectural nature is of course mutable based on other factors. The importance is to have all these components working towards the same ends, namely making as much of the world's knowledge available to as many of the world's people as possible. Without understanding the interrelation of these components we do not understand anything. I am of Rich Farmbrough, 14:14, 27 April 2012 (UTC).

Recidivism

2) Users who have been sanctioned for improper conduct are expected to avoid repeating it should they continue to participate in the project. Failure to do so may lead to the imposition of increasingly severe sanctions.

Comment by Arbitrators:

Yes. AGF requires editors who are appraised of the negative impact(s) of their actions to avoid those actions in the future; at whatever level this breaks down, the next level of dispute resolution is the next stop. Jclemens (talk) 05:47, 23 April 2012 (UTC)

Comment by parties:

From Betacommand 3, MZMcBride 2, others. This is perhaps a bit harsher than necessary, but given the recurrence of the problems this case focuses on something like this may be useful. Hersfold ^(t/a/c) 03:47, 13 April 2012 (UTC)

I think a one month block is pretty severe - clearly you don't since you decided to raise an Arbcom case in addition.

I also think that there is a basic flaw in the "increasingly severe sanctions" clause. I have pointed out elsewhere in the case that the validity of the restrictions under which I have been sanctioned is, to say the least, dubious. I have not gone into any detail on why they are also poorly thought out and badly worded. Here are some examples of edits that would be technical breeches.

• Removing swear words from an HTML comment
• Leaving a comment
• Adding a {{Pixie me}} template to a page
• Changing a deprecated parameter to a new parameter
• Removing hidden categories
• Fixing certain classes of vandalism
• Adding a template parameter that has recently been removed without my knowledge
• Reverting myself if I realise I have done one of the above unintentionally

These are just off the top of my head. And as Wnt implies below there is no proportionality. One such edit would get an "increasingly severe sanction" compared with the previous, possibly more problematic, dozens of edits getting a "less severe sanction"

The third point is that, despite claims that these sanctions are not punitive, they are clearly meant to be. They are not designed to protect the encyclopaedia, but to "punish".

Fourthly, no thought has gone into the effect of the use of sanctions. I see no evidence form past cases that blocking, for example, is a productive sanction. It can be useful as a cooling off device. And speaking for myself it is upsetting the first time, since a "clean block log" is rather nice. But if anyone thinks that my reaction to being blocked was "Oh dear, the lovely Elen has got mad at me because I have broken all of the Wikipedia I will try to do better." they have neither been paying attention to the case, nor are they a very good student of human behaviour. Blocks of established editors, I believe, create more bad feeling and distrust. In my case I don't particularly blame Elen, though I think she acted hastily. It does however show me that senior admins are prepared to block without due consideration, and in subject areas they don't understand. For myself I have either never, or almost never blocked an established editor, because, even in the cases where I thought it was justified, by the time I had carefully reviewed the evidence someone had already done it. I find it a shame that Admins are so willing, and even eager to block "I'll do it" Short blocks are even desirable in some circumstances - I have only resorted to blocking myself once, but this is not an uncommon occurrence for some editors. To characterise blocks in these circumstances as "sanctions" is then foolish.

Rich Farmbrough, 16:02, 20 April 2012 (UTC).

CORRECTION: "Short blocks are even desirable in some circumstances " should something like read "Short blocks are even something welcomed by the blockee in some circumstances" but worded more elegantly. Rich Farmbrough, 23:21, 24 April 2012 (UTC).

Comment by others:

I think this principle needs to be amended to somehow acknowledge the passage of time and for work done correctly in the meanwhile. If a person is "on probation" after some offense, that probation must end someday. If there is a mentality that penalties must automatically escalate, and that only demerits, never accomplishments, accumulate and mean something, then the participation of any person in Wikipedia must be temporary. Wnt (talk) 22:46, 13 April 2012 (UTC)

This is sage, and reflects the proposed principle you made elsewhere quite well. I would also question the value of keeping information indefinitely on-wiki to be used as a "badge of shame". I was assured that this would not be the use that would be made of combining al "my" AN/I threads into a sup-page, but undoubtedly it has been. Rich Farmbrough, 03:27, 20 April 2012 (UTC).

Good faith and disruption

3) Disruptive behavior driven by good intentions is still inappropriate. Editors acting in good faith may still be sanctioned when their actions are disruptive.

Comment by Arbitrators:

Comment by parties:

From Civility enforcement. Rich has stated that all of his edits address a need/fix a problem/are for the good of the project, including many of the ones pointed out as violations of his restrictions and/or useless and/or otherwise problematic. Hersfold ^(t/a/c) 03:47, 13 April 2012 (UTC)

I have never stated that. Rich Farmbrough, 08:22, 18 April 2012 (UTC).

In fact I specifically made it clear in my evidence that I accept that I make errors. Which has, of course, been used against me. Rich Farmbrough, 03:32, 20 April 2012 (UTC).

A good example of this is Fram's postings to AN, AN/I. If we AGF the were motivated by "good intentions" (although he varies between "I would never block Rich" and "Can't he be indef blocked" which seem to show his intentions are not always in agreement with what he states) - nonetheless they have caused an immense amount of drama and grief. Rich Farmbrough, 03:31, 20 April 2012 (UTC).

Ah, as usual, claims about me without any evidence to back them up or show context. My position is that I would have no regrets about seeing you indef blocked (although it would be much better if somehow your good editing could be retained while getting rid of the problematic edits that brought us here), but that, being clearly involved, I don't take any administrative actions wrt you (with the result that I will not block you or your bots any more, and that I don't delete pages you created, apart from pure housekeeping like empty older monthly maintenance cats created by Femtobot). These two positions are not at odds with each other and don't "vary", they are logically connected. 07:03, 20 April 2012 (UTC)

Comment by others:

Rich and Fram might be referring to this. I think Fram's interpretation above is consistent with the diff. Kingpin13 says Fram should not block Rich due to WP:INVOLVED, and Fram agrees. That doesn't conflict with Fram's desiring on other occasions for someone else to block Rich. 67.117.130.107 (talk) 22:21, 26 April 2012 (UTC)

Administrators

4) Administrators are trusted members of the community, and expected to lead by example and 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 poor judgment, multiple violations of policy (in the use of Administrator tools, or otherwise), or particularly egregious behaviour, may result in the removal of administrator status. Administrators are also expected to learn from experience and from justified criticisms of their actions.

Comment by Arbitrators:

Comment by parties:

From Civility enforcement, MZMcBride 2, others. This principle is the core reason why I requested this case; as demonstrated in the evidence (or soon to be, mine isn't quite done yet), Rich has time and again exercised poor judgment in the execution of his scripts and violations of his restrictions, and has repeatedly breached policies regarding bot and script use, civility, and use of the tools (mainly relating to unblocking his own bots). Hersfold ^(t/a/c) 03:47, 13 April 2012 (UTC)

It is standard practice to stop a bot when there is a problem with it's functioning. It is also standard practice to restart it when the problem is resolved. For later incarnations of SB/HPB the mechanism is block unblock (for earlier incarnations, until I was forced to drop AWB by the drama-mongers a talk page message sufficed). There is no abuse of privilege here, in general the blocking admin has explicitly said the bot can be unblocked when the issue is resolved. Barking up the wrong tree with this one. Rich Farmbrough, 08:28, 18 April 2012 (UTC).

Comment by others:

Offensive commentary

5) Repeated use of sarcasm, wordplay formulated to mock another user, casting aspersions on an identifiable group, or use of language that can reasonably be anticipated to offend a significant segment of the community is disruptive, particularly when it distracts from the focus of an ongoing discussion on communal pages such as those in the Wikipedia namespace.

Comment by Arbitrators:

Comment by parties:

From Civility enforcement. While again this goes much farther than necessary for this case, part of it are relevant; Rich has repeatedly responded to good-faith concerns with sarcasm, derision, or attacks. Hersfold ^(t/a/c) 03:47, 13 April 2012 (UTC)

Er... I don't think so. Firstly there is a difference between good faith concerns and long term stalking. Secondly I have by and large been extremely restrained in my responses. Thirdly you have failed completely to grasp the intent of my comments relating to process and one or two other matters, which makes me wonder if you understand any of my responses, except on a technical level. This also applies to the email Elen shared with you. Rich Farmbrough, 00:12, 18 April 2012 (UTC).

Comment by others:

This is a most unfortunate principle. Prohibiting expressions of prejudice doesn't work; consider the effects of formal bans on hate speech on progress in racial and religious toleration in the U.S. versus Europe. The harder you try to enforce civility, the more rancor ensues as people seek sanctions against one another. In any case, this principle, as admitted, is apparently mostly irrelevant to the case. Please, chop it down to what actually applies here - these should be relevant principles for the case, not promulgations of ArbCom legislation. Wnt (talk) 22:52, 13 April 2012 (UTC)

Surely casting aspersions on an identifiable group must be permitted where the aspersions are true and of relevance to the project? Past a certain point clearly they should be pursued only through dispute resolution though. You're casting aspersions on Rich Farmbrough, after all. I suppose you can say he isn't a group but there are plenty of similar cases that do involve identifiable groups of people. 87.254.68.92 (talk) 19:51, 15 April 2012 (UTC)

I can accept that some discussions get a little bit heated, so incivility per se doesn't come across to me as a big issue in this dispute. I'm more concerned about self-serving obtuseness similar to WP:CPUSH that we sometimes see in mainspace disputes. If Rich makes 20,000 edits with AWB that are all of a similar nature, those are automated edits and need a BRFA. "Manually approving" the edits by leaning on a "yes" button while the script does its thing doesn't change that. 67.117.130.107 (talk) 22:31, 26 April 2012 (UTC)

Role of bots and scripts

6) Bots are processes that modify Wikipedia content in a fully or partially automated fashion. Scripts are also computer algorithms utilized to automate or semi-automate certain types of editing. These tools are extremely valuable for the purpose of facilitating the making of multiple edits that would be unduly time-consuming or tedious for a human editor to perform manually. Approval from the Bot Approvals Group is generally required before an editor may use a bot for automatic or high-speed edits.

Comment by Arbitrators:

Comment by parties:

From Date delinking. Rich has repeatedly run automated scripts and bots from his main and bot accounts without approval for the edits being made. Hersfold ^(t/a/c) 03:47, 13 April 2012 (UTC)

No I haven't. Rich Farmbrough, 00:57, 18 April 2012 (UTC).

See Hersfold's explanation at point number 4. Whenaxis (contribs) ^{DR goes to Wikimania!} 00:59, 18 April 2012 (UTC)

What? that he hasn't had time to complete his evidence? Rich Farmbrough, 02:28, 20 April 2012 (UTC).

Comment by others:

Regarding Hersfold's statement. In March 2012 alone, R.F. made 2439 edits on his main account with the summary "general fixes"; 2142 "create wanted category"; 1679 "general fixes using" (that is not a typo, he did use that summary); 1180 "track this category"; and around 800 with summaries similar to "Add ref section minor fixes". None of these tasks was approved by BAG or, as far as I can tell, discussed anywhere before it was done. In January 2012 he made 7,792 edits on his main account with the edit summary "Metadata and general fixes using AWB", and 20,245 total edits from his main account. In January 2011 he made 24,000 edits from his main account all to tag empty sections, with 38,000 edits total that month on his main account. The claim that manual attention was given to each of these edits is incredible. Large-scale jobs with these magnitudes should have had BAG approval and should have been run on a bot account. — Carl (CBM · talk) 10:35, 18 April 2012 (UTC)

Responsibility of bot operators

7) Like administrators and other editors in positions of trust, bot operators have a heightened responsibility to the community. Bot operators are expected to respond reasonably to questions or concerns about the operation of their bot. An editor who (even in good faith) misuses automated editing tools, such as bots and scripts, or fails to respond appropriately to concerns from the community about their use over a period of time, may lose the privilege of using such tools or may have such privilege restricted.

Comment by Arbitrators:

Comment by parties:

From Date delinking. Rich has repeatedly failed to respond to good-faith concerns regarding his scripts and bots, and has repeatedly used them (in good faith) to violate his editing restrictions regarding their use. Hersfold ^(t/a/c) 03:47, 13 April 2012 (UTC)

Er... I was never a party to that arb case (or indeed any other arb case). Looks like you are just making it up as you go along now? Rich Farmbrough, 00:15, 18 April 2012 (UTC).

I'm not implying you were a party to the case, nor any other I've quoted principles from. I'm simply giving credit where it's due. It's common to reuse principles and other boilerplate text used in past cases, as often the same principles can be applied to multiple cases. In this case, I feel as though this principle regarding bot operators and expectations of them applies here. Hersfold ^(t/a/c) 01:28, 18 April 2012 (UTC)

I was working my way up the page for some reason. I since saw that these are just attempts to abstract principles. Rich Farmbrough, 09:02, 18 April 2012 (UTC).

I think that it is a stretch to say that I have "repeatedly failed to respond to good faith concerns." I invite you to look at some sample months and compare the number of "responses" to "failures to respond". Even if you restrict the scope to bot/script threads (or any other type of thread) you will have trouble finding more than the odd "failure to respond" in all my talk page archives. Rich Farmbrough, 09:02, 18 April 2012 (UTC).

Comment by others:

"Ignore all rules"

8) Wikipedia policies and guidelines are largely meant to be flexible and applied with common sense; limited violations of policies and guidelines are permitted when adherence to them prevents or impedes efforts to improve the project. However, this principle may not be used to defend actions considered to be disruptive by the wider community.

Comment by Arbitrators:

Eh, I'd word it differently. Something along the lines of "while rules can be ignored to make improvements, the community is empowered to decide what is an improvement and what is not." Jclemens (talk) 06:11, 23 April 2012 (UTC)

Comment by parties:

Self-written, although I'm sure a better version exists somewhere; just not in the cases I looked through, and I don't care to dig any further than I already have. Anyway, Rich has repeatedly claimed that his actions were done with the sole intent of improving the project, essentially invoking WP:IAR whenever challenged. While valid in a limited extent, Rich has stretched "ignore all rules" to and beyond its breaking point, hence the sanctions placed against him. Hersfold ^(t/a/c) 03:47, 13 April 2012 (UTC)

And as discussed the "sanctions" are based on one !vote, and are therefore pretty much ignoring any and all rules. So the purpose of bringing this up is to attempt to draw the teeth of IAR, however unless you clearly demonstrate significant harm to the project, over and above that caused by the continual obsessing by Fram and Carl, this principle is not relevant to the case. Rich Farmbrough, 01:01, 18 April 2012 (UTC).

Comment by others:

The key point of IAR as it is currently understood is that it allows editors in specific situations to bend or ignore the rules, but not to ignore the rules chronically as if they don't exist. The original meaning of IAR was that way (the "if rules make you nervous, just forget they exist" version [7]), but the meaning of IAR has changed since then. For at least the last 5 years, the general consensus has been that IAR is meant to apply to specific special cases, not to general everyday editing. In particular, except in very rare occasions, there is no reason a bot operator should need to invoke IAR to circumvent the bot approval process. — Carl (CBM · talk) 11:05, 18 April 2012 (UTC)

That is a very good phraseology. Rules do make me nervous and depressed. Particularly when I know the kind of absurd and in some cases very unpleasant consequences that flow from breaking them, even by accident or in immaterial ways. However it is a point of principle that bullying behaviour, even if it has nominal backing from "rules" (or should I say "actual rules"?) and defiantly where it doesn't should not be allowed to triumph. To paraphrase "rules were made for editors, not editors for rules". Anyone who doesn't understand this has no business editing here. Rich Farmbrough, 01:09, 20 April 2012 (UTC).

I'm quite pro-IAR myself. As stated elsewhere, I see the problem here not as ignoring rules, but as ignoring other editors. If you do an IAR for something sensible and it gets acceptance, that's great. If you get pushback, cut it out. You have gotten tons of pushback on this stuff. 64.160.39.210 (talk) 07:58, 20 April 2012 (UTC) Added: it's unseemly for an IAR practitioner to wikilawyer the way you've done in various places on this page. IAR requires being able to ground your judgment in the project's goals and principles, which transcend the wikilawyered arcana of the policy documents. "To live outside the law, you must be honest." (Bob Dylan, supposedly). 64.160.39.210 (talk) 08:04, 20 April 2012 (UTC)

I think you need to be precise on "You have gotten tons of pushback on this stuff" for it to be a useful comment. It is certainly true that there was a relatively large amount of pushback on some stuff I was doing two years ago which I stopped doing. Not necessarily saying I was right or wrong to do it, it simply was causing too much trouble for minor benefit (or as some said minor disbenefit). That could have been later resolved, maybe if anyone ever wanted to try and form consensus.

As far as "real errors" are concerned, I don't think "pushback" is the right term, I would expect anyone spotting an error to notify me if they thought there was a reasonable chance that it was systemic, just as I would (and have) for others. That is part of an improvement culture. I would also expect them to do it in a polite, non-loaded and preferably friendly manner. Maybe that's too much to ask.

Then we have aggressive hounding over trivia - Elen calls them "niggly" complaints. This isn't "pushback" this is behaviour which is not going to get the results it seeks, certainly not in the short term. Building this sand-castle of complaint around "Otheruses4" will not stop the tide coming in, and Canute will get his feet wet, the only question is how many bodies will be buried in the sand. Rich Farmbrough, 19:33, 23 April 2012 (UTC).

Rich, that sounds like IDHT. There are many ANI threads concerning your editing that are much more recent than 2 years ago.[8] This one by itself (Nov 2011) was opened by Xeno and has critical comments towards you from numerous editors besides CBM and Fram. Here is another (January 2012) with some comments from me. In that thread, there was a discussion (in my reading) heading towards increasing your editing restriction, before Roux stunk up the place with his obvious animosity and the direction changed. Of course there's plenty more.

As for "niggly" complaints, one theory is that come from people who don't share your editing philosophy and consider most or all of your bot tasks useless, including the approved ones, and they don't necessarily see BAG approval as also indicating community desire for the bot.[9] They know that even the bots that the community does think it wants shouldn't in all cases be approved (that's the point of turning approval over to "experts" after all). They may choose not to bother trying to contest the approved tasks, but instead respond vigorously when you overstep the approvals, something like when an admin breaks UNINVOLVED to do something that is otherwise not egregious on its face, parties who don't like the action will invariably kick up a fuss. So there is really sort of a content dispute underneath all of this.

I've started writing another post (I might not finish it today) discussing this issue further, and suggesting that you and some of the other parties agree to enter mediation after this case closes. I think it could be useful to get your philosophical and editorial differences out into the open and try to work out reasonable compromises, but arbitration isn't such a good venue for that. Without such compromises, what's left are the blunt instruments of policy enforcement (where we are now) and I don't think your long-term prospects are good in that setting. Let me know what you think. 67.119.3.53 (talk) 19:04, 24 April 2012 (UTC)

Of course I am open to Dispute Resolution and Mediation. And constructive criticism, constructively delivered. And straightforward query and discussion. Sure it is my fault that I mishandled the events of late 2010, got overzealous with what I was doing and blanked my talk page when a couple of complainants got overzealous with what they were doing, I do feel, though, that, given that as a result, I made massive changes (far in excess than actually corresponded to the original complaints) I should not still be suffering for these mistakes. But maybe I am wrong, I brought attention to my gnoming, and now it is going to be subject of scrutiny (which I don't mind) and perpetual drama (which I do) for all eternity. Rich Farmbrough, 23:36, 24 April 2012 (UTC).

I'm glad to hear of your openness to resolving issues through discussion. I can accept you had a change of approach around late 2010. For clarification I'd appreciate if you could indicate whether January 2011 (specifically this incident) falls under the new or the old approach. If it's the new approach, I think there's still a problem. 67.117.130.107 (talk) 17:56, 26 April 2012 (UTC)

It's a completely different sort of error, and something I don't intend to repeat. It has no relation to the changes in the way automated edits were done, which are largely about removing on-the-fly improvements until consensus can be documented, and ensuring that the 2-3 per thousand cosmetic only changes were not saved. I am of Rich Farmbrough, 00:50, 27 April 2012 (UTC).

Proposed findings of fact

Rich Farmbrough

1) Rich Farmbrough (talk · contribs · blocks · protections · deletions · page moves · rights · RfA) is an administrator on Wikipedia, active since 2004. During his time on the project he has made nearly one million edits, more than any other human editor on the project. Many of these edits have been made with the assistance of the semi-automated tool AutoWikiBrowser.

Comment by Arbitrators:

I understand that this language is no longer quite accurate. Jclemens (talk) 06:12, 23 April 2012 (UTC)

Comment by parties:

The usual "introduction" finding common in these cases. Hersfold ^(t/a/c) 03:47, 13 April 2012 (UTC)

Comment by others:

Wait, that second sentence is bogus. Those million edits were mostly made by Rich's bots, not by his editing in his human capacity. You're writing congratulatory language about Rich running bots on his main account, which I thought was one of the things we're trying to get stopped. 64.160.39.210 (talk) 08:09, 20 April 2012 (UTC)

No if you throw in the bots it's more like 5 million. Rich Farmbrough, 09:37, 21 April 2012 (UTC).

Editing restrictions

2) Rich Farmbrough (talk · contribs · blocks · protections · deletions · page moves · rights · RfA) is subject to two editing restrictions placed by the community as a result of separate discussions at the Administrator's noticeboard and Incidents noticeboard. The first of these, placed in October 2010, prohibits Rich Farmbrough from making cosmetic changes to wikicode beyond those enabled by AutoWikiBrowser's default settings and/or those explicitly approved by the community's consensus or the Bot Approvals Group. The second, placed in January 2011, prohibits Rich Farmbrough from mass creation of pages in any namespace without approval from the community. Rich Farmbrough has violated these restrictions on a number of occasions: <diffs to come, going to add more to evidence before the period ends>

Comment by Arbitrators:

Comment by parties:

The primary concern with Rich's actions from multiple editors is his repeated violation of and disdain for these editing restrictions. Hersfold ^(t/a/c) 03:47, 13 April 2012 (UTC)

Please note that the only support !vote for that ER was cast by User:Dirk Beetstra - I'm not sure if Dirk would stand by that vote today, but I am inclined to believe from his positive and helpful comments that he would not be terribly supportive of what the ER has been used for. If any member of the community feels that one support !vote is sufficient to carry an ER, I would ask them to think very carefully about what such a position would mean. Rich Farmbrough, 00:28, 18 April 2012 (UTC).

Comment by others:

I presented evidence on the validity of the edit restriction. R.F. has had ample opportunity to have the restriction text removed from WP:ER if it really is not valid, so its continued presence there after it was used to justify multiple blocks is prima facie a sign that the restriction is in effect. — Carl (CBM · talk) 11:10, 18 April 2012 (UTC)

I take it as a sign that when you take something to AN/ANI, people do not read the original ER discussion. On getting it repealed, surely you are either joking or have never seen an ER try to be repealed. Jenks24 (talk) 10:30, 19 April 2012 (UTC)

Once Hersfold posts the links to the AN discussions, I would strongly encourage the arbs to have a thorough read of the original editing restriction discussion. IMO there is no way that there was a consensus to place a ER on Rich. Jenks24 (talk) 10:30, 19 April 2012 (UTC)

Actually, there were 16 people in support of much harsher restrictions, while 10 opposing that harsher remedy: some of those opposed any restriction, and some wanted a less severe restriction, which he eventually got. Taking into account the total number of people supporting either the harsher restriction or the one actually implemented, vs. the number of people opposing either one, I do believe that it is fair to say that there was a consensus for restrictions, and that he came of lucky with the one he actuall got. (This refers all to the first restriction he got, the second followed three months later). I provided links in my previous comment[10]. The second restriction was implemented soon in the discussion (Wikipedia:Administrators' noticeboard/Incidents/Rich Farmbrough/January 2011, but got support from a fair number of people as well, and no clear opposition, so no problem with that one either IMO. Fram (talk) 11:03, 19 April 2012 (UTC)

Which discussion are you talking about? I'm talking about this one from October 2010. On a quick glance, there looks like less than 16 people in that whole discussion, so I think you must be talking about something else. Jenks24 (talk) 04:32, 20 April 2012 (UTC)

The one immediately preceding this one, Wikipedia:Administrators' noticeboard/Incidents/Rich Farmbrough/October 2010, which also contains the section you linked to. The section "Edit restriction proposal for Rich Farmbrough" is the one with the most participants and the one I took my tally from. Fram (talk) 07:11, 20 April 2012 (UTC)

Hmm, reading the AN archive is far different to reading ANI subpage where it seems that quite a few separate threads have been merged to make it appear like it was one big thread (nothing nefarious in it, I'm sure). The way I see it is that one very harsh proposal did not get consensus, although I will grant you that there was support for it, and then a second, separate, more lenient proposal got one support and was enacted. Not saying ArbCom should ignore this, but I don't think they should give it that much weight. Jenks24 (talk) 02:22, 22 April 2012 (UTC)

And this was interrupted - as the consensus was swinging away from it. So really is there any point bringing up ERs that even CBM does not think were enacted, except to be nasty? Rich Farmbrough, 02:30, 21 April 2012 (UTC).

The links are all here. — Carl (CBM · talk) 13:13, 19 April 2012 (UTC)

Thanks Carl. Jenks24 (talk) 04:32, 20 April 2012 (UTC)

As I am named here, I will comment on this. My support there, was purely for the point that I do think that purely cosmetic changes should not be saved. If they are part of a larger edit, then that is totally fine with me, and I would actually encourage them! Whether CBM, Fram, me, R.F. or whoever makes a truly cosmetic edit, I believe that that should be strongly avoided. That being said - I will not support any form of bashing of whoever if they have made in a series of edits an occasional cosmetic edit. They should do their utmost best to avoid it (and I have suggested, while it was implemented, an AWB-trick that should help to avoid it), but, since it does not break anything, there is, never a reason to hold that against an editor who does save purely cosmetic edits. I find it utterly pathetic that certain editors here keep bringing up that put fault on editors because they save cosmetic edits, and who rip the existence of some purely cosmetic changes totally out of the context of the total edits. That being said, unfortunately we are not here to see whether there is fault with edit restrictions, and I doubt whether ArbCom will rule on whether the implementation of an edit restriction has been reasonable, or even whether the implementation of the edit restriction was appropriate. --Dirk Beetstra ^{T C} 07:39, 22 April 2012 (UTC)

Rich Farmbrough's conduct

3) In responding to concerns about potential violations of policy and/or editing restrictions, Rich Farmbrough (talk · contribs · blocks · protections · deletions · page moves · rights · RfA) has repeatedly replied in a sarcastic or incivil tone, in some cases attacking the motivations of those expressing the concerns: [11] [12] [13] [14] [15]

Comment by Arbitrators:

Comment by parties:

This is the second of the three primary concerns I have with Rich's actions; Rich is unable to assume good faith on the behalf of those expressing concerns that he has erred, and instead of taking responsibility for or working to improve his actions, attacks or belittles the concerns or those expressing them. Hersfold ^(t/a/c) 03:47, 13 April 2012 (UTC)

So you continue to introduce more assertions and evidence? I'm not sure this is workshopping, it looks like abuse of process to me. Rich Farmbrough, 00:29, 18 April 2012 (UTC).

I'd strongly suggest you review the guide to arbitration and look through some past cases before accusing me of any further misconduct. Yes, there are allegations here, supported by evidence I and others have already posted on the Wikipedia:Arbitration/Requests/Case/Rich Farmbrough/Evidence page, because this is a finding of fact. It's what these are here for. Hersfold ^(t/a/c) 01:32, 18 April 2012 (UTC)

But it's not fact. It's a complete lack of empathy, and a sense of humour failure. When someone says...

Because obviously there are so many other pages starting with "Michael Schimmel Center" which would be incorrectly sorted without this defaultsort. Thanks, this is a prime example of a useless edit that follows the letter of a (also useless) rule, but does absolutely nothing to improve the encyclopedia.

(<== This really is sarcasm.) to respond with a <sigh> is surely much more relaxed than to take someone to Arbcom for being sarcastic. Especially when this editor has been beating me up with "the letter of a (also useless) rule" for two to three years. I get a lot of rough treatment, I was called a "Fucking liar" while attempting to prevent abuse of process, I did not respond to that. It's all very well for you to cherry pick a few diffs, and criticize my responses. I understand that you might not like wordplay, but it is clearly stated when you edit my talk page that you will benefit from understanding my sense of humour. Rich Farmbrough, 07:16, 18 April 2012 (UTC).

Comment by others:

Rich Farmbrough violated bot policy

4) On several occasions, Rich Farmbrough (talk · contribs · blocks · protections · deletions · page moves · rights · RfA) has run scripts from his main account at bot-like rates without approval from the Bot Approvals Group: [16]; at other times, he has run tasks approved to be run from a designated bot account from his main account, while said bot was blocked: [17]

Comment by Arbitrators:

Comment by parties:

This is the last of the primary concerns I have with Rich's actions; Rich will act against policy when he deems it necessary for the good of the project, even when the community has previously said that the actions he is taking are in fact not as beneficial as he believes. Hersfold ^(t/a/c) 03:47, 13 April 2012 (UTC)

Neither of these are against bot policy if they are done manually Rich Farmbrough, 06:44, 18 April 2012 (UTC).

Yes they are. WP:MEATBOT. Headbomb {talk / contribs / physics / books} 17:05, 18 April 2012 (UTC)

Read it. Rich Farmbrough, 02:44, 20 April 2012 (UTC).

Comment by others:

The bot policy does make an exception for manual edits. However, as I describe in my comment dated 10:35, 18 April 2012 above, R.F. has routinely undertaken jobs on his main account that should have gone through bot approval, and which are not "manual" in any significant way. I have warned him at least once [18] about running bot jobs from SmackBot on his main account. I realized later that SmackBot was blocked at that time (2010-11-15 to 2010-11-23). Using the bot operator's account to evade a block of the bot cannot be justified under the "manual edit" exception. — Carl (CBM · talk) 11:23, 18 April 2012 (UTC)

Running a script from a main account is not forbidden, in fact, we have many editors allowed to edit using AWB on their main account (which can be run in a bot-like way, position your mouse on the save button, your finger on left mouse button, close your eyes and move your finger up and down in a repetitive way). The key is that the task you are running on your main account should be with the utmost consideration of what is happening. That sentence in the bot policy is to prevent people from doing inconsiderate bot(-like) actions, it is not there to prevent people do do bot(-like) actions (that is, to prevent damage, or to allow to prevent damage - not to prevent editing). Reading the bot policy as if it is not allowed to do bot-tasks on your main account is overly bureaucratic (if not completely wrong) and in violation of WP:IAR. --Dirk Beetstra ^{T C} 05:52, 22 April 2012 (UTC)

Rich writes, "It's often easier to do something with AWB than have a BRFA approved. The break-even seems to vary between a few hundred edits and, currently, about 20,000."[19] I would hope any batch of 20,000 edits for a single task needs some sort of positive consensus. It's just not ok to do that large an operation unilaterally. 64.160.39.210 (talk) 06:33, 22 April 2012 (UTC)

No, I think that you do not need a consensus before you start a task (WP:BOLD and such). If you go through the BRFA and get the consensus, good, fine, perfect, great, but there is, per policy, no must. If you do a task which has 'apparent consensus', you do 50.000 edits to mainspace, fixing a problem or (significantly) improving Wikipedia using a self-written script, and get no complains (what, you get even praise), you do not break anything, then there is no problem. That does, per policy, not require a BRFA. If the edits contain significant real errors (and not, what is generally the complaint, edits that significantly improve the page but break something (or propagate something that was already broken), or that do not break a page, do not improve a page, but are merely cosmetic), then if those errors are too many, then that sequence of editing should be stopped. If you do that with an approved bot, the bot will be blocked, if you do it with an unapproved bot, the 'bot' will be blocked - if it is unapproved the trout-slapping will exceed the trout-slapping you receive when the task was approved, but, still, policy does not require you to approve the task.

That being said, this statement by Hersfold is true, and this, or something similar, will likely end up in the final decision, and will likely get a lot of support by the Arbitration Committee (obviously: it is true). There is of course nothing proven to be wrong, or nothing significant here, but that does not matter. The implication that R.F. is doing something wrong is enough. --Dirk Beetstra ^{T C} 07:13, 22 April 2012 (UTC)

BOLD is ok for one edit. Maybe even two or three edits. Not 20,000. We had that same discussion in Betacommand 3. BOLD only makes sense if it's something that can be easily reverted by any other editor (i.e. without a bot) if it wasn't the right thing. And the multiple blocks of both Betacommand and Rich shows the absence of "apparent consensus" for those operations. 64.160.39.210 (talk) 08:06, 22 April 2012 (UTC)

No, that is a wrong conclusion. Please read again what I said. --Dirk Beetstra ^{T C} 08:11, 22 April 2012 (UTC)

I did read it, I just think you are mistaken. See the many "fait accompli" principles from arb cases over the years. 64.160.39.210 (talk) 15:55, 22 April 2012 (UTC)

That does not make them right. --Dirk Beetstra ^{T C} 05:20, 23 April 2012 (UTC)

Further to 64.160.39.210: You know that the community generally approves of tasks which can very easily be done by a bot, and which are performed on thousands of pages at speeds way higher than what could be called 'script assisted', and which certainly were not overseen on a case-by-case basis? Those edits do not violate the bot policy more than what Rich (or Δ) have been doing on their main account. --Dirk Beetstra ^{T C} 10:39, 23 April 2012 (UTC)

WP:BOTPOL#Bot_usage clearly says, "Administrators blocking a user account suspected of operating an unapproved bot or an approved bot in unapproved ways should block indefinitely... Note that high-speed semi-automated processes may effectively be considered bots in some cases, even if performed by an account used by a human editor. If in doubt, check." (Italics per original). I don't know how you're helping your argument in saying "Those edits do not violate the bot policy more than what Rich (or Δ) have been doing on their main account." Rich and Δ both found themselves defending arb cases due to those very activities, with Δ's most recent case resulting in a long term site ban. 67.119.3.53 (talk) 07:45, 24 April 2012 (UTC)

Well, that's what I have been saying after Δ's case, and what I say now .. that ArbCom decided that does not make those decisions right, does it? Admins have the specific possibility to run bot-type actions on their account (add '&bot=1' to url), editors can run scripts on their accounts, or do many repetetive actions on them, Twinkle allows for mass actions, and no-one will complain that they are bot-like actions. --Dirk Beetstra ^{T C} 08:05, 24 April 2012 (UTC)

Block logs of Rich Farmbrough and his bots

5) The block logs of Rich Farmbrough (talk · contribs · blocks · protections · deletions · page moves · rights · RfA) and his two primary bot accounts, SmackBot (talk · contribs) (now inactive) and Helpful Pixie Bot (talk · contribs), show a large number of blocks, many for violation of editing restrictions and/or editing outside of Bot Approvals Group approval.

Comment by Arbitrators:

Comment by parties:

Evidence of recurring problems. Hersfold ^(t/a/c) 03:47, 13 April 2012 (UTC)

Lets examine HPB's block log.

• (del/undel) 01:24, 2 April 2012 Courcelles (talk | contribs | block) blocked Helpful Pixie Bot (talk | contribs) (account creation blocked, autoblock disabled) with an expiry time of indefinite (The main account unblock was only procedural for arbitration participation. Conditional unblock is for arbitration-related edits ONLY.) (unblock | change block)
  • Block for user blocked'
• (del/undel) 06:06, 31 March 2012 Courcelles (talk | contribs | block) blocked Helpful Pixie Bot (talk | contribs) (account creation blocked, autoblock disabled) with an expiry time of 1 month (Procedural block. The bot should not be editing while its owner is blocked. If owner is unblocked, please unblock this bot. ) (unblock | change block)
  • Block for user blocked'
• (del/undel) 02:16, 26 March 2012 Hersfold (talk | contribs | block) blocked Helpful Pixie Bot (talk | contribs) (autoblock disabled) with an expiry time of indefinite (https://en.wikipedia.org/w/index.php?title=User_talk:Rich_Farmbrough&oldid=483941317#Editing_restrictions_and_Pixie_Bot - bot violating operator's editing restrictions, blocked until assurances are received it won't continue to do so) (unblock | change block)
  • Block for ER'
• (del/undel) 18:45, 14 March 2012 NuclearWarfare (talk | contribs | block) blocked Helpful Pixie Bot (talk | contribs) (autoblock

disabled) with an expiry time of indefinite (Breaking URLs) (unblock | change block)

• • Block for bug'
• (del/undel) 12:27, 18 October 2011 MSGJ (talk | contribs | block) blocked Helpful Pixie Bot (talk | contribs) (autoblock disabled) with an expiry time of indefinite (previous assurances not being ahdered to) (unblock | change block)
  • Block for strange reason'
• (del/undel) 14:24, 17 October 2011 Rich Farmbrough (talk | contribs | block) blocked Helpful Pixie Bot (talk | contribs) (autoblock disabled) with an expiry time of indefinite (temporary halt) (unblock | change block)
  • Block for bug'
• (del/undel) 11:51, 23 September 2011 MSGJ (talk | contribs | block) blocked Helpful Pixie Bot (talk | contribs) (autoblock disabled) with an expiry time of indefinite (bot adding inappropriate templates to pages, no response from operator, and editing talk page does not stop bot) (unblock | change block)
  • Block for bug'
• (del/undel) 16:35, 9 September 2011 MSGJ (talk | contribs | block) blocked Helpful Pixie Bot (talk | contribs) (autoblock disabled) with an expiry time of 1 week (bot operator currently blocked) (unblock | change block)
  • Block for user blocked'

So here we see one (1) block for ER. Hardly "many". Rich Farmbrough, 06:51, 18 April 2012 (UTC).

Comment by others:

No, Hersfold, it is a block log, that is not evidence of recurring problems. --Dirk Beetstra ^{T C} 07:17, 22 April 2012 (UTC)

Previous request for arbitration

6) In November 2011, a request for arbitration was filed regarding allegations that Rich Farmbrough had repeatedly violated two editing restrictions placed against him by the community. The request was rejected with a vote of 7 in favor of accepting, 5 against, and 2 recused.

Comment by Arbitrators:

Comment by parties:

Not entirely sure this is worth noting, but does seem a good marker indicating that problems were noted in the past, and that the Committee did seriously consider reviewing the matter at that time, only narrowly voting against doing so. Hersfold ^(t/a/c) 03:47, 13 April 2012 (UTC)

Or.. it could be a good marker that the problems were seen as overblown then, and that this case wouldn't have been accepted this time if I hadn't asked for it to be. Rich Farmbrough, 06:54, 18 April 2012 (UTC).

Comment by others:

Needs copy editing. I'm guessing that the first sentence should end "was filed" or "was submitted" or something like that? Or maybe it should start "In November 2011 there was a request for arbitration..."? 87.254.68.92 (talk) 20:00, 15 April 2012 (UTC)

Fixed. Hersfold ^(t/a/c) 01:34, 18 April 2012 (UTC)

So? (That being said, it is true, so likely suitable as a finding of Fact showing that there is a problem). --Dirk Beetstra ^{T C} 10:40, 23 April 2012 (UTC)

Proposed remedies

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.

Rich Farmbrough desysopped

1) For repeated violation of community placed sanctions and Wikipedia policies, Rich Farmbrough (talk · contribs · blocks · protections · deletions · page moves · rights · RfA) is desysopped. He may reapply for adminship at Requests for Adminship at any time.

Comment by Arbitrators:

Is there any other active current administrator who has a block log as long as Rich Farmbrough's? Jclemens (talk) 06:32, 23 April 2012 (UTC)

Comment by parties:

This is the primary outcome I am seeking from this case. I feel that Rich's disregard for his editing restrictions, violations of the bot policy, and poor communication regarding concerns raised about both are all incompatible with adminship. Administrators are meant to be role models of sorts amongst the community, and responsible for enforcement of the same restrictions and policies Rich has himself violated. Hersfold ^(t/a/c) 03:47, 13 April 2012 (UTC)

And again you assume that which you have signally failed to prove. In every case there have been brave editors who have stood up even to the mob-mentality of AN/I back in 2009/10 (now, thank heavens, substantially improved), while in general those who have been "taking me to task" have been the same two or three (or I could even say one) editors. It is interesting that you came to my talk page and asked me to change my bot, which I did, and as a result you are asking for me to be de-sysopped.Rich Farmbrough, 01:06, 18 April 2012 (UTC).

Because the problems continued, hence why you were blocked. Obviously we're going to disagree on this, however. Hersfold ^(t/a/c) 01:37, 18 April 2012 (UTC)

A new issue arose, which I think has been conclusively demonstrated not to be a problem. That is completely different from "the problems continued". Rich Farmbrough, 02:47, 21 April 2012 (UTC).

Elen of the Roads is kind enough to say I am "not a particularly bad admin" - in British understatement that is quite a compliment (at least reaching as high as praising with faint damns). Of course she may mean it in another dialect. Rich Farmbrough, 04:02, 18 April 2012 (UTC).

Comment by others:

Yeah, unfortunately, I think Rich probably doesn't have the trust of the community to remain an administrator at this time. Sorry it had to end this way. Master&Expert (Talk) 21:57, 15 April 2012 (UTC)

Meh. I suppose this is justified by policy, but as far as I can tell, it's really the bots that are the problem. And as we saw with Betacommand, desysopping won't necessarily help. 67.117.147.20 (talk) 03:09, 16 April 2012 (UTC)

A good observation, 67.117.147.20 - "desysopping won't necessarily help" - it would not even address the issues. --Dirk Beetstra ^{T C} 05:39, 16 April 2012 (UTC)

However, it would remove automatic access to AWB, which has proven to be a problem in the past, as Rich has used it to engage in rapid-fire bot (or botlike) edits on his main account. If Rich is violating the bot policy this frequently, and yet is still allowed AWB access, then it will remain a problem even if he's restricted to his main account only, IMO. (This is why I'm not a fan of the whole "gadgets" panel concept; it makes it impossible to restrict access to them if someone is proving problematic with them.) rdfox 76 (talk) 11:36, 16 April 2012 (UTC)

Maybe you folks know something that I'm missing, but I don't see how limiting AWB makes much difference. AWB is just a tool, and if that particular tool becomes unavailable, there are lots more like it, or the existing one can be downloaded and modified; and Rich is also capable of developing his own tools. Imagine dealing with a serial axe murderer by restricting his/her future access to buying some particular model of axe. It doesn't sound like that promising a solution, IMHO. 67.117.147.20 (talk) 22:30, 16 April 2012 (UTC)

While I am not sure being compared with a serial axe murderer is very flattering, it is true that, ironically, User:CBM forced me (by threats and actual blocking of SmackBot) to download and tweak the code of AWB, so indeed it would be the work of a moment to remove the check, if I so desired. This is a red herring brought up by User:Elen of the Roads, who is not familiar with stuff like C# (a state of blissful ignorance I sadly miss). Rich Farmbrough, 01:53, 18 April 2012 (UTC).

Yeah, enough people brought up AWB that I started to wonder if I was missing something (it sounds like I wasn't). The axe thing was just a way to show how a "solution" can focus on an almost-irrelevant aspect of the problem, not to compare you personally with an axe murderer.

Getting back to the issue, though: Rich, may I ask your intentions following this case? Are you going to continue the unilateral bot operations? What if anything would it take to get you to cut it out? Why do you find it so important to run them? I think we are agreed that desysopping won't make any difference either way, so my own preference is that you not be desysopped, as your general admin work is mostly good as far as I can tell. But from my past observation of bot disputes, this case feels destined to become known as "Rich Farmbrough 1" sooner or later, no matter what decision is reached in the short term, if you get my drift. You do enough good stuff that the community probably will never siteban you like it did Betacommand, so I fear a Giano-like infinite drama loop developing instead, a pretty crappy outcome. I wonder if you have any advice about how to avoid that, presuming that letting the bot editing continue is not on the menu.

As for C#, I've never used it myself, but some people who use both have told me it's noticably better than Java. I guess that's not saying much. 64.160.39.210 (talk) 07:59, 19 April 2012 (UTC)

I'm not sure my intentions following this case involve editing Wikipedia at all. There have been several attempts to provide a way forward. Xeno suggested a generic BRFA for uncontroversial fixes, such as he has, that was laughed out of court by teh haterz. I applied for a BRFA for editing templates specifically modelled on VeblenBot's BRFA (number 4 I think it was) - Veblen bot is CBM's bot - and that was turned down, partly after CBM opposed it, despite having the same BRFA approved himself. Various baggers have turned round BRFAs quickly, but this has not been sustained.

Some changes are not suitable for bot attention at this time (spelling fixes for example), and some are too small. People probably don't realise that I might well use AWB for 10 or 12 edits. Often I write the reg-exes and submit a BRFA, by the time it is approved for testing I ahve to re-write them.

I built an automated system to submit BRFAs and streamline the process, but I can only streamline my end. (I did leave an (approved) automated message for an unattended BRFA on a BAG members talk page, and got a very hostile reaction. Funnily enough that BAG member runs a bot to leave automated messages on bot operators pages, and thinks that's really cool. Another great <meh> of Wikipeida.) Most of BAG is inactive, and BRFAs have also become a locus of Fram attacks.

So what are the ways forward? Well I think that a lot of the FUD can be dispelled by giving clear undertakings. Keeping the drama llama away from the situation will help too. And if CBM can be convinced that he does not need to be the guardian of obsolete wikicode that will be the icing on the cake.

And then I would hope we could go back to people who understand that, while one is constantly trying to improve, largely errors are a numbers game. The only people who don't make errors don't make anything. Every time I get some time to initiate some new projects I get a flurry of talk page messages - naturally enough a mixture of praise, thanks, queries and complaints. And while I like to treat them all with respect, and in particular resolve any complaints and queries, it does become a bit of a drag when 90% of the complaints are either Fram, CBM or someone they have left messages for. With that gone, I think everything would be rosy. Sure there would still be complaints, but they would be almost all amenable to reasonable and swift resolution.

Rich Farmbrough, 12:23, 20 April 2012 (UTC).

Thanks for that response. I'm in and out of here today and will be away most of the weekend, but I'll try to digest it and look at some more of the diffs you posted in the other section (I've already looked at a few). Overall you seem to be pursuing a mission that I still don't understand. I may try to discuss this with you some more later, if that's ok with you, in the hope that better understanding this background can be useful in figuring out (on all sides) what to do next. 64.160.39.210 (talk) 09:08, 21 April 2012 (UTC)

I see not even the claim that Rich abused admin powers, much less evidence that he did. "But he does not always interpret all policies the same way I do" is a lousy argument. Oppose. --Stephan Schulz (talk) 09:22, 21 April 2012 (UTC)

Agree with everything Stephen said. Jenks24 (talk) 02:26, 22 April 2012 (UTC)

Not that I have even the faintest of hope that it will matter .. but I agree with Stephan Schulz as well. --Dirk Beetstra ^{T C} 05:43, 22 April 2012 (UTC)

@JClemens: EditCountitis, but well: Hersfold's blocklog is shorter (as is their contribution-list), but significant. --Dirk Beetstra ^{T C} 07:02, 23 April 2012 (UTC)

Not sure I would consider that significant. However, there are at least a couple of current admins with a similar or longer block log than RF: [20][21] Jafeluv (talk) 11:52, 23 April 2012 (UTC)

That was actually my point .. the length of the blocklog is not significant, what for, in what context, and seen in perspective is way more important. --Dirk Beetstra ^{T C} 12:21, 23 April 2012 (UTC)

Community sanctions confirmed

2) The Arbitration Committee acknowledges the editing restrictions regarding cosmetic changes to wikicode and mass creation of articles that Rich Farmbrough has been subject to, and confirms their validity.

Comment by Arbitrators:

More diffs please. I realize they're embedded above, but I would like to see this fleshed out so that the remedy contains evidence of its own accuracy. Jclemens (talk) 06:24, 23 April 2012 (UTC)

Comment by parties:

Some reference to the sanctions likely does need to be made to make it clear that the lack of additional sanctions here does not render those void, as the precise opposite was done in Betacommand 3, where the Committee made clear it was overriding all existing sanctions. Some users may view that as precedent; this remedy, in essence, allows the existing community sanctions to stand without directly making them sanctions from the Committee that would need to be appealed solely to the Committee. Hersfold ^(t/a/c) 03:47, 13 April 2012 (UTC)

See my comments elsewhere, on the fact that the primary so-called editing restriction completely fails to pass any reasonable test for community consensus. In particular only one !vote was in support of the motion. Rich Farmbrough, 01:08, 18 April 2012 (UTC).

Comment by others:

Opposed to these in all cases. Can you imagine someone trying to get ArbCom-endorsed restrictions repealed at ANI? Either leave the community sanctions alone, repeal them or take them on yourselves, so they can be appealed to ArbCom. This will just cause more problems than it solves. Jenks24 (talk) 02:42, 22 April 2012 (UTC)

Procedural oppose - let ArbCom try and see whether the sanctions by itself are all suitable or necessary, I oppose blanket confirmation. --Dirk Beetstra ^{T C} 07:43, 22 April 2012 (UTC)

Why not give the restrictions back to the community (excluding the editors active in this case), and see what the community thinks about the restrictions? --Dirk Beetstra ^{T C} 05:56, 25 April 2012 (UTC)

This may well be what I have to do. Xeno tried it but got nowhere, only two editors actually answering the question he raised. I am of Rich Farmbrough, 00:53, 27 April 2012 (UTC).

Proposed enforcement

Anti-proposal

No enforcement provisions

Comment by Arbitrators:

Comment by parties:

As I am recommending the Committee allow the existing sanctions to stand without assuming authority for them, and not proposing any further sanctions, I recommend that no explicit enforcement provision be enacted. Passing even the standard enforcement provision implies that the existing community sanctions may be enforced through Arbitration Enforcement; this should not be necessary, given that the current restrictions already include an enforcement provision of escalating blocks, which has already reached the one-month mark. Hersfold ^(t/a/c) 03:47, 13 April 2012 (UTC)

Comment by others:

Proposals by User:Rich Farmbrough

Proposed principles

Sanctions are preventative not punitive

1) Sanctions are preventative not punitive

In particular use of sanctions, simply because they are available, is bad practice.

Comment by Arbitrators:

Comment by parties:

Comment by others:

As a general principle this is fine. However, although some restrictions may appear to be punitive, because they are focused on a particular editor, they are actually preventative, because in the community's judgment they prevent continued disrtuption by that editor. Thus, for example, editing restrictions, community bans, and arbitration remedies do not become invalid solely because they might appear to be punitive. — Carl (CBM · talk) 12:18, 18 April 2012 (UTC)

Applying a sanction because of a technical breach is not preventative. It is damaging. It damages the community as well as the enyclopedia more directly. It has a chilling effect. It makes for division and conflict.

Applying a sanction because of an activity that is not ongoing, is punitive, and hence against Wikipedia ethos. For this reason, among others, threats should not be used. Rich Farmbrough, 00:09, 23 April 2012 (UTC).

Consensus is plural

2) Consensus in serious matters requires explicit support of more than one editor.

Comment by Arbitrators:

Comment by parties:

Comment by others:

Coöperation

3) Wikipedia works better as a coöperative effort than a combative arena

In particular Wikipedians should be wary of giving commands to their colleagues, of assuming bad faith, and of assuming that the other editor is wrong, before a full discussion has taken place. Wikipedians should be very wary of "piling on" to existing disputes without taking extremely careful steps to understand the situation. Wikipedians should be mindful that helping them with their queries may take a volunteer a significant amount of time and effort, and should reflect this in their intercourse with other members of the community.

Comment by Arbitrators:

Comment by parties:

Comment by others:

Proposed findings of fact

Fram vexatious

1) While Fram has reported real errors (as well as false alarms), his manner in doing so and ongoing attacks by AN ANI, ARBcom, CSD, AfD etc etc. render his interaction with Rich Farmbrough a net negative to the project. Moreover it is unlikely that anything positive will come out of interaction between these two editors.

Comment by Arbitrators:

Comment by parties:

A finding of fact without any evidence presented to back it up? Fram (talk) 07:14, 20 April 2012 (UTC)

Comment by others:

I don't have a real clear read on the situation between Rich and Fram. If there was anyone I thought was really trying to hassle Rich, it would be Roux, who is completely absent from this proceeding. 64.160.39.210 (talk) 22:19, 23 April 2012 (UTC)

You get several million points, no one else has noticed that. Seems to have sprung from this. Rich Farmbrough, 19:43, 24 April 2012 (UTC).

Edit restriction invalid

2) The editing restriction is invalid because no consensus was in place.

Comment by Arbitrators:

Comment by parties:

Fram claims the editing restriction is valid because some previous discussion started off against me.

CBM claims it is valid because I didn't oppose it or some such nonsense.

Elen just baldly states it, with no reasoning.

No one else advances any support for it being validly passed. Several people think it is ridiculous or not subject to consensus. Some of them have been cited or spoken up in this case. I have not searched for these opinions or asked for them. To me this seems crystal clear, and should the ArbCom wish to endorse this finding that might well prove valuable in reducing conflict and tension. However if they wish not to get involved in community decisions, then that is quite understandable.

Rich Farmbrough, 05:25, 25 April 2012 (UTC).

"Fram claims"? Please just provide a link to my comments, instead of making half-baked claims on my behalf. I claim that the editing restriction is valid because in the discussion about the editing restriction, many people supported a much stronger one, some supported an intermediate one, and some supported nothing but a wristslap or similar. The admin who installed the editing restriction choose a middle ground, a compromise. That's what dispute resolution and consensus-seeling is all about, no? Fram (talk) 06:52, 25 April 2012 (UTC)

As I said in a different, previous section a discussion started off against me. Big whoopi-doo. Rich Farmbrough, 05:15, 26 April 2012 (UTC).

How come we are discussing a "finding of fact" with a lot of claims (in the comments) but not a single diff or link? Not much "fact" there for the moment... Fram (talk) 06:52, 25 April 2012 (UTC)

It's all there in the Workshop or evidence. I'm sorry you either can't or won't see it. Rich Farmbrough, 05:15, 26 April 2012 (UTC).

Comment by others:

Unfortunately, the edit restriction was put in place, hence it is 'valid' (claiming otherwise could be a breach of WP:BUREAUCRACY). It does however show how narrowminded editors like Fram and CBM are: they enforce the rules because of the rules, not in spirit of WP:IAR, and without looking back whether a rule is appropriate or even necessary (oh, they will claim back, if someone breaks a rule, the rule is valid). However, ArbCom could ask the community to review the edit restrictions and their implementation while this case is ongoing, and take it from there. Maybe with the injunction that editors who have interacted with this case, who are parties to this case, or who are Arbitrators, are not to !vote or comment unless specifically asked for comments (and that those comments then are, if necessary, moderated towards neutrality). --Dirk Beetstra ^{T C} 05:39, 25 April 2012 (UTC)

Please give some examples of where I started AN or ANI discussion "because of the rules" and not because actual problems and errors resulted from the violations? And I have never claimed that "if someone breaks a rule, the rule is valid"; if someone causes a lot of problems while breaking a rule, then the rule is valid. Fram (talk) 06:52, 25 April 2012 (UTC)

No, that is again a wrong assumption. A rule is not valid because someone is making mistakes and/while breaking the rule (I already said that in my post of UTC 05:39, 25 April 2012 that that would be the defence). What I mean is, can you show me diffs where you actually questioned whether the rule was valid, suitable, useful, doing the job as it was intended? Maybe the edit restrictions are just to the detriment of the Wikipedia, and Rich Farmbrough (or Δ) does not make more errors then any other editor, but just because there are edit restrictions, there is a reason to bash such editors around for it and ignore all other problems that are there, and even, ignore all problems that come from not having Rich Farmbrough (or Δ) improving Wikipedia --Dirk Beetstra ^{T C} 08:05, 25 April 2012 (UTC)

Not so much a wrong assumption as a badly expressed idea, I think. A rule is valid on Wikipedia if it has community support, no matter whether it is useful, helpful, ... or not. A rule is useful if when followed or applied, it would make Wikipedia better (or would prevent it becoming worse). A rule is "effective" if it achieves this result. The rule in this case is IMO valid and useful, and partly effective, since it has reduced the number of incidents and problems. RF is making errors while violating the rule, and (at least in the case of the "no mass creation" rule) wouldn't be making these errors if he wasn't breaking the rule. The other restriction, against cosmetic edits, is less obviously effective, but is more of a gain-loss balance rule; the errors made while making cosmetic-only changes are not in themselves worse than errors made during more indepth, content-adding edits, but the net result of the edits is worse, since the gain from the cosmetic edits is so small (if any, the gain of changing e.g. "reflist" to "Reflist" is not really obvious). Coupled with the number of errors (and the number of times one error is made, due to the repetitive nature of many of the edits) and the issues regarding the handling of such errors afterwards, the problem with RF gets more serious than with many other editors. If we could restrict RF to make only edits that clearly, substantially improve Wikipedia, and to handle errors and problems better than has often been the case, then most of the problems being discussed here would be gone. Fram (talk) 08:31, 25 April 2012 (UTC)

Still, that edit restriction on cosmetic changes is held against him, and while I wholeheartedly agree that 'changing {{reflist}} to {{Reflist}} does not improve Wikipedia, when done as a part of a overall non-cosmetic-only edit, it does not degrade Wikipedia either. It does not degrade Wikipedia even it is part of a totally cosmetic edit, and while I also wholeheartedly agree that such cosmetic edits should not be made, I detest the fact that showing that Rich Farmbrough is occassionaly performing purely cosmetic edits (or even, presenting non-cosmetic edits which incorporate cosmetic changes as such) in breach of that edit restriction is used against him in stead of having the sanity to see that having that rule is, plainly, an obstruction to the general improvement of Wikipedia. And, in a way, that same goes for the mass-creation tasks, Wikipedia improves more when occasional mistakes (even on a mass scale) are made, then from total disallowing it. Restrictions of this type are degrading Wikipedia. --Dirk Beetstra ^{T C} 08:45, 25 April 2012 (UTC)

Apart from the fact that the vast majority of his mass creation tasks since the restrictions (and immediately prior to it) contained major errors or problems (like the sockpuppet categories, and the "Wikipedians who like X" ones) or were totally unnecssary (like creating hundreds of DNB templates, deleted after CfD)? Perhaps people need the "sanity" to see that the rule is necessary to improve Wikipedia, since the occasions were it was breached have not often been successful. Want another example, not yet included in this case anywhere AFAIK? Wikipedia:Miscellany for deletion/Portal:Trinidad and Tobago/On this day had RF creating lots of pages during an MfD for the group, only to get 70 of them deleted immediately at the close of that MfD again.

Is there any reason why he continues making these changes (like changing "reflist" to "Reflist")? Any reason at all? There are easy ways that he can significantly decrease the amount of criticism he gets, without Wikipedia losing out on anything, if he would stop making those edits that are not an improvement, are not guideline-prescribed, and are getting on the nerves of some people. People are lauding RF for making so many contributions, and feel sad that due to his block, he wasn't the first to reach 1 million edits. Instead of this editcountitis, perhaps people could focus more on the quality of work and the actual benefit edits present. The improvement to Wikipedia is a lot less impressive than the number of 1 million edits seems to indicate.— Preceding unsigned comment added by Fram (talk • contribs)

The non-existence of the Sockpuppet categories is utterly stupid, they should be created. But because R.F. is creating them in violation of an edit restriction, it is deemed a bad idea, it is not a bad idea because it is a bad idea (which it is not).

Is there any reason why he should not make the changes? Is there any reason you have to change {{see}} to {{further}}? Yes, just as much as that! And regarding quality, because 10 edits repair something, and one of them repairs it wrong, it are still 9 repairs. Instead of these flawed statistics, perhaps people could more focus on what is actually improved, and focus on improving further what has not been improved perfectly etc. --Dirk Beetstra ^{T C} 10:29, 25 April 2012 (UTC)

We'll have to agree to disagree on the esockpuppet categories then. I fail to see a reason to link two IP adresses to one another permanently because they may have been used by one editor years ago. I fail to see the reason why an editor in good standing should be linked to another editor through a category now, when an investigation years ago concluded that there was no evidence for a link between them. I fail to see a reason why an editor who made one edit while logged out, and afterwards signed that edit, should get a "suspected sockpuppet category" years later for that. At best, these categories no longer serve any purpose at all at the time of creation; at worst, they are wrong or extremely outdated. You have not given one reason what the benefit was of creating these years after the fact and indiscriminately. It would have been a bad idea, no matter who did it, it just happened to be RF who implemented this bad idea. Fram (talk) 10:55, 25 April 2012 (UTC)

Oh, so you do not have a problem with a user being wrongly tagged as a sockpuppet, or obsoletely tagged as a sockpuppet or whatever, you do not have a problem with a user being wrongly categorised as a sockpuppet of another editor, or obsoletely categorised as a sockpuppet of another editor or whatever, but you have a problem with the existence of the category itself, which already exists, whether or not there is a Wikipedia page at that position, because the category is populated. Hence, either it is utterly stupid that the category that is populated does not have a proper page, or it is utterly stupid that the editor is still categorised there, but it is certainly not a bad idea that an existing category gets a page. --Dirk Beetstra ^{T C} 12:38, 25 April 2012 (UTC)

In some cases, the problem probably was that someone had simply forgotten to create a needed sockpuppet category. In other cases, the actual problem was tagging the user invalidly as a sockpuppet in the first place. In the latter case it would certainly have been a better solution to simply remove the invalid tag instead of bluelinking the category. However, the claim that it was Rich Farmbrough who created the link between the tagged accounts does look unfounded to me. Redlinked categories function in the same way bluelinked ones do, which means the link between the users had existed long before RF created the category pages. Jafeluv (talk) 13:06, 25 April 2012 (UTC)

Calm down, please. Where have I indicated that I don't have a problem with any of these? That there were errors before doesn't mean that we have to double them. If an admin comes across incorrect or outdated tags, the solution is not to create the accompanying categories, but to remove the tags. Furthermore, you are totally wrong when you claim that the category "already exists"; a redlinked category doesn't exist, now they do. Please open Category:Suspected Wikipedia sockpuppets: see any redlinked categories hare? On the other hand, do you see things like Category:Suspected Wikipedia sockpuppets of 119.30.39.37, a page created by RF in this run, linking an IP address still in use to another IP address, based on a tag placed in 2009? Before RF created that category, it wouldn't have showed up in Category:Suspected Wikipedia sockpuppets, now it does. Contrary to what you say, it is a very poor idea to give a redlinked page an existence simply to remove redlinks. If you enter "Category:Suspected Wikipedia sockpuppets of 117.0.117.202" in the search box, you'll get a result. If you enter "Category:Suspected Wikipedia sockpuppets of 07757109672a" in it, you don't get a result.

When you enter "Category:Suspected Wikipedia sockpuppets of 117.0.117.202" in Google, you get Category:Suspected Wikipedia sockpuppets as the very first result. If you enter "Category:Suspected Wikipedia sockpuppets of 07757109672a" in it, you don't get any results. Both categories have one entry on Wikipedia, but the former has been created by RF, the latter hasn't been created. Now please explain to me how creating these makes no difference, and how these categories already existed before they were created? Please indicate how creating these categories didn't make the link between the people a lot more visible? And on the other hand, please indicate what "advantage", what improvement to Wikipedia, these creations made? Fram (talk) 14:09, 25 April 2012 (UTC)

This is paper thin. Even if that category hadn't been noindexed (which is a moments work - and again I note that slinging mud here is ssooo much more fun than taking a minute to resolve the problem, as Beetsra did) and even if someone is going to type that exact phrase into Google (more likely to type it into Wikipedia search) there would be no change to whether that item got a hit - it gets two hits on WP. Rich Farmbrough, 05:27, 26 April 2012 (UTC).

You don't need to use the "exact phrase", it works just as well with e.g. "sockpuppets 117.0.117.202" vs. "sockpuppets 07757109672a". And "there would be no change to whether that item got a hit"? I just demonstrated that it makes a difference both when searching inside Wikipedia, and when searching through a search engine like Google. 06:57, 26 April 2012 (UTC)

No, it still gets a hit, even discounting the cached one. And "Sockpuppets Rich Farmbrough" gets hits too - should I be suing for defamation? There are far worse things than that posted as a result of my work on Wikipedia, even on this page. Sheesh some people need to get real. I am of Rich Farmbrough, 01:02, 27 April 2012 (UTC).

I agree with Fram here, that searching for the term "sockpuppet 07757109672a" does NOT find the category that the editor is in, while searching for the term "sockpuppet 117.0.117.202" does find the category the IP is in. That is likely true for a handful of similar search-terms when using the Wikipedia built-in search facility. This is of interest, as this becomes an equilibrium between 'having the categories and hence being able to find them would unnecessarily taint the editors who are in those categories', vs. 'not having the categories, and hence not being able to find the categories would make it very difficult to actually find the sockpuppets when they are needed'.

The Google search for "sockpuppet 117.0.117.202" currently finds the holding category "Category:Suspected Wikipedia sockpuppets, but not the category itself, because the noindex has not taken hold yet, the Google search for "sockpuppet 07757109672a" does not find the holding category, nor the exact category, neither does the Google search "sockpuppet 90.195.131.12". That last result is of interest, as it shows that only the first results are found by Google, whether or not Rich Farmbrough created any categories, that would have been true for the first 200 that were there. In other words, if Rich would have created 15.000 categories for 15.000 username socks (and have excluded the IP-socks), none of the created categories would have appeared in the results (and in a couple of days, that is also true for any of the results that are currently obtained from Google. --Dirk Beetstra ^{T C} 09:07, 28 April 2012 (UTC)

And again, that is not a mistake by Rich Farmbrough. Obviously, Category:Suspected Wikipedia sockpuppets should be no-indexed. --Dirk Beetstra ^{T C} 14:12, 25 April 2012 (UTC)

I have {{noindex}}-ed the category, as it makes no sense that the subcategories can not be found on Google directly, but the master category can be found, unnecessarily tainting the sockpuppets, and making them visible to the outside world. Moreover, the many cats that are created (not in a bot run by Rich Farmbrough), and which do contain wrongly tagged names, are also indexed on Google. --Dirk Beetstra ^{T C} 14:16, 25 April 2012 (UTC)

Proposed remedies

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.

CBM injuncted

1) CBM injuncted to stop reverting "cosmetic" parts of others' edits

Comment by Arbitrators:

Comment by parties:

Edits such as this against the consensus that this template shall be called "{{About}}" are counter productive. Moreover they are incredibly hypocritical, and since CBM often rolls back a complex edit to restore one of his "fave" redirects, constitute de facto vandalism.

Comment by others:

The {{otheruses4}} template had a deletion discussion which closed as keep. If there was consenus to replace all uses of the template, a bot would already have done it. The general principle of WP:NOTBROKEN is that redirects are harmless. In this case the specific redirect name "otheruses4" says which message the template is supposed to be generating, unlike the more general {{about}} which can generate several different messages depending on how it is invoked. — Carl (CBM · talk) 21:26, 18 April 2012 (UTC)

The idea is to SAVE EDITS. That is the supreme irony here. We fix the small stuff as we go. And yet we get fluff implying that we are making "pointless edits" instead of having a bot fix one thing at a time we FIX THEM AS WE GO. We keep the edit counts DOWN, we keep the history SHORT, we keep the dumps SMALL. Occasionally Yobot, or historically SmackBot might make 2 or 3 per thousand "cosmetic only edits" but it is people like me, Rjwilmsli, ReedyBoy, Magliodites, OhConfucious, Kumioko and many others that have been WORKING TOGETHER over many years to produce, as well as millions of minor (but needed) fixes, tens or hundreds of millions of "cosmetic" fixes. Without our work there would be twice as many edits to the English Wikipedia by now. And yet at every turn we come across roadblocks in the form of editors who have "turf" they want to protect - mostly they are amenable to discussion, I fail to see why you aren't. Why do you think it is your duty to protect "otheruses8" for extinction? What greater cause is served other than showing those dang fixers who's boss? How does it help me, you or anyone? The answer is it doesn't. It just makes the whole operation look like a bunch of idiots who fight over trivia. That's why I generally don't comment when you decide to go and undo the work of others, because in the end the extinction probability of these outdated redirects is asymptotically close to one, and regardless of your actions they will be gone in five or ten years time (yes this is a long term cleanup). But when you decide to sabotage BRFAs, drop status updates on Elen's page, despite saying that you only look at my edits when they are on your watchlist, and finally bring all the same baggage here, I'm afraid you have to take responsibility for your obstructive habits. Rich Farmbrough, 02:42, 20 April 2012 (UTC).

CBM, there was a deletion discussion on the redirect, not the template (the template itself has been named {{about}} since 2009). A keep decision at RFD does not mean that all uses of the redirect must be preserved. Reverting articles to link to the redirect instead of using the template directly seems entirely pointless. Even more so when you revert other useful edits while doing it. Jafeluv (talk) 07:28, 20 April 2012 (UTC)

The removal of the date was an undetected edit conflict with AnomieBot; the edit I was responding to was [22], which I avoided completely undoing. The context is that on March 24, R.F. did an unapproved series of edits to about 30 pages all of which used otheruses4 (e.g. [23]). There is no chance that he accidentally edited that many articles in a row all of which use an uncommon template. He has done similar things before, performing unapproved tasks on his main account. If RF wants to replace all uses of the otheruses4, he should get consensus to do it, and I will abide by any such consensus. In fact I would even do the replacements with my own bot. But at the moment he is just him trying to enforce his personal preference about the template name, in violation of WP:NOTBROKEN. There is not a first-mover advantage in such things: if an edit is made without consensus, it can be undone, otherwise we would have a fait accompli. — Carl (CBM · talk)

His "personal preference about the template name"? That's the actual name of the template since 2009, no? (The move discussion is here.) Running a bot for the sole purpose of replacing all uses of otheruses4 would probably be pointless – however, reverting such replacements with no other changes made to the article is doubly pointless, and it seems you've been making such edits since at least July 2010 (diff, discussion). Jafeluv (talk) 12:00, 20 April 2012 (UTC)

It's actually worse than that, as well as these reverts and partial reverts to no end, he has also mass reverted, then run VeblenBot to reinstate the edits he reverted. Still mostly harmless - but in the context of his complaints against me, rather making clean hands a moot point. Rich Farmbrough, 04:35, 22 April 2012 (UTC).

The last time VeblenBot edited mainspace, it was in Dec 2011 to handle some offical stub deletion discussions (e.g. [24]). Before that, I seem to have done 6 test edits in 2009 related to now-banned user User:Erik9bot, and before that the last time VeblenBot edited mainspace was 2007 (bot contribs in mainspace). — Carl (CBM · talk) 19:45, 24 April 2012 (UTC)

Indeed, 28 December 2011 was your "revert Rich" day.

• Here I correct a dash and replace a general fix deprecated stub template, and fix the spacing between the stub and the rest of the article.
• Here CBM partially reverts with a "NOT BROKEN" edit summary - it is broken now, since the spacing is wrong.

Another case:

• Here I fix the category ordering and replace the stub template.
• Here CBM reversts with the same edit summary.
• Here VeblenBot reverts CBM.

VeblenBot then goes on to make hundreds, possibly thousands of similar edits - which CBM would have kicked up merry Hell about had I made them.

Maybe the issue here is that CBM can only see two types of fixes, ones which are worthy of an edit, and ones which are not worthy to be done at all. The general approach is that there is a class of edit which are worth doing only as part of a more substantial edit, ideally. A more nuanced approach still is probably best, but the triage approach is the standard model.

Rich Farmbrough, 01:07, 25 April 2012 (UTC).

Had you marked your edits as implementing the deletion discussion, via the edit summary, I would have looked it up and left them alone. The edit summary "small fixes", on the other hand, is not a reason to bypass a template, as that is not a "fix" (nor is rearranging categories). The only edits that VeblenBot made in the year 2011 were all clearly marked with a deletion discussion in the edit summary. It's still not clear why you were trying to close the deletion discussion with your own individual account instead of letting a bot do it. That is exactly the sort of thing that bots are intended to achieve, and it's another example of you running bot-scale jobs on your main account. In this case I explicitly sought consensus here before using VeblenBot on the task, just to be sure there was consensus to replace the template. — Carl (CBM · talk) 01:59, 25 April 2012 (UTC)

I was not trying to "close the deletion discussion" - and rearranging categories is a fix, though I see you are currently engaged in a dispute with another editor over that. There were maybe 10 or 20 such edits, calling that "bot-scale" is crazy. The template replacement was standard AWB, therefore legal even according to your interpretation of the rules. Had you just taken the sane approach that this is a minor edit you did not need to worry about, and walked on there would have been no problem.

And you say I should have used a bot, yet you opposed the very BRFA which would have allowed me to do such things.

I'm sorry but these arguments do not stack up, and are just more evidence of a desire to protect obsolescent wiki-source, which you need to find a way to get away from, ignore or loose. Rich Farmbrough, 06:30, 25 April 2012 (UTC).

There were over 3,000 pages in the job of closing that "stubs for deletion" request. If you did due diligence before starting to implement it (I assume you did) then you knew that already. If you were not trying to implement the deletion discussion, there was no reason to start replacing those templates in the first place. The AWB list has templates that may be replaced, but replacing them solely for the sake of doing so has never had consensus, because there has never been consensus that bypassing redirects is a "fix". Replacing all the rail-stub templates in particular, to implement the deletion discussion, did have consensus, but that's 3,000 edits, too many to just do them with a non-bot account. — Carl (CBM · talk) 10:50, 25 April 2012 (UTC)

Well I could not have done it with a bot since you would have (and did) sabotage that. All AWB bots and editors with GFs on would be replacing that template, and in every case that I did so there was a rendered change to the page. And to say doing part of the task has no consensus but doing at all does is just another absurdity. Rich Farmbrough, 23:49, 25 April 2012 (UTC).

Interaction

2) Rich Farmbrough and Fram not to interact.

Comment by Arbitrators:

Comment by parties:

See finding of fact above. It is time for Fram and I to say goodbye.

Which "finding of fact" would that be? Fram (talk) 07:18, 19 April 2012 (UTC)

Comment by others:

This would be interesting, though. If problems would be reported without Fram being allowed to report them, then we really know there is a problem. Should be combined with a 'Rich Farmbrough and CBM not to interact' restriction. I should have thought about this in the Δ-cases. --Dirk Beetstra ^{T C} 10:20, 24 April 2012 (UTC)

I see the benefit for Rich Farmbrough in this, but what is the benefit for Wikipedia? Less errors and problems being spotted and corrected? Fram (talk) 10:22, 24 April 2012 (UTC)

Thank you. --Dirk Beetstra ^{T C} 10:26, 24 April 2012 (UTC)

I don't believe I was involved at all in the recent discussions that led to the block. Nor was I involved in the creation of either of the edit restrictions. I don't believe that simply reporting violations is a problem - the problem is that the violations continue to happen. Blaming one of the messengers will not solve the problem. — Carl (CBM · talk) 16:59, 24 April 2012 (UTC)

So, Carl, we are talking here about an interaction ban between Fram and Rich Farmbrough, you have your own section below - maybe you are right, but that can be established there. --Dirk Beetstra ^{T C} 03:52, 25 April 2012 (UTC)

I created the complementary 'Rich Farmbrough and CBM not to interact proposal below. --Dirk Beetstra ^{T C} 10:31, 24 April 2012 (UTC)

Errors

3) Rich Farmbrough to fix errors
Rich Farmbrough to fix any errors brought to his attention by members of BAG. (To exclude CBM and Fram should they ever become members.)

Comment by Arbitrators:

Comment by parties:

It seems that this is what most of the parties think is the key problem. While I am frankly gobsmacked by the way people attempt to resolve disputes, there is really little actual meat at the heart of this, just some problematic editors spreading FUD, with their own agendas. So for clarity, while I would do this anyway, and I not just for BAG members, let this be part of the resolution of the problem. Note here that I am looking at proposals that will nail down the actual and claimed causes of the dispute. Of course I make errors, and of course I will fix them, while there is still breath in my body. But if we need to make it a Arbcom injunction to get the work back on the rails, then so be it. Rich Farmbrough, 04:33, 20 April 2012 (UTC).

Comment by others:

This doesn't seem so wise. CBM and Fram are very good at identifying errors, and they have good sense about when automation is appropriate. BAG is a group of bot enthusiasts that by its nature has a predilection for bot operations, so it approves a lot of them that the project doesn't need. Having a closed, self-selected cabal like BAG deciding this stuff is like putting a cabal of garage band members in charge of the encyclopedia-wide notability guidelines. I don't have any reason to think that Fram's or CBM's judgment about bot errors is worse than that of BAG. 64.160.39.210 (talk) 06:15, 20 April 2012 (UTC)

CBM's is worse, because he has no objectivity over items that he has personally latched on to. Look at this edit. This is sabotage pure and simple, driven by protection of obsolete redirects. Here he loses focus, reordering categories (sending stub cats to the end notably, but other changes in order too) is a perfectly valid rendering change to the article. On HEADBOMB's talk page he says of Magiolatitis edits "I would even do them with my own bot if they were approved". (This is the bot that has made, I believe, about a million needless edits, and which he is proposing to use to do the task which was sabotaged at BAG recently.) CBM is constantly tilting at windmills, making articles worse than they were, either by reverting perfectly good edits in whole, or in part. Furthermore those fixes, that were done with another fix, will now have to be re-done without that fix, resulting in three edits where one would do. And yet this editor complains about unnecessary change. It has got to the stage where the majority of CBM's edits are fighting other users. And while he is not obnoxious about it, like some, it is still disruptive editing, and needs to be stopped. Giving him carte blanche to make me fix what he calls errors would be totally ludicrous.

Fram's is worse because he takes several mutually contradictory positions at the same time. I am still willing to look at issues he brings up, but part of the idea here is that there would be an interaction ban. If Fram was unable to stop stalking my edits, despite an interaction ban, and found what he considered a "real problem" - despite years of banging on about the ER, he claims that he doesn't care about that - he would be able to raise it (politely) with a BAG member, who would give it some careful consideration, and then discuss it politely and calmly with me.

I haven't seen BAG members go off half-cocked, and, although of course they make errors the same as the rest of us, they are generally more willing to discuss something without getting defensive. Rich Farmbrough, 11:53, 20 April 2012 (UTC).

There has never been a consensus that these cosmetic changes need to be done, or even that they are improvements. If there was such a consensus - e.g., that a bot should go through and replace all these redirects, or that categories need to be ordered alphabetically - a bot would just do them all, and these other bots wouldn't have to worry about it. However, the general principle about these minor, optional things on Wikipedia is that they should be left they way they are unless there is a consensus for widespread change. Otherwise, we get a first mover advantage and the ability for bot operators to present a fait accompli by making too many changes for others to undo. If there is no consensus for a change, editors are free to reverse the edits per WP:BRD and usual editing rules.

This is sheer nonsense. If a widespread change is made as an "also fix" it is not a fait accompli unless you count five or ten years as fait accompli which is rather stretching it. And you are completely missing the point we are trying to avoid these incredibly wasteful bot runs you are so fond of, that is the whole purpose of bundling as many fixes into each edit as possible. Rich Farmbrough, 12:40, 20 April 2012 (UTC).

The problem is that these "also fix" are being done without community approval, and even without community notification that they are being done. Even the edit summaries fail to indicate what is happening. Thus, by the time anyone notices, thousands of articles have been changed, and the changed state appears to be the status quo, giving a false sense that the edits had widespread agreement. That is what I mean by alluding to fait accompli from the date delinking arbcom case. Making a change to one article is editorial discretion; making the same change to 1,000 articles is not, and requires some positive consensus before it's done. This is true for all sorts of optional stylistic edits; the general rule is to leave each article as it is unless there is positive consensus to make the change, for example in the MOS. When you edit 30 articles in a row, and every one of them has the same "also fix" in it [25], it is clear that the "also fix" is the main purpose of the edit, and the claimed "main purpose" is secondary. — Carl (CBM · talk) 12:54, 20 April 2012 (UTC)

The vast majority of bot operators act responsibly and never draw any complaints. For example I have never seen complaints about Cydebot, which has 3.5 million edits. But, as might unfortunately be expected, a few bot operators lack the necessary discretion and respect for community norms that are needed to operate a bot. In the case of R.F., this lead to an edit restriction, which lead to this arbitration case. This was not the result of one or two users alone, it was the result of a long-term pattern of complaints about edits by R.F. and his bots. — Carl (CBM · talk) 12:12, 20 April 2012 (UTC)

Really? Count then, the number of AN and ANI threads started by you and Fram about me, compared with all 1 million plus other editors of Wikipedia combined. Count, then the complaints about me left on Elen of the Roads page by you compared with anyone else. Count the irrelevant and hostile contributions to BRFAs by you and Fram compared with everyone else. Rich Farmbrough, 12:34, 20 April 2012 (UTC).

Rich, diffs? links? It's hard for people to judge whether I have left many (or any) irrelevant) contributions to BRFAs if you don't at least provide some examples. It's also hard to check your statement that I take "several mutually contradictory positions at the same time" when you don't indicate to people where and how I have done this. We can go down into a "I haven't", "You have", "I haven't" shouting contest, but that is not really how this is supposed to go. Please, please, either provide some evidence for your statements or don't make them at all. Fram (talk) 12:44, 20 April 2012 (UTC)

I see Fram and CBM made some edits in the HPB 46 request that Rich moved to a subpage.[26] Fram also made a number of other comments in Rich-related BRFA's that exhibit some acrimony, but I don't see obvious errors or distortions in them. 64.160.39.210 (talk) 04:08, 21 April 2012 (UTC)

Here's one comment from Fram from that BRFA

I'm glad you consider your misrepresentation worse than my supposed tag teaming. The block log of a bot is quite relevant when discussing requests for approval of new tasks for that bot. But the only reason I posted here is that you felt the need to make an unwarranted personal attack against another editor while ignoring the other admins that blocked the same bot (if you had renamed Smackbot instead of creating a new bot, this would have been more obvious, but for some reason you chose not to rename it)

First he starts with accusing me of misrepresentation. I.E. lying. That alone is incredibly bad behaviour for an admin - yes Fram is an admin, hard to believe though it is.

Then he imputes that I made an unwarranted personal attack, as an excuse for his previous post. He is actually familiar, as we have discussed it before, and some of it was caused by his AN-AN/I shenanigans, with the reasons for most of the data he presents, but he presents it as if it were all due to reprehensible behaviour.

Then he says "for some reason you chose not to rename it" - implying very clearly that I was hiding the block log. As anyone who looks at HPB's user page will see it is clearly linked to the old identity, and the old pages are redirected to HPB. And anyone who has been around Wikipedia will know that accounts with more than 50,000 edits are never renamed.

When I first met Fram I would have assumed that he didn't know this, and explained it to him. Now I know him better I realise that I would be wasting my time, because he would either not believe me, or make out still that I was somehow "hiding" something, or maybe he knew all along. It really doesn't matter which, pointing out his error would not change his opinion, if anything it would make him more adamant that I was a bad guy. He would say something like "Rich trots out the same old excuses".

For these reasons I have given up explaining stuff to him, and prefer avoid and ignore where possible. This does not seem unreasonable. Rich Farmbrough, 03:18, 22 April 2012 (UTC).

Rich, do you actually read the pages you use as evidence. You claim "First he starts with accusing me of misrepresentation. I.E. lying. That alone is incredibly bad behaviour for an admin - yes Fram is an admin, hard to believe though it is.". Have you read the line that I responded to? "Bringing it up when you two have been tag teaming me for such a long time discredits you (plural) not me, making such a misrepresentation is even worse. Rich Farmbrough, 13:02, 11 January 2012 (UTC)."(here) Yes, you accused me or misrepresentation, I only threw it back into your own face. If you feel that accusing someone of misrepresentation is "incredibly bad behaviour for an admin", then feel free to resign your admin bit.

Secondly, while you claim that "anyone who has been around Wikipedia will know that accounts with more than 50,000 edits are never renamed.", how do you explain that Kumioko, with a few 100,000 edits, was very recently renamed? "It really doesn't matter which, pointing out his error would not change his opinion, if anything it would make him more adamant that I was a bad guy." I'll let everybody decide who this applies more to in this discussion... Fram (talk) 09:31, 22 April 2012 (UTC)

Exactly, you throw things back in people's faces. Instead of addressing the issues.

You said "In the last 12 months alone, it was blocked by CBM, Ucucha, Angusmclellan, RD232, MSGJ, me, and Deacon of Pndapetzim." You were perfectly well aware that your block was the cause of two or three re-blocks, which were wrong since you had explicitly said I could unblock. So to represent these as independent condemnation is misrepresentation. And in this case it is extremely clear, whereas a lot is by implication, which is why you have got away with it for such a long time.

As to the limits on moves, you will have to ask MBianz who both did the move and changed the guidelines on 19 April. Rich Farmbrough, 23:54, 22 April 2012 (UTC).

@64. They can, unless subject to an interaction ban, still report errors to me, or to a BAG member. I simply don;t want to make myself hostage to their claims. If something has been considered and discussed with a BAG member, then, even though there have been a couple if instances of impatience, I am sure we will have no difficulty resolving any issues. The idea behind this is, funnily enough, to resolve matters and move forward.

My detractors claim that I don't fix my errors, I say that I do. This remedy means that if they are right the problem is solved, if I am right the problem is solved. Even if, as someone commented, the truth is somewhere in between, the problem is solved. Rich Farmbrough, 00:03, 23 April 2012 (UTC).

Rich, fixing most errors is trivial once they are found, and I'll accept that you're generally willing to fix them after other people find them. The error complaint as I see it isn't that you don't fix the errors, it's that you don't find the errors, because you don't check your edits closely enough. If you're making more edits than you're able to check, the solution is to make fewer edits. And the whole idea that only BAG can determine that something is an error is ridiculous. 64.160.39.210 (talk) 22:32, 23 April 2012 (UTC)

Yes, mysterious IP, you correct on almost all points.

1. Fixing errors is trivial - sorta - in an example I cited elsewhere I trawled through 160,000 edits to identify about 9 (minor) errors. This involved writing new code and several days run time, plus the actual fixing.
2. The complaint being that I don't "find" my own errors - possibly, but this seems to be a moveable feast.
3. There are no statistics for errors avoided, naturally, and if I find an error and fix it I don't always log that anywhere (although I could) - there are some bug reports for my bots on their user pages, and I have just built a rudimentary bug reproving system on my talk page. The assumption is that I don't find any/all/most of my own errors, and even if valid it is fairly pointless comment - why would I deliberately save an error? I can assure you, though, that there have been times when I have both found and fixed my own errors, sometimes trivially, sometimes with a great deal of sweat.
4. The idea is not that only BAG can determine errors, but that they are a bunch of reasonably clueful people who understand not only the technical side of Wikipedia and, I assume, the community side. More to the point they will, by and large, listen to what I have to say. I could throw in a couple of dozen other names, PlasticSpork, RD232, Haeb, Alison, Amalthea, Gerogewilliamherbert, Rjwilmsi, Siebrand, WerespielChequeres, Moonriddengirl, Magnus Manske, Fences & Windows, JDForrester, EdJohnston, EvilIPaddress etc... (apols for any spelling errors) But the point is not to create a worse restriction than the old one. I don't want some random editor turning up on my page and saying "ZOMG Mixed Martial Arts should be capitalised, it's the name of a sport!" (based on a true story) and then saying I should be sanctioned for not "fixing" his flawed understanding of the universe. As I said I will fix real errors brought to my by anyone, even as things stand. This proposal is not designed to make a difference, but to give comfort and assurance, and a binding guarantee. If no one wants it, fine, toss it out, it certainly doesn't do me any good. Rich Farmbrough, 23:02, 24 April 2012 (UTC).
   

Transparency

4) Rich to create bug-tracking system, so that the effort he does or does not put into identifying and resolving problems can be clearly seen.

Comment by Arbitrators:

Comment by parties:

 Done Rich Farmbrough, 15:40, 25 April 2012 (UTC).

Comment by others:

Proposed enforcement

Fixing errors

If Rich Farmbrough fails to make good faith attempts to reasonably fix systemic errors introduced by his editing then he may be restricted from continuing the task that caused the errors, or commencing any new tasks until the problem is resolved. Breaking this restriction would be considered blockable.

Comment by Arbitrators:

Comment by parties:

I was going to make this enforcement more swinging in the first instance, but considering the abuse of previous restrictions I think that a two stage enforcement is better. I am still concerned that, just as unreasonable time limits for talk page response have been used in the past, there is scope for abuse, deliberate or un-intended, but I am reluctant to make the wording more complex. Rich Farmbrough, 12:01, 20 April 2012 (UTC).

Comment by others:

It's not just the systemic errors that are a problem. The isolated, unsystemic errors are problems too, which is why all the edits have to be checked. The ANI threads are full of reports of those errors. 64.160.39.210 (talk) 22:52, 23 April 2012 (UTC)

I'm not sure that there are many complaints about one-off errors, but of course they should be fixed too. They are just <fix> and it's done though. I would think anyone taking one-off errors to ANI would get pretty short shrift. Or perhaps not. Rich Farmbrough, 23:08, 24 April 2012 (UTC).

Are you kidding? The one-off errors found by Fram and CBM (and maybe others) resulted in countless ANI's and are a substantial part of the case against you. They matter, and it's not up to you to decide that they don't. 67.119.3.53 (talk) 04:36, 25 April 2012 (UTC)

That's not my take on it. Perhaps we are talking at cross purposes here. For example Fram makes great play of the "Living people" error, I would call this a systematic error since it affected maybe 20-25 articles. Maybe this is what you mean by a one off error? If not I'm not sure what you mean. Rich Farmbrough, 04:56, 25 April 2012 (UTC).

Template

2) {text of proposed enforcement}

Comment by Arbitrators:

Comment by parties:

Comment by others:

Proposals by CBM

Proposed principles

Template

1) {Text of proposed principle}

Comment by Arbitrators:

Comment by parties:

Comment by others:

Template

2) {Text of proposed principle}

Comment by Arbitrators:

Comment by parties:

Comment by others:

Proposed findings of fact

Violation of bot policy

1) Rich Farmbrough has violated the bot policy. He has run large-scale tasks on his main account without approval [27]; has performed unauthorized edits as part of otherwise approved tasks; and has run approved tasks on his main account while his bot account was blocked [28].

Comment by Arbitrators:

Comment by parties:

This is a repeat of what Hersfold has said - repeating something three times before breakfast may allow you to believe it but it does not make it true. Specific accusations 2 are not against bot policy and one is specifically allowed by WP:COSMETICBOT. Rich Farmbrough, 15:44, 18 April 2012 (UTC).

Comment by others:

It is not prohibited to run large-scale tasks from an own account without approval, so there is nothing to violate. --Dirk Beetstra ^{T C} 10:42, 23 April 2012 (UTC)

I added two links to help clarify what is meant. Re Beetstra, the WP:BOTPOL says "Note that high-speed semi-automated processes may effectively be considered bots in some cases, even if performed by an account used by a human editor. If in doubt, check." The exception for manual edits is not intended to allow editors to bypass the usual consensus-gathering process for large-scale tasks. — Carl (CBM · talk) 12:01, 23 April 2012 (UTC)

I know Carl, but you know very well that editors do such large scale tasks, high speed tasks from their main accounts without getting that approval or pre-edit consensus (they base their actions on the editor'; s interpretation of existing policy or guideline). AWB and Twinkle do allow for high speed (for AWB limited to the speed of AWB and your 'save-button-click-finger-speed'), massive volume actions without waiting for consensus, without having approval, and the community does not have any problem with such (generally until they find one, or more errors in the process). --Dirk Beetstra ^{T C} 12:07, 23 April 2012 (UTC)

There are a few editors who do violate the bot policy without too many consequences, but in general the use of AWB or semi-auto tools does not on its own eliminate the need for prior consensus for large-scale tasks, as arbcom has made clear in the past. R.F. is in a small minority of AWB users who abuse the tool to avoid the need for community discussion of large-scale changes. The section WP:MEATBOT is also relevant. — Carl (CBM · talk) 13:50, 23 April 2012 (UTC)

They are not violating, the community is aware that those editors do that, and when there are no problems with the task, they are allowed to do it without consequences. That means that there is no violation when they are doing the large scale edits like that. --Dirk Beetstra ^{T C} 14:07, 23 April 2012 (UTC)

By the way, and again, You say: "Note that high-speed semi-automated processes may effectively be considered bots in some cases, even if performed by an account used by a human editor. If in doubt, check." - that does not make the fact that someone is doing a script assisted set of edits in a bot-like manner a 'violation' of the bot policy, the editor is very much allowed to do these edits. --Dirk Beetstra ^{T C} 12:11, 23 April 2012 (UTC)

In the end, arbcom is the one who would decide whether the policy was broken. I think that the point of the policy is clear, and R.F. had adequate warnings that he was pushing things too far. — Carl (CBM · talk) 13:50, 23 April 2012 (UTC)

I presume you mean '.. whether the policy was violated' - ArbCom is not writing policy. And well, that is why we have this workshop .... but again, they are not violating .. that sentence does not say that one is not allowed to do large scale edits, and an ArbCom ruling that way does not make it so, that is for the community to decide. --Dirk Beetstra ^{T C} 14:07, 23 April 2012 (UTC)

If everyone could simpy ignore the bot policy and run the tasks on their own accounts, claiming that the edits were manually made, the policy would be pretty much meaningless. The point of the bot policy is that approval is needed for large scale jobs, and this has been clear since at least the date delinking arbcom case. As with everything on Wikipedia, someone who violates the rule on rare occasions, and causes no problems, is unlikely to face criticism. But that is not the situation we're dealing with here. — Carl (CBM · talk) 14:49, 23 April 2012 (UTC)

Leaving aside the nasty "claiming":-

This is entirely my point in opposing bringing humans under control of BAG. There are many hundreds of large scale jobs done on WIkipedia - large scale jobs are often not even permissible to do by bot "Encyclopaedia Britannica" for example, is mis-spelled in a number of book titles we cite, and a number of quotes. While I know have enough specialised knowledge about that one phrase to probably be able to automate it entirely, that is only as a result of doing a very tedious replacement job. There was no requirement to submit that task for a BRFA.

(Incidentally this job was started as a side effect of Fram claiming that "Encylopedia" was a wholly unlikely misspelling and deleting a redirect I had created. I established that it was a very common misspelling. While he may have meant well deleting the redirect, it was simply wrong. If he did not mean well it was even more wrong.)

Suggesting that all large scale jobs be subject to bot approvals is foolish. How would one ever draw boundaries? I have made over 5000 edits to my talk page, to give an example. Is that a large scale job? How about when I corrected all the elements that were labelled "non-magnetic" that are actually paramagnetic? How about the people who built all the monarchy infoboxes? There are literally hundreds of editors doing these large scale systematic jobs, from a dozen up to hundreds of thousands of edits. Some are amenable to automation, some not so much. It really doesn't matter. The purpose of BOTPOL is, or ought to be, primarily to assure that automated edits meet certain criteria - and even these are based on outmoded models of Wikipedia and to placate misunderstandings and fear of technology.

Rich Farmbrough, 04:33, 24 April 2012 (UTC).

I didn't claim that Encylopedia as such was an unlikely misspelling, I claimed that "Template:Jewish Encylopedia" was an unlikely misspelling. You restored the template redirect, but afterwards another admin deleted it anyway, so it seems as if I wasn't the only one believing that creating a template redirect for that misspelling was overkill. Fram (talk) 08:28, 24 April 2012 (UTC)

The bot policy says, "Bot policy covers the operation of all bots and automated scripts used to provide automation of Wikipedia edits, whether completely automated, higher speed, or simply assisting human editors in their own work." It does not cover only automated edits. When the project was smaller, people could do more in the way of large scale jobs without discussion, but for at least the last few years that has not been the case, and we expect positive consensus in favor of all large-scale jobs, regardless of editing method. — Carl (CBM · talk) 11:40, 24 April 2012 (UTC)

Well, CBM, only a few editors are interpretating the Bot Policy as such. Admins can run such tasks on their own accounts, on thousands of pages, blatantly working with a fully automated script that is available to administrators through Twinkle, from a self-compiled list of pages, claiming that it is not a bot but an assisted script, and when confronted with that no-one cares because it is not in violation of the bot policy. Other editors do the same with AWB, or with self-written scripts. Yes, it would be better that they go through a BRFA, but, again, it is not a violation of the bot policy if they do not. It is utterly pathetic that you (and Fram) are using that argument as (an major part of) a finding of fact that Rich Farmbrough is at fault. --Dirk Beetstra ^{T C} 05:33, 24 April 2012 (UTC)

Twinkle has general community approval, and it is a reasonable application of IAR not to file a BRFA for each user who runs it. Many of the changes that R.F. makes do not have that sort of general community approval, and in many cases he never raises them for discussion at all, simply making them based on his personal preferences. — Carl (CBM · talk) 11:40, 24 April 2012 (UTC)

Even if that user is using Twinkle to edit 50 pages? Even if the actions that are then performed were performed before, widely discussed on at least 2 different forums, consensus was reached that it should not be done, they were undone, but since Twinkle can do it, now it is IAR and can be redone without discussion? Give me a break. If I repair manually 50 pages one typical typo, that is a bot-like action (and, that is likely uncontroversial), if I use built in Twinkle script that does a bot-like action on 50 pages, it is also by definition uncontroversial? But if R.F. uses AWB with additional scripts to do something on 100 pages, it MUST have a BRFA? --Dirk Beetstra ^{T C} 14:15, 24 April 2012 (UTC)

First AWB has general community approval too. Secondly you claim that they are based "on my personal preferences", this is simply false. They are based on sound logic, good sense, implicit and explicit consensus, and some of them are very much against my personal preferences. Simply because I choose not to contest some of the chaff you throw up, it does not mean that it is incorrect, as we saw with the rail stubs. Other items I have wasted time creating consensus, where it is immediately apparent to anyone who gives it a moments thought what should be done, but not to someone who only cares about rules for their own sake. Rich Farmbrough, 06:02, 26 April 2012 (UTC).

Template

2) {text of proposed finding of fact}

Comment by Arbitrators:

Comment by parties:

Comment by others:

Proposed remedies

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.

Large-scale tasks restricted

1) Rich Farmbrough is prohibited from performing large-scale tasks with his main account. In particular:

• He may not use his main account to edit more than 50 distinct pages, from any combination of namespaces, in any 24-hour period.
• He may not use his main account to create more than 5 pages, in any combination of namespaces, in any 24-hour period.

Comment by Arbitrators:

Comment by parties:

A stupid and vexatious limitation. For example I created stubs for many of the directors of MI5, in one day, see for example Vernon Kell. Preventing this would not be a good thing. In the evidence phase a problem with {{Cv-unsure}} was mentioned - it will require up to 400 talk pages editing and in some cases revdels. (I would be doing this today if I weren't blocked.) Am I supposed to stop and start on the 24 hour anniversary of each page edit? What happens if I fire up Huggle or Stwiki? Rich Farmbrough, 03:03, 20 April 2012 (UTC).

Comment by others:

Gets to the heart of the matter of R.F. evading bot policy by using his main account for unapproved large-scale tasks. He would still be free to run bot jobs after getting them approved by BAG. The motivation for the particular type of sanction is that it is very difficult to tell how an edit is being made (script, AWB, etc.) but the count of edits is objective, and limiting the edit count has essentially the same effect in terms of reducing automated editing. — Carl (CBM · talk) 13:55, 18 April 2012 (UTC)

And would you be on the BRFA saying "Please note that this editor..." and "For the record this editor...."? Rich Farmbrough, 02:56, 20 April 2012 (UTC).

Placing 400 cv-unsure templates on article talkpages with a bot or script (no matter who does it or how it is done) should obviously require some sort of prior consensus, maybe not formal BAG approval if the person doing it hadn't been in repeated bot drama. Rich could seek BAG approval for such an operation, or else someone else could place the templates, or non-BAG consensus could be good enough. Yes this would prevent Rich from going into the Huggle business. Lots (i.e. almost all) editors don't use Huggle and Rich can be one of them. One of Rich's better qualities as an admin is that he's (when he wants to be) good at individualized, non-scripted communication with users. Huggle if anything is counterproductive to that. I'd frankly be delighted if Huggle and everything like it were simply banned from the project, as it would help with this situation.

I could see exempting user talkpages from the page creation restriction, since there are various ANI incidents and the like that require human communication with multiple editors, some of whom are new and don't yet have talk pages. 5 is really a pretty low number. Or exempt any user talk edits that are obviously non-automated (because they contain a significant amount of original text). Also, if Rich creates a non-talk page anyplace, then the associated talk page shouldn't count against the limit. Rich's personal userspace should also be exempted. Other than these tweaks, the proposal seems on the right track and I've elsewhere supported implementing something like it it in MediaWiki so that it applies to all editors (not just Rich) unless they are flagged for high-volume editing. Added: mass creation of stubs (I don't know if there were enough MI5 directors to qualify as "mass") is not a good thing per WP:MASSCREATION. Better to just list them all in one place. 64.160.39.210 (talk) 06:35, 20 April 2012 (UTC)

• I don't intend to place new CV-unsure templates, merely to review those already in place. In most cases the copyvio has been identified subsequently and the problem resolved. These need changing to Cv-investigated, and possible revedeling. In other cases there is clearly no copyvio. In a small remainder concerns might remain with no evidence.
• I do use Huggle (and Stwiki), and using Huggle correctly is not easy. Huggle is however great for "Josh is ghey" vandalism, though the templates need working (and are being worked) on. And when I have used Huggle the other Hugglers have been great, extremely professional and proficient. I don't think that their use of Huggle can be related to new-editor retention problems. I think the culture of conflict might be, but I also think there are far more subtle issues which are not being addressed, as well as the great big ones that are. And the risk we face is that we put a massive amount of effort into areas that are not going to produce results, which is, to some extent the Wiki way - we have after all over written the vast majority of our content.
• BAG approval can be considered in terms of days at best and weeks or months at worst. Moreover it is not a good thing for ARB to issue instructions to BAG, even indirectly - this successfully delayed date-delinking for about a year. Requests, on the other hand might be considered favourably.

Rich Farmbrough, 12:56, 20 April 2012 (UTC).

Note that MASSCREATION was written substantially after those articles - which have all flourished - were created. And Xeno, the moving force behind MASSCREATION said it was to avoid "masses of sub-stubs" or words to that effect. It is an interesting case where a BAGGER created a piece of policy that puts BAG in charge of human editing. It is also interesting to see it being perverted to be used against substantial articles in other venues, when the intention was to control stubs. Back in the day we thought stubs were a good thing. This for example would now doubtless be speedied, and had the creator also made similar stubs for other nations they would be subject to all sorts of sanctions. We suffer from a form of "recentism" if we believe that Wikipedia is "almost done" and therefore we should oppose the creative drive of her editors. But that is, I think one thing we do see, the strait-jacketing with rules and instruction creep that leave potential new editors extremely nervous, and provide ammunition for those inclined to Wiki-war. Whether this is inevitable is hard to tell, but it is certainly something that we should fight to to delay and reduce as much as we can. Rich Farmbrough, 04:01, 22 April 2012 (UTC).

Apropos of new user retention User talk:Rich Farmbrough#Crushing a statistical shibboleth may be of interest to the IP. Rich Farmbrough, 03:50, 22 April 2012 (UTC).

Oppose - does not define 'large scale tasks', which can (and will!) then be interpreted at random. --Dirk Beetstra ^{T C} 05:45, 28 April 2012 (UTC)

Template

2) {text of proposed remedy}

Comment by Arbitrators:

Comment by parties:

Comment by others:

Proposed enforcement

Template

1) {text of proposed enforcement}

Comment by Arbitrators:

Comment by parties:

Comment by others:

Template

2) {text of proposed enforcement}

Comment by Arbitrators:

Comment by parties:

Comment by others:

Proposals by User:Fram

Proposed principles

Etiquette

1) Wikipedia's code of conduct is one of the five pillars of Wikipedia that all editors should adhere to. Wikipedia editors are expected to behave reasonably and calmly in their interactions with other users, to keep their cool when editing, and to avoid acting in a manner that brings the project into disrepute. Administrators are expected to adhere to this at a higher standard. Unseemly conduct—including, but not limited to, edit-warring, personal attacks, lack of respect for other editors, failure to work towards consensus, disrupting Wikipedia to make a point, offensive language (including rude, offensive, derogatory, and insulting terms—whether in English, a language other than English, or using invented terms), trolling, harassment, gaming the system, and failure to assume good faith are all inconsistent with Wikipedia etiquette. Editors should not respond to such behavior in kind; concerns regarding the actions of other users should be addressed in the appropriate forums.

Comment by Arbitrators:

Comment by parties:

This is taken straight from Wikipedia:Arbitration/Requests/Case/Civility enforcement#Principles. Fram (talk) 14:27, 20 April 2012 (UTC)

Comment by others:

Disruption

1) Continuing with the same behaviour that has led to restrictions and blocks is a form of Wikipedia:Disruptive editing. From that page: "Disruptive editing is not always intentional. An editor may be accidentally disruptive because they don't understand how to correctly edit, or because they lack the social skills or competence necessary to work collaboratively. The fact that the disruption occurs in good faith does not change the fact that it is harmful to Wikipedia."

Comment by Arbitrators:

Comment by parties:

Comment by others:

Proposed findings of fact

Etiquette violations by Rich Farmbrough

1) From the last 48 hours: "Endless fun for you, of course, and other wiki-lawyers. "[29], "Vandalism"[30], "incredibly hypocritical"[31], "trolling"[32], "his usual rudeness"[33], "is generally oppobrious"[34], "skewed mindset", "poisonous in the extreme", "incredibly stupid", "I doubt that Fram should be allowed on any volunteer project"[35], "some problematic editors spreading FUD, with their own agendas."[36], "teh haterz", "the drama llama"[37].

2) Repeated requests to support his allegations and statements with diffs or links[38][39][40][41] are largely ignored, making it impossible to discuss them in any meaningful way, and basically turn what could in some cases be decent arguments in his favour, or actual errors someone else made, into personal attacks. In general, and certainly in an ArbCom case, every reasonable request to provide supporting diffs or links for claims made should be provided, or the claims and statements withdrawn. (Note: there is no need that both parties agree that the diffs support the statements: this is something that every reader can judge for themselves; but without such supporting evidence, there is no way that impartial readers can judge, making the statements useless)

Comment by Arbitrators:

Comment by parties:

Comment by others:

Disruptive editing by Rich Farmbrough

1) Rich Farmbrough makes large-scale runs of automated or semi-automated tasks on his own account, but in many cases without adequately checking the results. This goes back some years (hence the 2010 problems and restriction), and continues until now. As an example, before this Arbcom case, he started creating redlinked but populated categories, a worthwhile task in itself which I had started doing as well a bit earlier. However, instead of careful and thoughtful deliberation of each individual redlinked category, he started creating them rapidly (about 100 categories every 8 minutes), taking them from five groups: "Category:Immediate children/", "Category:Immediate step-children/", "Category:Suspected Wikipedia sockpuppets of", "Category:Wikipedia sockpuppets of ", and "Category:Wikipedians who like"

Whether these categories were correct individually hasn't been checked, considering the speed of creation. Whether they were wanted as a group doesn't seem to have been considered as well. The use of creating a category in 2012 for e.g. two accounts who edited from August to October 2006(Category:Suspected Wikipedia sockpuppets of Bwikified), or to link a never blocked editor with an IP address that last edited in 2010(Category:Suspected Wikipedia sockpuppets of Cachemagic), or an editor who made a few edits in 2008 (and was never blocked) and his IP address at the time(Category:Suspected Wikipedia sockpuppets of Bvanvliet), or linking a user who made one edit as an IP, and then apparently noticed that he accidentally edited while locked out ad signed that comment([42]), back together two years later(Category:Suspected Wikipedia sockpuppets of 122.162.209.67) Or linking together two IP addresses(Category:Suspected Wikipedia sockpuppets of 119.30.39.37) based on a tag from 2009, even though with IP addresses you can hardly be certain that they still belong to the same user and so still have anything to do with one another three years later.

A cleanup of these tags, with a reasoned, individual approach, including in some cases the creation of these sockpuppet categories, would have been beneficial: the mindless rapid creation of these categories was not only a clear violation of his editing restriction, but also disruptive and a net negative.

2) When errors in his large scale automated or semi-automated edit runs are brought to his attention, he often only corrects the given examples, without checking whether there are more similar problems, or corrects them only haphazardly. The end result is that ifwe want to get rid of such systematic or recurring problems, other editors have to check his many edits thoroughly and either bring them all to his attention, or correct them all for him. The end result is disruption of the collaborative process and (of left unchecked) the creation of errors and problems, often within tasks of little benefit (MOS changes, replacing redirects with targets, and so on).

An example, again from the mass category creation: Rich Farmbrough created a few dozen categories from the "Category:Wikipedians who like" format. Three of them were individually listed in Wikipedia:Categories for discussion/Log/2012 March 28: Rich Farmbrough deleted those three as G7 on March 30. He then edited the template underlying the page where these categories came from, with the result that the many redlinked categories on that page no longer appeared[43]. This indicates that at the time of the deletion of these three cats, he was fully aware where they came from, and that there were dozens of categories from that same page, User:Lady Aleena/Media franchises, which he had also created. However, he made no attempts to delete them and went on to edit other things.

On April 17, a second bunch of these categories were nominated for deletion, Wikipedia:Categories for discussion/Log/2012 April 17#Wikipedians who like X. Only then, when someone else has made the effort of finding and nominating these, does he agree that they can be G7 deleted[44], afterwards only caring about his reputation apparently[45], but ignoring the fact that sadly even this mass nomination only concerns part of the problem, since it only is about the categories R-Z (User:Lady Aleena/Media franchises, R-Z), and not the other four groups, 1-B, C-E, F-H, and I-Q.

The end result is that a bunch of categories which shouldn't have been created in the first place, and otherwise should have been deleted after the first three CfD discussions, has now led to a fourth CfD discussion and still has the majority of these categories remaining.

Comment by Arbitrators:

Comment by parties:

" afterwards only caring about his reputation apparently" How can you say stuff like this and live with yourself? I know I have already given examples like this but every time I see one of these Framisms I am completely gobsmacked, not only that you would say such a thing, but that you would think it would advance your cause!

Talk page regulars will know that I will always G7 something I created if it is no longer needed, or is an error. In fact I always ask editors not to waste time with XfD but to talk to me first, since often either I can explain what something is for, or delete it, avoiding bureaucracy. Rich Farmbrough, 00:31, 25 April 2012 (UTC).

Comment by others:

Proposed remedies

Rejection of unsupported claims

1) All statemente and claims which have been challenged and have not been supported by evidence in the form of diffs or links will be removed from the case pages.

Comment by Arbitrators:

Comment by parties:

This should apply to all editors here, not just to Rich. Note that there doesn't have to be agreement that the evidence actually supports the statements, only that the original poster must present what he or she believes to be relevant evidence. Fram (talk) 14:27, 20 April 2012 (UTC)

Uh.. yeah right.. When attempts at oppression fail, reach for the censor. Rich Farmbrough, 01:59, 21 April 2012 (UTC).

Removing mistakes and removing personal attacks is hardly "censoring". It's not because this is ArbCom case that you suddenly are allowed to claim whatever you want, in whatever way you want to. This is an example of the mistakes you make in your posts and the extremely uncivil conclusions you then draw from your own mistakes. In this case, I could find out what you meant and where you were wrong; in many cases, I don't have a clue what you are even talking about, making it rather hard to provide evidence to the contrary or to explain what I did or said or meant, or even to admit that in such or such case I was wrong or overreacted or whatever. If your statements are left to stand but are disregarded by the ArbComs, or used as evidence against you, then fine by me, but the easiest way to get people to stop with this kind of unproductive page filling on already-long cases is to remove such statements. Fram (talk) 17:26, 21 April 2012 (UTC)

Comment by others:

Please do not do this. Those of us who wander into these proceedings every once in a while are expecting to see something like a court transcript, something assumed to be inviolate. Retroactively changing the proceedings means that impartial observers will be misled about what actually went on. True, the original proceedings might remain in the history, and if any redaction is done at the very end then this is readily viewed - but if so, then what is the point? Far better for whatever impartial party is supposed to do the redactions to instead add replies to the conversations, saying "the above allegation was unsubstantiated and should be disregarded", or something like that. Thanks. Wnt (talk) 18:52, 24 April 2012 (UTC)

Clerk Comment' I agree with Wnt. I do not have time to go through the case diff by diff to see what has been added and removed in each --Guerillero | My Talk 20:25, 24 April 2012 (UTC)

Fair enough. There have already been some things removed by people, when they realised they made mistakes, but that's about it for now. It's just rather frustrating that while people want a court transcript, they on the other hand don't apply other basic court rules. As far as I know, if people in court claim "I have evidence of X!", the judges will normally ask them to present such evidence or to withdraw their claim, no? Here, people can make almost any claim they like without needing to provide evidence. I sincerely hope that the ArbCom members will notice this and give it the weight it deserves, but I fear that when people encounter the same unsubstantiated claims over and over again, they will start to be influenced by it. Fram (talk) 07:00, 25 April 2012 (UTC)

Is that why you use the phrases like "cares only for his reputation"? If you believe that repetition of unfounded accusations has an effect on public perception, why haven't you stopped doing it? It can only be because you believe that. I am of Rich Farmbrough, 23:33, 26 April 2012 (UTC).

Civility enforcement

2) Rich Farmbrough is placed under civility parole for six months.

Comment by Arbitrators:

Comment by parties:

Suggested text and length may be changed according to precedent where needed. Fram (talk) 14:27, 20 April 2012 (UTC)

I have no idea what civility parole is, but if Fram will keep away from me then I am happy to be under it. Rich Farmbrough, 02:00, 21 April 2012 (UTC).

Wikipedia:Requests for arbitration/Betacommand 2#Community-imposed restrictions point 4 gives an idea of what I have in mind. Fram (talk) 17:30, 21 April 2012 (UTC)

Comment by others:

Aren't civility paroles, in essence, a license for others to freely poke and snipe at the subject without any fear that they'll snipe back? Such a thing seems like it would escalate the issues presented here instead of resolve them. The Betacommand restriction certainly seemed to have that effect. Given the subsequent acrimony and eventual reappearance at ArbCom I think it's safe to say the Betacommand restrictions are not a model we'd want to reuse. 28bytes (talk) 17:59, 21 April 2012 (UTC)

It is intended as an extra guarantee for people to know that they can mention errors and problems, or in general discuss things, without having to withstand the abuse and repeated severe WP:CIVIL violations we witness here from RF. It doesn't mean that other people can simply ignore WP:CIVIL when dealing with him, but that he has used up his quota of warnings, so to speak, and that further violations will result in blocks. IF you know of a better way to get this result, feel free to propose any alternatives. Fram (talk) 07:10, 23 April 2012 (UTC)

Agree with 28bytes here. This restriction simply never works, and is only counterproductive. --Dirk Beetstra ^{T C} 10:12, 24 April 2012 (UTC)

Meh, this is a pretty minor issue in the overall scope of the problem. 67.119.3.53 (talk) 04:45, 25 April 2012 (UTC)

Confirmation and continuation of the two existing editing restrictions

1) The two editing restrictions currently placed on Rich Farmbrough as noted in Wikipedia:Editing restrictions (one prohibiting cosmetic edits, one prohibiting mass creation of pages) are confirmed as being valid and being continued indefinitely.

Comment by Arbitrators:

Comment by parties:

Clearly they were imposed arbitrarily, the editor who imposed them said as much at the time. Moreover there was no actual consensus, only one supporting !vote, which seems very equivocal as of today. Ideally ArbCom should point out the lack of any reasonable support and instruct that they be struck. But they certainly should not be endorsed! Rich Farmbrough, 05:16, 25 April 2012 (UTC).

Apart from the fact that a much stronger restriction had much more support, the one actually implemented was in fact supported by Xeno (who proposed it), Dirk Beetstra, and Kingpin. And the restrictions were considered to be looking "reasonable" by yourself as well, in Wikipedia:Administrators' noticeboard/Incidents/Rich Farmbrough/October 2010#Suggestions from xeno and Kingpin. So can you stap repeating over and over again that there was only one support vote? Fram (talk) 07:12, 25 April 2012 (UTC)

Comment by others:

I thought I saw this also being proposed somewhere else. Again: procedural oppose, let the Arbitrators decide on the individual edit restrictions, or better, give them back to the community for reconsideration. Note: I am equivocal - my support was (and is) merely that purely cosmetic edits should not be performed. To quote myself further in the banningedit restriction decision: "I still refuse to see the capitalisation/whitespace change as an error, I do define it as 'useless'. And though I also think that of the fuss about it is equally useless: ..." (and that statement was made after my 'support' for the restrictions; [46], yet we are again here arguing about this utter nonsense) --Dirk Beetstra ^{T C} 05:37, 28 April 2012 (UTC)

Due diligence

2) Rich Farmbrough is expected to check for errors in his edits proactively, and to check and correct all similar edits he made for problems when a certain problem or error he nevertheless missed is brought to his attention.

Comment by Arbitrators:

Comment by parties:

This is to be expected from everyone, but needs special attention here. This is also separate from the two editing restrictions, because the same problems have happened in edit runs that didn't violate the restrictions as such, but still caused too many errors. Fram (talk) 09:45, 24 April 2012 (UTC)

So this is all very applehood and motherpie. But just because you like to think I don't check stuff, it doesn't mean it's true. Examples have been given, including checking 160,000 edits for a dozen (or less) minor problems. Rich Farmbrough, 05:12, 25 April 2012 (UTC).

Comment by others:

I expect that of everybody, and as for many other restrictions and edits, how would you check whether Rich did - will only come back to 'you made an error here, you obviously did not check this proactively'. The latter is already done by Rich. --Dirk Beetstra ^{T C} 05:42, 28 April 2012 (UTC)

Proposed enforcement

Template

1) {text of proposed enforcement}

Comment by Arbitrators:

Comment by parties:

Comment by others:

Template

2) {text of proposed enforcement}

Comment by Arbitrators:

Comment by parties:

Comment by others:

Proposals by Beetstra

Proposed principles

Template

1) {Text of proposed principle}

Comment by Arbitrators:

Comment by parties:

Comment by others:

Template

2) {Text of proposed principle}

Comment by Arbitrators:

Comment by parties:

Comment by others:

Proposed findings of fact

Template

1) {text of proposed finding of fact}

Comment by Arbitrators:

Comment by parties:

Comment by others:

Template

2) {text of proposed finding of fact}

Comment by Arbitrators:

Comment by parties:

Comment by others:

Proposed remedies

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.

Interaction 2

1) Rich Farmbrough and CBM not to interact.

Comment by Arbitrators:

Comment by parties:

I would not lose any sleep over this (unlike the amount I've lost over the arb case) were it enacted, but I would suggest limiting it to a six month period. Rich Farmbrough, 05:19, 25 April 2012 (UTC).

Comment by others:

If problems would be reported without CBM being allowed to report them, then we really know there is a problem. See also Wikipedia:Arbitration/Requests/Case/Rich_Farmbrough/Workshop#Interaction. --Dirk Beetstra ^{T C} 14:10, 24 April 2012 (UTC)

How would R.F. know which pages are on my watchlist, and how would he keep his bots from editing them? For ordinary editors, interaction bans are straightforward. But because R.F. and his bots edit more or less randomly, there is no practical way to avoid his edits. It's like putting a restraining order on a FedEx driver. My practice has been, repeatedly, to let things drop under the assumption that R.F. would fix the problems- only to see the same problem re-appear on my watchlist a little while later. — Carl (CBM · talk) 10:59, 24 April 2012 (UTC)

'not to interact', not 'not to edit the same pages'. But if you want to change it: 'You are not to confront Rich Farmbrough with erratic edits that he performs, leave it to others', or something along those lines. --Dirk Beetstra ^{T C} 11:05, 24 April 2012 (UTC)

The difficulty is that R.F. has a habit of re-doing the edits, even when (because they were reverted) he knows there is not consensus for them. As a concrete example, I once undid an out-of-process move of a fully protected template that he had performed, and he simply re-did it [47] (at which point I walked away in amazement). It seems like, unless some other restrictions on R.F. are passed simultaneously, this would serve as a sort of first-mover benefit in his favor, while the community in general avoids making rules that have a first-mover benefit. — Carl (CBM · talk) 11:13, 24 April 2012 (UTC)

So, that does not mean that you have to go after them again. Let someone else solve it - thát is what I meant. --Dirk Beetstra ^{T C} 14:09, 24 April 2012 (UTC)

Could you provide evidence of reports I have made that were not actually violations of the editing restrictions? Surely the solution to erratic edits is to stop them rather than to blame the messenger. — Carl (CBM · talk) 11:18, 24 April 2012 (UTC)

That is not the point, Carl. What I am seeing is that there seem to be certain editors who target other editors for 'their mistakes'. What my feeling is, is that there are actually not that many mistakes (with respect to the total number of edits), but just that you seem to keep disagreeing with edits that have been done. Also with to the edit restrictions in themselves (for Δ you were involved in the process of setting up the edit restrictions) - and I seriously wonder whether the having those edit restrictions written as they were was correct, whether the implementation was correct.

I wonder, really, that if you and Fram are taken out of this equation, whether the general public still has the same problem with the mistakes (which, certainly, will still occur - as with ANY editor on Wikipedia), but that all of that will be solved in less dramatic ways, and that that will actually benefit Wikipedia. I am not saying that you are wrong, I am just saying that because you and Fram are two of the editors who keep going after the mistakes certain editors make, that that frustrates the whole situation, while that situation, maybe, could be solved in other ways. --Dirk Beetstra ^{T C} 14:09, 24 April 2012 (UTC)

The effect will be that errors and problems remain on Wikipedia for even longer. When I pointed out that many of his DNB creations (biographies written in 1900) placed the subjects in the "category:Living people" (and not a single other mainspace category in most cases), a number of these had stood for two months already, without anyone noticing this or the many other problems with these pages (or at least without anyone correcting it or noting it on Rich's talk page). A number of other errors I found were also weeks old. Wikipedia:Articles for deletion/The Queen's Award for Enterprise: International Trade (Export) (1966) was a mass nomination of pages that were auto-created and wrong. No one contested them for three months. After a year and a half, we still are stuck with things like The Queen's Award for Enterprise: International Trade (Export) (1976), with no content and a self-invented title. The reality is that we need to have more people checking his edits, not less (assuming that nothing else changes of course). We could start by not making his creations auto-reviewed, so that at least new page patrollers will look at them, but that won't help with the many edits to existing pages of course. Fram (talk) 14:32, 24 April 2012 (UTC)

Wrong, so wrong. You had many hours work deleted for "wrong title" because there was no published clear distinction between the three categories before a certain year (although every single award says explicitly "For export" or "For technical achievement" or "Combined award") - these should clearly have been merged or left as they were. For the later articles you could find no such excuse, or you would have tried to delete them too, as indeed you had already scattered prods everywhere. This was a really unpleasant act, and is a supreme example of why you appear to be vindictive, repeatedly trying to get work destroyed for your own satisfaction. Rich Farmbrough, 00:02, 25 April 2012 (UTC).

As I noted in the AfD already, you only started coing "many hours work" on these articles once they were at AfD (months after their creation, and after they had been prodded and deprodded) and clearly headed for deletion, and when you should have been well aware that they were incorrectly named and split. You could have done those "hours of work" on correct articles, but choose to spend it on doomed ones. I invite everyone to read that AfD and draw their own conclusions. Fram (talk) 07:21, 25 April 2012 (UTC)

Re Beetstra: the difference of viewpoint here is that the cosmetic changes that you call "mistakes" are not accidents, they are intentional choices by R.F. to violate his editing restriction. They are not infrequent - he has programmed them into his scripts and into Helpful Pixie Bot. It would be very easy for him to remove that code, at which point the problems would stop. Alternatively, he could get positive consensus in favor of his changes, at which point nobody could complain about them. What you seem to be suggesting is that, even though R.F. is specifically forbidden from performing these edits, he should be permitted to perform them anyway. That seems strange to me. — Carl (CBM · talk) 15:27, 24 April 2012 (UTC)

That is fine, CBM, but it does not need to be you that reports them to AN/ANI/R.F./whereever. Let others pick up that there is a significant problem. --Dirk Beetstra ^{T C} 08:07, 25 April 2012 (UTC)

Oh, and please note, "What you seem to be suggesting is that, even though R.F. is specifically forbidden from performing these edits, he should be permitted to perform them anyway". Could you please indicate where I write that between the lines? --Dirk Beetstra ^{T C} 08:15, 25 April 2012 (UTC)

It is utterly silly to propose an interaction ban between a bot operator and a user reporting issues with the bot because the bot owner simply doesn't like them. If the bot owner cannot deal with user reports of problems, then he simply is not fit to run bots per WP:BOTCOMM. Especially when considering CBM's reports are both legitimate and backed by policy. Headbomb {talk / contribs / physics / books} 15:40, 24 April 2012 (UTC)

"It is utterly silly to propose ..." Beetstra is proposing no such thing. Beetstra is drawing a distinction between the type of complaint CBM typically makes which Elen of the Roads describes as "niggly" and Fram says he is not bothered about, and what we are calling "real errors". Beetstra argues, reasonably, that should CBM become silent on these matters and no one else take them up, we might assume that they are problems only for CBM. This is the sort of experiment that CBM might well try himself, being basically sound.

"Especially when considering CBM's reports are both legitimate and backed by policy." A lot of them are neither - not to say they are bad faith. Where we have had the conversation consensus has often supported that which he complained of. Moreover I have made it quite clear that I can deal with CBM's reports, even when he makes them in miscellaneous places without informing me. Rich Farmbrough, 00:02, 25 April 2012 (UTC).

I was not a participant in either discussion that led to your edit restrictions, nor in the recent discussion that led to the month-long block, nor did I comment at the request for this arbcom case. The latter two I intentionally avoided. Of course I have pointed out your policy and edit restriction violations, but I am far from the only person to have done so. It seems to me that might be trying to avoid taking responsibility for your edits by shifting blame somehow to editors who point them out. If the edits were indeed violations of policy or your edit restrictions, there's nothing wrong about pointing them out. — Carl (CBM · talk) 02:07, 25 April 2012 (UTC)

Fine, CBM, but you are part in a significant number of those reports. And for Δ, you were also regularly part in them (and at least there, a significant factor in the set-up of the edit restrictions - I hope you recall that we there also discussed whether the edit restrictions themselves were properly crafted, properly set up, etc. etc. Not that the Arbitration Committee picked anything up on that, but it starts to be a recurring thing, you are involved in the restrictions). IIRC somewhere was claimed that you and Fram are, by far, the most often reporters of the perceived problems to Rich (I'll leave it to R.F. if there are more editors who have been overly active there, to a point where it becomes a constant factor). If for Δ cases of years ago, and single violations of one of the bans a long time ago are bad enough to get them into findings of facts again, then every case where you feel the necessity to interact with the bot operator on their errors and problems is counting. If it is really such a problem what is going on, then the community without you is also capable of handling it. I'm sure you can find other things to do on Wikipedia, having an edit restriction on you does not end your career on Wikipedia, CBM. --Dirk Beetstra ^{T C} 04:06, 25 April 2012 (UTC)

There is not necessarily value in pointing out technical breeches. In fact doing so makes you less credible in the eyes of others, than if you reserve your pointing out for things that are real problems. Moreover pointing out or reverting things that aren't breeches at all puts your credibility on the line further. There are significant things that need doing, microscopic examination of my edit history or creating a preserve for endangered template redirects are not two of them. Rich Farmbrough, 06:40, 25 April 2012 (UTC).

There are indeed real problems - you have been been given edit restrictions and blocked because of them, and now you are at an arbcom case because of them. The edits themselves may be somewhat trivial, but they demonstrate a lack of respect for the community which has been asking you to stop for years. — Carl (CBM · talk) 10:24, 25 April 2012 (UTC)

Well, the consensus on the implementation of one of the edit restriction by the community was .. overwhelming. I can really see that the community there was asking for Rich to stop 'for years'. --Dirk Beetstra ^{T C} 10:38, 25 April 2012 (UTC)

Template

2) {text of proposed remedy}

Comment by Arbitrators:

Comment by parties:

Comment by others:

Proposed enforcement

Template

1) {text of proposed enforcement}

Comment by Arbitrators:

Comment by parties:

Comment by others:

Template

2) {text of proposed enforcement}

Comment by Arbitrators:

Comment by parties:

Comment by others:

Proposals by {User}

Proposed principles

Template

1) {Text of proposed principle}

Comment by Arbitrators:

Comment by parties:

Comment by others:

Template

2) {Text of proposed principle}

Comment by Arbitrators:

Comment by parties:

Comment by others:

Proposed findings of fact

Template

1) {text of proposed finding of fact}

Comment by Arbitrators:

Comment by parties:

Comment by others:

Template

2) {text of proposed finding of fact}

Comment by Arbitrators:

Comment by parties:

Comment by others:

Proposed remedies

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

1) {text of proposed remedy}

Comment by Arbitrators:

Comment by parties:

Comment by others:

Template

2) {text of proposed remedy}

Comment by Arbitrators:

Comment by parties:

Comment by others:

Proposed enforcement

Template

1) {text of proposed enforcement}

Comment by Arbitrators:

Comment by parties:

Comment by others:

Template

2) {text of proposed enforcement}

Comment by Arbitrators:

Comment by parties:

Comment by others:

Analysis of evidence

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

Personal attack by Hersfold

"... acts with disdain towards ..." I have never acted with disdain towards anyone, on Wikipedia or off. I may have become irritable after years of low-level sniping by certain editors, that is completely different. This is a really nasty thing to say about someone in order to further an attempted de-sysopping. I am deeply hurt by this assertion more than anything else. Rich Farmbrough, 01:17, 18 April 2012 (UTC).

Comment by Arbitrators:

Comment by parties:

Comment by others:

'Disdain' is a lot milder than a lot of other personal attacks circling around Wikipedia about other editors involved in disputes. In his assertion, he's merely dramatizing the community's lost of respect for you. Regards, Whenaxis (contribs) ^{DR goes to Wikimania!} 01:32, 18 April 2012 (UTC)

Drama is, of course, 90% of the problem. Rich Farmbrough, 01:55, 18 April 2012 (UTC).

Alas, it is. Whenaxis (contribs) ^{DR goes to Wikimania!} 20:33, 19 April 2012 (UTC)

Abuse of private email by Hersfold

"... that those opposing his actions are making similar changes yet remain unsanctioned..."

This is a misreading of a comment in a private email that I sent to Elen of Roads, requesting an unblock. Elen asked permission to share the email with Hersfold, to which I replied, that she could but it would only make him more angry (this case is a sad testament that I was correct).

This was never meant as a tu quoque defence, it was merely a remark about a strange phenomenon that is part of the whole matrix within which this affaire is set. Notably, for example, Fram has made thousands of edits, one assumes manually, emulating what SmackBot used to do. (There are many other examples of Fram doing something that I have done, or very similar to it.) Of course I don't think Fram should be sanctioned for doing these things, they are far more useful than threatening, or carrying out useless block actions or malicious seeming deletions and reversions. Rich Farmbrough, 01:27, 18 April 2012 (UTC).

Comment by Arbitrators:

Comment by parties:

Finally, an explanation of what you meant by that accusation. I recall I asked for such an explanation two weeks ago, but never got a response.

Now that you have clarified things, I do apologize. I've taken a brief look through the numerous discussions I've referenced in Evidence and can't find any comment made by you that makes such an assertion. Rereading the email, I do have to conclude that it came from my reading of that email. It was not my intention to refer to an email sent to me in confidence, and when I'd posted the request for this case several days later I'd forgotten that was the source. I'd remove it from my original statement, however at this point that likely will do little good, and at no point did I directly quote the contents of the email, merely paraphrased at the quoted point.

To address Whenaxis's point below, the policy regarding the Committee's handling of correspondence does not apply here. Elen and myself were both involved in this matter as administrators, not as arbitrators; it is for that reason we are both recused on and participants in this case. At no point (until now) was this email sent to any of the Committee's mailing lists. However, again, I do make an effort to keep private emails private.

I say "until now" as now that that details of this accusation have been made clear, and I concur that it did in fact form part of the foundation for my arguments in the original statement, I would like to send a copy of the original email I received to the arbcom-en-b mailing list for the Committee to consider as private evidence. I'll await for Rich's consent before doing so, however, unless directed to do so by a non-recused arbitrator. Hersfold ^(t/a/c) 02:09, 18 April 2012 (UTC)

Since the mail contains references to matters discussed with me in confidence, albeit obliquely, I must decline, though with some regret, since it would, perhaps, put details more clearly before the committee. Rich Farmbrough, 04:32, 18 April 2012 (UTC).

Comment by others:

According to the main landing page for the Arbitration Committee (see here) under the 'Communications and privacy' heading: "Arbitrators usually seek to treat your communications, including emails, as private when possible. We however cannot guarantee against public disclosure for a number of reasons, including potential security limitations. Accordingly, you should not disclose sensitive personal information in your communications with us. Once received, your communications may be shared with committee members and, in some limited cases, with third parties to assist in resolving issues or other purposes." Whenaxis (contribs) ^{DR goes to Wikimania!} 01:30, 18 April 2012 (UTC)

Hersfold is not acting as an arbitrator in this case. Rich Farmbrough, 01:44, 18 April 2012 (UTC).

Analysis of Whenaxis evidence

Note

Whenaxis has retracted his evidence, a bold and righteous move for which he should be commended. I invite those other parties who have had major flaws shown in their evidence to do the same, but with little hope of compliance.

Bot Approvals Group (BAG) flaws

We all (including BAG) know that there are flaws with BAG. However, at the moment, the composition and activity level of BAG is one of the best we have had for some years, and BAG is responsive and does its job well.

In his first bullet Whenaxis confuses a bot and a task.

Wheel warring over blocks

Bullets 2-6: The block placed by Fram explicitly allowed me to remove the block when the task was stopped.[48] This is not uncommon with bot blocks.

Rich reacts inappropriately and breaches bot policy

• In the diff cited by Whenaxis I move ad hominem attempts by User:CBM, supported by User:Fram to a sub page. If there were doubt that these are ad hominem, it can be seen that the user attacking the BRFA has had an word-for-word identical BRFA approved. The same diff also shows examples of using the block log as a badge of shame, by the person who was the proximate cause of the block.
• In Madman's closing of the BRFA he is probably correct in saying, ab initio that I have been uncivil, however in the face of an unremitting 2-3 year campaign of obstruction, I was the model of discretion.

Rich Farmbrough, 01:46, 18 April 2012 (UTC).

Comment by Arbitrators:

Comment by parties:

The "cause of the block" is the one making the errors, not the one reporting them or the one blocking. Continued failure to take responsability for your actions. Fram (talk) 09:30, 18 April 2012 (UTC)

Read up on causality a bit. The fact remains that no one raised the issue on my talk page, and only you raised it on AN. And the fact also remains that you have repeatedly, repeatedly made reports to AN and AN/I often without any discussion. There are sufficient Admins that you occasionally succeed in getting some punitive action, and usually enough people that you can spread some more muck. Rich Farmbrough, 18:22, 18 April 2012 (UTC).

As far as I recall, I made reports to AN without prior discussion with you twice, the one that lead to your one-month block and this ArbCom case, and this one in January, where you eventually blocked yourself for 8 hours. In all other cases, I first discussed it at your talk page, and only took it to AN when you didn't respond or when the problems continued, and when there was an actual problem, not things like your continued changes of "reflist" to "Reflist" and similar useless and guideline-prohibited but not actually "harmful" edits. In a few cases, I was wrong and you were right, and I acknowledged this, e.g. in User talk:Rich Farmbrough/Archive/2012Mar#WP:ANI discussion about ISBNs and Helpful Pixie Bot. In most cases, the people at AN agreed that there was an actual problem with your edits, sometimes resulting in restrictions and/or blocks. Yoru opinion of how easy it is to dupe admins into blocking you is noted... Fram (talk) 07:33, 19 April 2012 (UTC)

Attempting to turn this around to make at me attacking generic admins is pathetic. This is not merely my opinion, nor is it a question wholly of duping. It is obvious that with a sufficiently large audience some of them will believe even the most unlikely tale. If they are inclined to block then that is what John Vandenberg called "First mover advantage" - and drew attention to quite rightly as a systemic flaw. Rich Farmbrough, 03:13, 20 April 2012 (UTC).

Comment by others:

I do not concur that my comments in the bot request were ad hominem. However, when BAG reviews a bot request, they consider both the technical aspects of the task and the suitability of the bot operator to conduct it. In the end BAG makes the final decision, but all Wikipedia users are encouraged to comment and discuss both aspects on the BRFA page. The history of R.F.'s bots, and his editing restrictions, may call into question whether he is the most suitable operator for particular tasks. — Carl (CBM · talk) 12:52, 18 April 2012 (UTC)

Ad hominem means against the person. Splattering your opinion that I was unfit to run a bot across the page is an ad hominem argument - though you may attempt or not to justify it, there is no doubt that that is what it is. Moreover BAG are all quite familiar with my work, and I am sure my failings, they do not need an involved editor trolling on the page to tell them of that. I am amazed that you even thought that was acceptable behaviour. Rich Farmbrough, 18:22, 18 April 2012 (UTC).

Rich, I removed the BAG flaws part of my evidence because of request by Headbomb on the evidence talk page. Whenaxis (contribs) ^{DR goes to Wikimania!} 20:44, 18 April 2012 (UTC)

OK, well it would be great if you removed the rest of it. Not really it does no harm, although I understand how you reached your conclusions, and note too that Hersfold reached the same conclusions from the same evidence. Checking my talk page archive for the time in question, I think, would have disabused you both. However it may surprise you to learn that some admins do not even leave a message when they block. Rich Farmbrough, 04:41, 20 April 2012 (UTC).

Re "BAG are all aware" - that is like saying that the person closing an RFA will be aware of the candidate, so there is no need for anyone to vote. If objections are not raised, BAG may well point to the lack of objections, and avoid inserting their personal opinion if it is not raised in the discussion. So it is necessary, at times, to point out in a BRFA if there are concerns with the operator's suitability. — Carl (CBM · talk) 11:23, 24 April 2012 (UTC)

So you were making an ad hominem argument. And, explicitly, the purpose of that was to stop me from running a bot task. Going forward do you intend to similarly sabotage BRFAs with attacks on me, or do you intend to restrict your comments to the actual task in hand? Rich Farmbrough, 15:14, 27 April 2012 (UTC).

The problem I have hear is one that I have mentioned numorous times over the years about bot policy in general. It is so ambiguously worded that the interpretation of many of the rules (particularly the number of articles edited threshold) can be tailored case by case and only enforced when we feel like it. If we like the edit or the person BAG and others are more than happy to look away and pretend its approved or they don't see it. This is true of edits done by Twinkle, AWB, HOTCAT and a host of others. IF an editor does an edit that don't like then they start waiving the bot rules saying they have been violated. This isn't how we should be doing business. It should be clearly worded and evenly practiced.

For example, Under Bot like editing the rules say that without relation to bot, script or human editing if the action is disruptive and someone mentions it then the editor has to stop. This one is used all the time but they conveniently leave out the next sentance. "Note that merely editing quickly, particularly for a short time, is not by itself disruptive." IMO, just filling a users watchlist is not disruptive.

Another commonn one thats thrown in people face a lot falls under the volume of edits: "While no specific definition of "large-scale" was decided, a suggestion of "anything more than 25 or 50" was not opposed." This actually falls under Mass creation of content. Now this has been argued by CBM and a couple of others as being for any editing including AWB but IMO it was being used out of context. It would also seem to indicate that the creation of Categories and talk pages (tagging for WikiProjects primarily) would be included in this but most users aren't required to get BAG approval for these, until they irritate someone that knows about this rule then it becomes an issue.

Now these rules only partially affect RF due to the volume and types of edits but I have seen the inferred here so I wanted to clear them up

My last point is, depending on how these rules are interpreted and if arbcom decides to enforce them, is BAG ready and willing to take on the responsibility of approving the onslought of BRFA requests they are going to get? Kumioko (talk) 14:24, 24 April 2012 (UTC)

Exactly, Kumioko. And that is exactly why the many reports issued by Fram and CBM about Rich Farmbrough (and earlier Δ) is a form of 'wikistalking'. BECAUSE they see an edit by Rich Farmbrough (or earlier, Δ) on their watchlist, they would specifically check that edit. If anyone else, a trusted editor or maybe even an unknown editor, is doing a spree of automated edits, it gets ignored, even if that edit is a blatant mistake that violates not only the bot policy, but is also not supported, in any form, by other policies. That is why I elsewhere suggest, that maybe CBM and Fram should (consider to) take a break from checking other editors' automated edits, and for now, especially Rich Farmbroughs. Let the community solve it, but prohibit that the majority of the negative reports are coming from one small subset of the community. --Dirk Beetstra ^{T C} 05:12, 25 April 2012 (UTC)

Where are the "many" reports I issued about Delta? Apart from that: because many blatant mistakes are not picked up by the community already, you suggest that the ones by Rich Farmbrough should get the same treatment? Again, how does this improve Wikipedia? Furthermore, I have also contacted people like Anomie when their bots made errors, and have had some 1,000 articles made by Dr. Blofeld deleted when it turned out that the source used to mass create them wasn't reliable enough and many of the articles were simply wrong. I could give other examples as well. I am not focusing solely on Rich Farmbrough, I try to notify everyone about their errors, certainly when they are systematic or recurring ones. With most people this happens less often and the reaction I get is a lot better though, so these don't end up at AN or ANI or ArbCom. Fram (talk) 07:29, 25 April 2012 (UTC)

"because many blatant mistakes are not picked up by the community already ...". Everybody makes mistakes, it is not up to you alone to fix everything, this is a collaborative project. And believe it or not, this project will go on also without you. Yes, errors will still be made, with bots and without bots, with Rich Farmbrough and Δ and without Rich Farmbrough and Δ. It is an utopia if you think that when Rich Farmbrough and Δ are not fixing problems (that is what they mainly do, maybe creating others .. but fixing 10 things and making one mistake in the process still removes 9, so unless you can show me that the edits increase the number of things to be fixed your point is nowhere), that Wikipedia will improve. And it is excessively self-centric to think that Wikipedia revolves around you, and that without you, Wikipedia can not improve (I've been away for most of the time around X-mas, and when I returned, low and behold, Wikipedia was still the same old mess as before). --Dirk Beetstra ^{T C} 08:31, 25 April 2012 (UTC)

Can we get back to what has actually has been said, and not your bizarre reading of it? You accuse me of wikistalking, so I point out that I don't wikistalk RF or excessively focus on him, but that I try to check for errors in general (among many other things). You claim that many editing sprees get ignored, even when they contain errors. True, but how is restricting me from checking some sprees make this any better? I seem to be damned if I do check other people's edits as well (because to bring this up is "extremely self-centric" and shows that I think that Wikipedia revolves around me), and damned if I only focus on RF's edits, because then I am "vindictive", "wiki-stalking", and "selective". Basically, your defense seems to be that everyone is making mistakes, so there is no point in finding them and getting them corrected.

As for RF mainly fixing problems: no, he is mainly making very minor edits which are hardly worth the name "problem". Your math is not convincing: when you have 9 edits that make an article slightly more MOS compliant (by changing an instance of USA to US) but don't add one iota to the understanding or usefulness or content of it, and you have one edit that breaks a link (because the linked article has "USA" in the title, and not "US", then the end result is not 9 problems improved and 1 problem created, but 9 extremely minimal benefits against one clearly worse article. In such a case, it would have been clearly better if none of the ten edits had been made. Fram (talk) 08:44, 25 April 2012 (UTC)

I can not help to see, that you were there with Δ's restrictions, and now here with Rich Farmbrough. You may be right, if you don't report it, Wikipedia may not get 'better', but the point here is: does Wikipedia get worse when you would not report them?

And that is where I disagree. It it still improves 9 articles, even minimally: those 9 articles had the same error before the edit as that was now introduced in the 10th one. --Dirk Beetstra ^{T C} 08:59, 25 April 2012 (UTC)

No, those 9 articles didn't have the error that was introduced in the 10th, where do you get that idea? Fram (talk) 09:26, 25 April 2012 (UTC)

Eh, they did not comply with the MOS before the change, did they? --Dirk Beetstra ^{T C} 10:10, 25 April 2012 (UTC)

At least, that is the conclusion that I draw from '... make an article slightly more MOS compliant (by changing an instance of USA to US) ...' and '.. 9 extremely minimal benefits...' (of course, 'extremely minimal' is utterly subjective, but you could not say '... 9 benefits...', because that would thwart your statistics). --Dirk Beetstra ^{T C} 10:15, 25 April 2012 (UTC)

I think the point was that not complying with the MOS is a less serious error than a broken wikilink. For example, you wouldn't change Communist Party USA to Communist Party US. Jafeluv (talk) 10:17, 25 April 2012 (UTC)

Well, actually, if it was a broken wikilink, then it is a missing redirect, because that is an obvious missing one (and still is). And 'less serious' is just as subjective as 'slightly more'. It still are 9 improvements (and the USA-US improvement is often combined with other improvements as well). --Dirk Beetstra ^{T C} 10:36, 25 April 2012 (UTC)

"if it was a broken wikilink, then it is a missing redirect" Perhaps in the example given, but not in the actual cases involved. Also a fictitious example, but much closer to what actually happened: if a link to Born in the U.S.A is changed to Born in the US, then the problem is not a "missing redirect", but a stupid edit. You may consider it subjective whether this is more or less serious than not complying with the MOS, it is pretty clear to me that the error is significantly more serious though. Fram (talk) 11:02, 25 April 2012 (UTC)

Yeah, but still all the other pages improved, and most on certainly not one point. This gets back to the same friggin' percentages again. How many wrongs, how many rights. I really wonder how many of these cases are breaking that wrongly on that specific task. 1 in 1000? --Dirk Beetstra ^{T C} 12:46, 25 April 2012 (UTC)

There were some problems here, it is true. But they were relatively rare and I went to great efforts to find and fix them all. This particular problem has been brought up several times (there's obviously a shortage) even though it is pretty minor. I'll attempt to get some statistics. Rich Farmbrough, 01:41, 26 April 2012 (UTC).

Well I spent several hours searching for these with no success, I particularly wanted to note whether I reviewed 25,000 or 50,000 articles, since I can't remember which it was. But I suppose it doesn't matter greatly. Rich Farmbrough, 06:05, 26 April 2012 (UTC).

The discussion of them was at User talk:Rich Farmbrough/Archive/2011Oct#Please don't change correct links to redirects (or worse, to redlinks), don't "correct" quotes, and don't remove the end of lines, where I checked 15 edits and found 5 of them containing errors. You continued making the same errors after I had pointed them out, and after I had pointed them out again. Fram (talk) 07:05, 26 April 2012 (UTC)

The error [[Addicted_to_noise|Addicted To Noise]] => [[Addicted To Noise]] was a completely different wholly manual error, which I think anyone could have made. For some reason I have put a lot of effort into analysing the edits of which you were complaining, the error rate of those so far analysed is approximately 1.3%, all of them fixed. I consider 1.3% far worse than I normally achieve, even if they were all fixed. I'm sure that this will all be wasted effort as those who already believe me when I say I put effort into resolving problems, and have basically a low error rate, will need no further evidence, whereas my detractors will use anything to damn me. It will be amusing to see what tactic is adopted though. Rich Farmbrough, 15:07, 27 April 2012 (UTC).

Analysis of Fram's evidence

Links to evidence

These go way, way beyond what is allowed by the rules of Arbcom, I have asked on the talk page that they be struck.

Wanting to mass create articles despite known and unresolved problems

When I created new articles at the rate of less than 1 per hour, Fram invoked the putative ER. According to this I need BAG approval to "mass create articles". To label following the ER that was a result his previous vexatious posting to AN/I as "lack of good judgement" speaks more about Fram than it does about me.

BAG, quite rightly, respond on the BRFA that this is not for them to approve. You will also see on that BRFA my correct prediction that the BRFA would be cited against me in future - which was seen as fanciful at the time.

Expects other people to find his errors for him

Fram totally fails to support this with evidence. Rather the contrary. The edit to the template was designed to quickly show any cases where the categories had been emptied since creation, (the only case where there could be a smidgen of justification for the original complaint). The edit summary of the self-revert is not some kind of arrogance, but a reference to "the wrong version" - it could also be noted that the editor who nominated the category for deletion had offered a compromise on my talk page[49] so I had good reason to believe that the category had been saved.

With the "Wikipdians who like..." categories Fram knows full well (as illustrated in this diff) that I took steps to resolve the source of these categories. He chooses, though to ignore that here, whereas at AN, where he describes it as "hiding some of the problems" which makes me seem mendacious. A lesser editor might have simply ignored it, but Fram uses it to further his campaign.

The last two point have nothing to do with the assertion, and are just an attempt at muck-flinging. Nontheless:

• With the "Statesman" stub, I invite the arbitrators to simply look at the talk page in question. There is clearly no substantive content, much less any that belongs at"Politician".
• Economy of Xiguan was split from Xiguan which was causing errors due to its size and nature. It was nominated for speedy, which was declined, it was then speedied anyway. Since there was a disagreement over the deletion it was not speediable.

Comment by Arbitrators:

Comment by parties:

I note that Rich finds fault with the format of my evidence section, but has failed to correct the error I pointed out in his evidence section, or to provide any evidence for claims he makes, e.g. "I have repeatedly requested Fram not to interact with me, and he has repeatedly refused." (repeated in his replies to Kirill Lokshin on this page, "It is perhaps noteworthy that I asked for a mutual interaction ban with User:Fram several times."). It is of course easier if you can silence your critics, but how that would help Wikipedia is beyond me.

He also claims here that "According to this I need BAG approval to "mass create articles"." This is incorrect, his restrictions states: "unless prior community approval for the specific mass creation task is documented." BAG approval is only one possible method of getting community approval, but far from the only one.

He claims in the "Expects other people to find his errors for him" section that I fail to support my assertion, and then goes on about unrelated things, instead of showing that he ha indeed corrected these errors or that there were no errors in those edits. He doesn't do either of those, but just claims that the edit to the template had good intentions. Yes, well, that wasn't disputed, the problem is that while we assume good faith, we notice bad execution and follow-up. As for the Wikipedians who like X, Wikipedia:Categories for discussion/Log/2012 April 17#Wikipedians who like X is the kind of thing I am talking about. If Rich had deleted these together with the three earlier nominated ones, this CfD would have been unnecessary.

• If I weren't "blocked" I would have deleted them. Rich Farmbrough, 18:12, 18 April 2012 (UTC).
  

Nonsense. You had time enough to delete the other ones, but the block hours later prevented you from deleting these? I would AGf on this if you had actually cleaned up after yourself at other occasions, but there as well you only did some effort on the examples pointed out to you, and failed to check other similar edits (e.g. when you listed biographies from the 1900 DNB as being about "living people"). Fram (talk) 07:53, 19 April 2012 (UTC)

I was notified and I gave permission for them to be G7'd. And you would not AGF. I don't think you have ever AGF'd. If you would AGF then unblock me. Rich Farmbrough, 04:17, 20 April 2012 (UTC).

I have good faith that you have good intentions, I never claimed otherwise. I no longer have good faith that your intentions are sufficient to prevent damage to Wikipedia, as evidenced by the multiple incidents and subsequent restrictions and blocks. And that you magnanimously gave permission for your errors to be G7'd changes nothing about your lack of actions. You have no problem with other people cleaning up your mess, but you are not looking whether you will make or have made a mess with your edits. Fram (talk) 07:40, 20 April 2012 (UTC)

See "magnanimously gave permission for your errors to be G7'd" is the same attacking style, the sarcasm Hersfold detests when he imputes it to me, and then you say "changes nothing about your lack of actions" - it changes everything, or would if you had any objectivity. I have done all I can short of breaking my "block" - actually I would be well within my rights to delete those categories, but there are enough people who would scream blue murder if I did. And you are one, so you are also being hypocritical if you think I could do more. And then you say "you are not looking whether you will make or have made a mess with your edits" this is just your opinion, you have no idea what steps I take either to prevent errors, or to resolve them. You will have seen cases where I went through tens of thousands of edits to check for rare errors and fix them, when this is brought up on my talk page - but you ignore those. On the other hand you don't see, and it never occurs to you, that there are also errors I spot that no one else is ever aware of. And not only my errors but thousands upon thousands of other errors. Moreover you have no idea - and why should you - of the steps I take to avoid errors in the first place. Nonetheless the fact that you ignore the stuff you should know about, shows that you do not only not AGF, but ABF, and even do so contrary to evidence. Well, that's just you and I'm sure you are a great person in other respects, but it is not a suitable attitude to interact with me, nor I venture to say most Wikipedians. Rich Farmbrough, 02:15, 21 April 2012 (UTC).

And now he says the requesting G7 was "to protect my reputation"! I think that I have very little left after Fram has finished with it anyway. Rich Farmbrough, 00:36, 25 April 2012 (UTC).

No, that's not what I said. Your attempt to protect your reputation was "Please also be careful using expressions like "the author created them using an automated process" - this sort of thing, while technically (almost) correct, becomes "Rich runs unapproved bots on his account". Rich Farmbrough, 00:47, 18 April 2012 (UTC)." Fram (talk) 07:33, 25 April 2012 (UTC)

Finally, he still blames the problems on others, and fails to see his own fault in this: "following the ER that was a result his previous vexatious posting to AN/I": no, the ER, was a result of the many previous problems when Rich Farmbrough did similar tasks before, and a means to stop him from creating more of the same. People at WP:AN don't support or impose sanctions because I ask for them, but because they have looked at the evidence presented and concluded that there is indeed an actual problem that needs to be solved or prevented from reoccurring. Fram (talk) 07:51, 18 April 2012 (UTC)

The point you rather miss is People at WP:AN don't support or impose sanctions at all. Rich Farmbrough, 18:09, 18 April 2012 (UTC).

Have you read the actual discussions leading to the sanctions and the blocks? A lot of people supported harsher sanctions, the actual sanctions were a fair and not very strict compromise. They didn't come out of the blue. Let's take a look at Wikipedia:Administrators' noticeboard/Incidents/Rich Farmbrough/October 2010. There was an actual proposal to restrict you like this:

"Proposed: Rich Farmbrough is indefinitely restricted: - from using AWB or any other mass-editing tool; - from running bots of any sort; - from making bot-like edits; - from making more than four edits per minute;"

This was supported (rough count) by 16 people, and opposed by 10 people (some of them opposing any sanction, some wanting a weaker one). I think you should consider yourself lucky with what was the eventual sanction, instead of this much harsher one. Your repeated complaints about the lack of support for the current sanctions, and your rather absurd claim that "People at WP:AN don't support or impose sanctions at all." go against all available evidence. Fram (talk) 07:53, 19 April 2012 (UTC)

You said AN not AN/I. And your count is invalid since the proposal was dropped and several of the editors had fully or partially recanted. Rich Farmbrough, 04:13, 20 April 2012 (UTC).

AN/I is a subpage of AN... My count is invalid since the proposal was dropped? How does the count depend on the status of the proposal? As for the "recanted", I believe I took these into accounr, feel free to provide your own count though. Fram (talk) 07:34, 20 April 2012 (UTC)

The distinction is important. AN does not have the same bad reputation that AN/I has had over the past few years. I note, though, that AN/I is improving, for example by insisting more neutral headings. Perhaps that is why you have moved to AN?

Moreover if you look at the structure of the discussion, as it progresses increasingly editors come to post "oppose" !votes. The last "support" !vote is based on my making a satirical and somewhat self-deprecating comment on a talk page. So the consensus as the discussion develops shifts against the proposal, but no consensus was reached. That is why a new proposal was tried. That failed by any objective measure also, or if you prefer was prematurely closed. The fact that even without significant defence from me (mainly a third party was kind enough to put the work in) , and with a lot of misleading evidence, no consensus to take sanctions was reached shows that the issues are at the very least not as cut-and-dried as you would like to make them. Rich Farmbrough, 15:07, 20 April 2012 (UTC).

Moreover you are right I blame the problems on you. Not the errors, I take responsibility for errors. But the problems are fairly and squarely at your door. Rich Farmbrough, 04:19, 20 April 2012 (UTC).

Comment by others:

Fram, Rich is correct when he states he would need BAG approval to mass create articles; this is stipulated in the bot policy. — madman 16:55, 18 April 2012 (UTC)

Only when it is "automated or semi-automated article creation", not when you do actual manual mass creation. He wanted to create pages with a script (had actaully done more than a hundred of them before, with lots of errors and problems), not manually, despite the clear problems this presented. It was not like the creation of articles on populated places from a database, this was the creation of biographies, with wikilinks, categories, ... all script-based. If he wanted to create these manually but fairly rapidly, there wouldn't be a problem, he could get permission at other places than BRFA. But if he wants to do it with a script, then he indeed has a problem. Fram (talk) 07:53, 19 April 2012 (UTC)

And BAG decided it was too manual to fall under their purview. So I am left with a putative ER that says I need BAG approval and BAG saying I don't. And Fram checking my edits ready to launch an AN/I thread at the drop of a hat. Very cool. Rich Farmbrough, 04:44, 20 April 2012 (UTC).

Sorry, you are of course correct in your clarification; I thought that was implied. I hadn't seen that BRFA, but now that I have, it seems to me that the request was closed because it was seen as a request for de facto modification of the editing restriction and BRFA isn't the proper venue for that, not because BRFA wasn't the proper venue for approving that kind of mass page creation (it may or may not have been in this case; some clarification would have been needed). Obviously, I can't speak for Hellknowz, however. — madman 17:56, 23 April 2012 (UTC)

I provided background on the Statesman deletion at Wikipedia:Arbitration/Requests/Case/Rich Farmbrough/Evidence#Rich Farmbrough's G6 deletion of Statesman and later response were suboptimal. Flatscan (talk) 04:21, 23 April 2012 (UTC)

Yes, sorry we have rather ignored that. But Fram has already introduced it elsewhere. And there's really not a to to add, except that when I went and looked last, there was still no sign that anyone deems anything from the "article" worth merging anywhere. Practically were were better off totally shot of the whole thing, even the few lines of text were glib and in some cases dubious generalisations, not worthy of a dicdef. Rich Farmbrough, 07:15, 25 April 2012 (UTC).

Analysis of evidence presented by Hersfold

Violation of "editing restrictions"

• • Whitespace changes at the ends of lines:

This change used to be an AWB change. It would therefore have been legitimate.

• [50] Here, Hersfold has had to go back 18 months, to a time when I am fairly sure AWB had a separate talk page list, listing the majority of WikiProject templates with their redirects.

List of blocks

Once again linking to links, means over-supplying evidence.

Note that the list of blocks is pretty much all based on items initiated by CBM or Fram.

Hersfold accuses me of running bots from my main account

• He states that these edits must be automatic because of the "variation in edit summaries while running at such a high rate".

The irony here is User:Xeno has accused me of running bots because the edit summaries were the same!

But seriously Hersfold's evidence supports no such statement, because AWB is moderately dynamic in its edit summaries (listing typos etc.) and has a drop down list of recently used summaries. Moreover this is back in 2010!

• Then some rapid AWBing gets the "it must be a bot" treatment.

• "When SmackBot was blocked once, he transferred its operation to his main account" (relevant diff)

Quite simply there were broken articles with references and no references section. CBM had blocked the bot until I agreed to edit the source code for AWB to remove a general fix he disagreed with. I saw no reason not to use AWB manually for a minor task, just because CBM was being a pain in the rear end about some changes there was widespread consensus for. I still think that fixing articles with big red errors on them is preferable to not fixing them.

• Hersfold cites the block logs of SmackBot and HPB. What he does not say is that, ignoring errors, almost all of the blocks were either by CBM or Fram, directly or indirectly. Moreover he promoted the canard, already covered by Whenaxis, that I "repeatedly unblocked his bots in violation of policy" - while Whenaxis has an excuse to think this is the case, Hersfold, as a BAGGER, Admin and Arb, should know better. Every unblock was with explicit or implicit permission.

Rich responds to concerns and criticism with derision

Hersfold says My statement requesting this case was "character assassination" and characterized him "as spawn of the devil"

Firstly there is a complete failure to be literate here. Obviously the statement about "characterising as spawn of the devil" means "assigning a level of wrongdoing completely in excess of not only what was done, but of what was alleged to be done". ZOMG.

Secondly character assassination, it is, if only by degrees. And there is a key point here. I don't really care if "everyone" thinks that I am a bot, a bad editor, or "spawn of the devil" per se. When, however, these unfounded accusations (or even founded ones - for example my typing is not what it should be) become common currency, so that I am faced with people on IRC thinking I run bots on my main account, and, moreover, these beliefs can result in harm to the project by preventing me from contributing, or imposing unreasonable limitations on my contribution, then there is a problem. Let me here demonstrate some of the phrases from Hersfold's "Statement" that are false, and constitute character assasination.

1. "linking dozens of users to IP addresses in a manner that can be parsed by internet search engines"
2. "I had to threaten to block Rich – and did block his bot – to force an end to yet another series of violations."
3. " Rich has proven to be extremely dismissive of the community’s concerns"
4. "insisting ... that those opposing his actions are making similar changes yet remain unsanctioned"
5. "Rich has adhered to neither of these policies."
6. "On two separate occasions, the community has placed restrictions on his ability to edit"
7. "on numerous occasions, he has continued to defy the community on these grounds."
8. "He has repeatedly violated policy"
9. "acts with disdain towards any who oppose his actions."

(There are more that are expressed in ways that make them non-falsifiable, but still present a negative view "I believe that Rich has lost the trust.." "I urge.." etc. etc.)

Taking these by turns,

1. Is simply factually wrong
2. The blocking of the bot was, apparently, caused by Hersfold's impatience and he mentions his trouble controlling his temper.
   1. It is moot that the edits were a "violation"
3. The community had concerns back in 2009, which I spent a great deal of time over. The archives to my talk pages contain literally thousands of threads where I have helped and responded cordially, and many barnstars. Sadly not every visitor can go away happy, but the vast majority do. If there's a problem I fix it, if they need help I provide it, if they are wrong I tell them cordially and gently.
4. This is discussed elsewhere. I have never insisted anything of the sort.
5. Broadly I have been far more of an adherent to WP policies than those who chose to make process the king, and have no respect for other editors.
6. This is discussed elsewhere.
7. There is no defying the community. I am not happy to have two editors following me around, creating the situation where I am worried with every edit I make that I am going to be castigated for "no change to the rendered page", whereas if I make an edit that results in something actually breaking, that is, in theory, perfectly fine.
8. Again an unjustified and unsupported assumption. Assuming this to be true, Hersfold then believes his evidence to support it.
9. This is discussed elsewhere.

• Hersfold says Responds to concerns with sarcasm or derision:

Let us look at one of Hersfold's diffs, [en.wikipedia.org/w/index.php?title=User_talk:Rich_Farmbrough&diff=394350364&oldid=394350245 chosen at random]: Here I would say that the user I am replying to is being sarcastic "Because so many pages begin with....".

In another Fram is informing me he is reporting me to ANI (yet again). I don't see that any response more than a sigh is appropriate.

WP:POINT-y conduct during this case

The common mis-use of WP:POINT...

• Hersfold says Sarcastic comments directed at other parties: title=Wikipedia_talk:Arbitration/Requests/Case/Rich_Farmbrough/Evidence&diff=485637063&oldid=485575492

This isn't sarcasm. I'm sorry that you don't like having your bot's errors reported. It is nonetheless ironic that the bot of the editor raising the Arbcase over issues allegedly to do with mis-performing automation, should malfunction on the very arb case in question, even if it was a trivial malfunction (a bit like removing an un-needed space when another piece of software has had its specification changed not to do so?).

• Hersfold says Frivolous clerking requests:

"As stated in the last diff, Rich is making these requests to ensure strict adherence to what he observes as ArbCom's "rules", since he is accused of ignoring actual rules and restrictions elsewhere on Wikipedia."

This is to completely misunderstand due process. I don't care if the rules are strictly enforced or laxly enforced, but I do care that all parties should know what to expect. If one party is attempting to perform to the letter of the rule and the other's aren't then that party is at a disadvantage. It is somewhat worrying that you cannot see this.

There is no suggestion of linkage here, as should be plain to all readers.

I'm not sure either what the distinction between 'Arbcom's "rules"' (as if they were not real rules) and 'actual rules' are. This sounds like an attempt to have one's cake and eat it.

Hersfold says Rich's restrictions have community support

(section "Review of Rich Farmbrough's cosmetic changes restriction")

Seriously? Let's analyse the contributions to that section:

• Xeno raises two questions in one "Was this restriction duly imposed and should it remain in effect?"

The first question is answered by precisely two of the twelve editors;

• Jenks24 says "Reading over the restriction, I do not see a consensus – rather the closing admin deciding "This Gordian knot clearly needs cutting, and I hereby cut it: the restriction proposed above is enacted" with seemingly no regard as to what the actual consensus is"
• CBM says The fact that the restriction remained on WP:Editing restrictions is evidence that there was consensus for it, as it would have been removed otherwise.

I may be biased, but, to me, one of these statements is rooted in fact, one in wishful thinking.

As well as the angry comments, we have

• DJSasso "at this point I think you are just picking at him."
• Kumioko says "I have seen edits blaming Rich for violating a rule that the accuser either doesn't understand or is choosing to interpret in a way that misrepresents the problem."
• Od Mishehu " I don't think that him changing the "r" at the beginning of a {{references}} between upper and lower case, as part of a relevant edit, is harmful"
• 28bytes says "After reading this whole conversation, I have become convinced that the editing restrictions are more trouble than they're worth and should be dropped. I'm really not seeing a benefit to the 'pedia in keeping these restrictions active."

• We also have Fences and Windows quoted as saying "Blocking for capitalisation changes from {{reflist}} to {{Reflist}} is punitive and provides no benefit to the encyclopedia."

There is much much more meat here for those interested in Wikipedian sociology. For example, why does Rich Farmbrough not put up more of a fight, as CBM says he should? And why do three of the editors who seem to be condemning him at the beginning of the section change to defending him by the end? Could this be WikiHonour? And who is the mysterious Elen who appears so briefly and is never heard form again?

(Note: 28bytes also accurately prefigures my plaint that the ER does nothing about real errors, while penalising over trivia. "A worthwhile restriction would do precisely the opposite.")

Regardless, if this is section is described as "community support" then "one wall and no roof make a house".

Summary

A mixture of misunderstanding (one hopes) and extensive digging into the past that has failed to come up with anything concrete or much that is recent. If diffs from 3 years ago are needed to damn me, then the evidence is scant indeed. When in addition those who are presenting evidence against me need to share some of their diffs, it might be said to be vanishing.

Comment by Arbitrators:

Comment by parties:

Comment by others:

Analysis of evidence presented by CBM

Edit restriction not supported by community

We have already seen all these diffs cited above. They are no more valid here

Once again CBM advances the slightly ludicrous "status quo" argument, that since I did not go and remove the ERs from the list I must have agreed to them. Simply because I choose to avoid confrontation, it does not make that which is wrong, right.

Cosmetic changes

CBM uses the link-to pages of links method to avoid the limits on links. And I trust anyone following the links will draw their own conclusions to the reasons someone would want to have these sort of minor fixes suppressed, unless they are in AWB. Also note that at Elen's request I stopped many of the changes Carl was complaining about, and got consensus for others. Every one of these edits was making a substantive change to the page.

Mass creation

R.F. was endeavoring to create practically every red-linked category on the wiki.

We need to not have red-linked categories. In some cases the category needs to be created, in others the members need to be removed. I have done both. On the last attempt to open an ArbCom case, John Vandenberg remarked "The creation of the ISO templates, and edit to template:interwiki, is absolutely ridiculous to be brought up here. Rich did the right thing by creating those templates." I am not arguing that Jon would necessarily say the same thing here, I do however think that given that comment, it shows that there should not be an automatic presumption that everything I do is wrong.

One example is the joke category Category:Wikipedians in red-linked categories. R.F. created it anyway, and it went to CFD, where R.F. said that by creating it he was actually fixing a problem [51]. He said that someone else needs to actually address the underlying problem, which supports the claim that R.F. sometimes expects others to clean up after he makes a series of blind edits.

This is simply untrue. I said someone needs to address the underlying problem, not someone else. I would have been glad to do it myself. Note also that no-one has addressed the underlying problem - which I would have done if not psuedo-blocked.

Multiple users have attempted to counsel R.F.

CBM uses this header to attempt to dig up dirt from 2006. And yet the damage is done by the weasel words "attempted to counsel". How different the identical content would have been with a more friendly wording! "R.F. discusses his edits with anyone who cares to visit his talk page".

Best practices for bots

• CBM cites these two edits: [52] [53]

This is a known MediaWiki bug. The edit conflict with AnomieBot is not picked up, otherwise these would be valid edits/not done. Unfortunately AnomieBot tunes itself to the same edit delay as HPB maximising the number of edit conflicts.

• CBM says "Bot policy requires preapproval for mass article or article category creation" but actually it says "where those categories are visible in the article space".

Infeasible to 'cross the street'

CBM says An editor cannot simply 'cross the street' to avoid R.F. Here "an editor" presumably means CBM. Historically CBM has stated he only looks at my edits if they are on his watch-list. This statement would seem to be a roundabout way of saying the same thing - without actually saying it. Nonetheless it has been shown recently that CBM does stalk my edits, despite his earlier assertions.

Comment by Arbitrators:

Comment by parties:

Comment by others:

Comment by CBM. Bulleted, including some of my analysis of my evidence:

• In my evidence, I do link to three reports I made to Elen of the Roads about the chronic edit restriction violations by R.F.. However, those three posts together have under 40 diffs if I count correctly, and I have 9 diffs in my evidence, so I could copy all those diffs into my evidence and still be under the 50 diff limit. In any case I did not expect arbcom to look at all the diffs within my diffs, I just linked to the summaries I reported to Elen of the Roads, and expected the arbitrators would do their own investigation into the editing restriction. I will happily make any changes to my evidence that are requested by a clerk.
• The edit restrictions for R.F., unlike those for Betacommand, apply regardless whether there is also a substantive change in the same edit.
  • That doesn't stop people implying that I make a significant number of cosmetic only edits. Rich Farmbrough, 17:59, 18 April 2012 (UTC).
    
• Rich says that, at Elen's request, he stopped "many" of the changes that were inappropriate. This on its own demonstates that he did not make an effort to simply eliminate all the cosmetic changes, as the edit restriction requires. However, if you look at the text I wrote in the three reports linked in my evidence, you will see that the same sorts of problems occur in all three, which goes against the claim that the initial reports were acted on.
  • Here is a typical chopped logic. One would think that successfully using proxy block threats to prevent the type of edit that CBM finds so offensive that he regularly reverts any editor making them (thus making cosmetic only edits that do nothing other than what he complains of, or often reverting real fixes to vandalize articles) would result in his laying off and getting on with something useful. Instead if I stop I am damned and if I don't stop I am damned. Rich Farmbrough, 15:54, 18 April 2012 (UTC).
    
• Rich says that [54] is a "known Mediawiki bug". But the "bug" is that his bot changed the whitespace of the page. If it had not, and had made an edit identical to the one that AnomieBot made, then the edit by his bot would have appeared to be a null edit and would not show in the page history. R.F. is blaming Mediawiki for a problem that would not have occured if he followed his editing restriction.
  • Er... no. The combined edit could be perfectly permissible even under your rules. Get a grip. Rich Farmbrough, 15:54, 18 April 2012 (UTC).
    
• The point of the "infeasible to cross the street" paragraph is that it is impractical to impose a sanction that asks R.F. not to interact with other editors, unless there is also a sanction that drastically restricts the number of pages that he edits. There is no way that any other editor (me, Fram, Hersfold, Elen of the Roads, or anyone else) could avoid encountering R.F. when R.F. and his bots edit thousands of distinct pages per month.
  • Ah I see. You want to be allowed to continue your campaign? Because despite your protestations, even when I have avoided editing any articles that you edit, you still complain about what I do. Moreover you are not compelled to follow me to pages I create, such as BRFAs.
  • Strange that you cannot see that by sabotaging BRFAs you are forcing me to make my edits manually - which you apparently dislike. Rich Farmbrough, 15:54, 18 April 2012 (UTC).
    
• The point of the "attempt to counsel" paragraph is that R.F. has been asked may times to change his editing, and has declined to so to. I view this as demonstrating a long-term pattern of incollegiality and disruption rather than a temporary mistake or an argument between only one or two editors. I personally decided a while ago to recuse from any more admin actions on R.F.'s bot to see whether other admins agreed that there were problems -- and they did.
  • I have not declined to do so. I have made innumerable changes to my editing patterns. And yes, you recuse from "admin intervention" but you still post on other admins talk pages, and on BRFAs and other places, until you get a response. I would rather you had the guts to say that you fight improvement wherever you can, as shown by mindless reverts of other editors which litter you contribution history. And I notice since you derailed my recent BRFA - leading indirectly to this Arb case - you are now taking over the task yourself - wasting the time and effort that have gone into coding the task. Rich Farmbrough, 17:58, 18 April 2012 (UTC).
    
• Regarding Category:Wikipedians in red-linked categories, it was deleted and emptied, and then other editors added themselves to it.
• — Carl (CBM · talk) 12:44, 18 April 2012 (UTC)

BAsically edits such as this are edit warring and vandalism. And CBM has made many of them. Rich Farmbrough, 18:02, 18 April 2012 (UTC).

If you consider that to be vandalism, I am even more in favour of desysopping you. Fram (talk) 13:30, 19 April 2012 (UTC)

IMO FRAM, CBM and several other editors are just as guilty for perpetuating the cycle. You and others have adamantly opposed these types of minor edits as being pointless. Regardless of whether they are or are not, if that is the case, then reverting it is even more wasteful than simply leaving it alone. If these would have been left then the bot would not go back and try and change it again and half the problems with arguments and edit warring would be solved. You are helping to perpetuate the cycle by doing these reversions. In the end, regardless of wether they are necessary or not there is no harm done by doing these minor edits and therefore no reason to revert them other than to instigate an argument. Kumioko (talk) 17:08, 19 April 2012 (UTC)

I don't revert such minor edits if they don't contain actual errors. Please don't accuse me of things without backing it up with some evidence. Fram (talk) 18:50, 19 April 2012 (UTC)

In fairness that comment was directed mostly at CBM but I have seen you revert a few that I thought were petty. I am not going to go mining through thousands of articles to find them though. Kumioko (talk) 19:50, 23 April 2012 (UTC)

Analysis of evidence presented by Charles Matthews

Charles Matthews claimed that I at first only raised a "technical issue", which is incorrect: what I wanted was information on how we could best attribute the articles created by Rich Farmbrough, since they weren't properly attributed despite being near word-for-word copies. This is not really a "technical issue" but a basic requirement.

When PBS gives two examples of articles RF made in that context, congratulates him on the effort, and notes some problems with them. I then don't "change tack" or use "a single instance" as "proof" of anything, I take one of the two examples given by PBS and indicate that there is more wrong with them than may at first be obvious, and that he IMO should not really be congratulated for these. This is a normal response to what was said right before, not a change of direction (if anyone changed direction, it was PBS; Note that I hadn't even mentioned RF, the bot request, or anything else related to these problems before PBS started about it. My reply to PBS caused him to change his position[55]: "Silly me, I had assumed that these were proofread pages. Given that they are not, we seem to have to things that we can do. Either fix them here on Wikipedia, or put them up for a mass deletion."

For some reason you then interject your edit between my edit and PBS' response (making your evidence claim that "the threading now makes it hard to reconstruct the chrnology" rather ironic)[56], and start out of the blue with "Please assume good faith", without indicating where or how I haven't done this.

Next you claim that "despite this spat[46] prompted by Fram saying we didn't care, PBS helped Fram with the solvable problem." This conveniently ignores the pôst before that[57] which has the edit summary "A thankyou for the work of PBS, the response of the others leaves me rather disappointed though" and also the text "Thank you to Philip Baird Shearer, who effectively did something to tackle this and works on it, but the other responses here leave me rather disappointed." I think this makes it rather clear that I did appreciate the work PBS had put into it, contrary to what the evidence by Charles Matthews seems to suggest.

As for "There was no mystery at any point which the "pilot project" articles were, so it began to look like a personal attack wrapped in technical issues. Purism on Rich's part meant he didn't apply {{DNB}} to his pilot articles as WP:PLAGIARISM would suggest. But he gave the sourcing: Fram's point was about tracking. That's it, for policy.": wrong, there was no easy way to find the articles created by Rich Farmbrough in this manner, and using a reference instead of an attribution is a serious problem, which is easily understandable for a newbie, but not so much for an admin who wants to mass create such articles. Attribution is not optional but policy-required, despite what you seem to believe.

Finally, when you do a pilot, you don't create 100+ seriously deficient articles, don't check for errors, don't correct errors after they have been pointed out by others, and then request permission to use a bot to created thousands more of the same. If you can't see what the problem is with this approach and why it needs to be stopped, then I can't help you. Fram (talk) 14:40, 19 April 2012 (UTC)

Comment by Arbitrators:

Comment by parties:

I never requested a bot to do thousands more. That is plain untruth. It is typical of Fram's style of arguing. He also weasels "100+ seriously deficient articles" in there. Elsewhere he says "No one is arguing that no good came out of it" when I point out articles have made DYK on the main page. But of course he still throws around these sweeping generalisations, untruths and half-truths. Rich Farmbrough, 01:47, 20 April 2012 (UTC).

Wikipedia:Bots/Requests for approval/Rich Farmbrough (mass article creation) claims that you would create "3,000 - 4,000" more such pages "manually", when in reality it was a script that would run and create the pages. Not a bot by name, but a bot by nature. Fram (talk) 08:09, 20 April 2012 (UTC)

And the difference between a bot and a script is the difference between a bot and a human. I used a script to link several thousand Wikisource articles to their Wikipedia counterparts, a bot would not have been suitable, since many required detailed investigation. Similarly using a script to create these articles in small handfuls or individually then doing extensive manual editing is completely different from using a bot to create thousands of them. If you can't see the difference you should not really be trying to contribute to these discussions. If you can then you should stop being disingenuous. Rich Farmbrough, 14:12, 20 April 2012 (UTC).

I can't see the difference in the articles you actually created when using this script, no. [58] This was the state of an article you created with the script, and then abandoned (in the mainspace). You didn't make a single further edit to the content, only some very minor cosmetic ones (including, of course, a technical error in your orphan tag[59], fixed not by the intervening edit by Helpful Pixie Bot, but by AnomieBot[60]). The actual changes from a bot generated article to a decent one were done only months later by other people like PBS, Slowking4, and myself[61], despite the fact that you were made aware of the problems with this (and similar) articles before this. I would understand a bot doing such a poor job, I have trouble though to see an intelligent human editor doing this (not once, but many times in a row). Fram (talk) 14:44, 20 April 2012 (UTC)

And the idea behind Wikipedia is collaboration. While the correction of the category by you is welcome, you could have told me of the problem -since fixing anything feeds your obsession with "Rich expects others to clean up". Instead you fix them, play the martyr card and use that to sabotage a BRFA (the reason you were looking in the first place). Moreover you complain about the lack of a common category, which had been removed as a result of your complaints. There is no good faith here, just rather transparent attempts by you to wiki-war. Rich Farmbrough, 01:12, 21 April 2012 (UTC).

Rich, "you could have told me of the problem -since fixing anything feeds your obsession with "Rich expects others to clean up"." Actually, I changed it on October 10; but I had told you on September 8, more than a month before, that the articles you created in that bunch were listed in that specific, incorrect category, in Wikipedia:Administrators' noticeboard/Archive227#Rich Farmbrough violating editing restriction: "a number of these pages are categorized not by year of birth and death, but as living people, even though all of these people are very dead; e.g." with 7 examples following. As I stated in my evidence section, but which you dispute; you didn't check whether other articles had the same problem, you didn't even correct the problems with that article that I did mention ("Similarly, John Barrow (fl.1756) links to the surnames of explorers, with pages like wafer bluelinked, while e.g. Van Noort is a redlink that could easily have been turned into Olivier van Noort.) You did claim in that discussion in a reply that "Fourthly it is an assumption that these people are dead and not caught in a time rift, or ascended, or simply very very lucky. However I agree marking them as living is unwise and that is easily fixed. (Now fixed.)"

Not only did you not check beyond the examples given, even among those seven, you missed one, which I fixed a month later as well[62]. So I did tell you of the problem, you did claim it was fixed, and in the end it came down to others to correct it anyway. Which is what I have claimed all along, but which you state is an "obsession" and "playing the martyr card". Could you, if you want to defend your actions, stick to the truth instead of making up things which are easily shown to be completely false? Fram (talk) 17:19, 21 April 2012 (UTC)

Again, you assume that I did not check "beyond the examples you give" this is easily falsifiable with one of several examples.[63] And you could simply (at the time) have verified this. And since you were stalking my edits, you probably were aware. How the 7 other items you found were missed I don't know, nor do I much care (probably due to the AN/I stink you were creating at the time), but I find it repellent that you used them to attack me, rather than pointing them out to me top be fixed. Rich Farmbrough, 04:26, 22 April 2012 (UTC).

How many times do you actually want things to be pointed out to you then, since once is apparently not enough? YOu claimed in the ANI discussion that they were "easily fixed" and that they were fixed. Apparently the fact that this wasn't true is my fault somehow, but I'm quite used to that by now. As for stalking your edits, I only look at a sample, you make too many to trawl through them all, and usually looking at a sample is more than enough to spot the errors. Fram (talk) 09:39, 22 April 2012 (UTC)

As I say your statement that I did not check "beyond the examples you give" is easily falsifiable with one of several examples.[64] What you are supposed to say now is "Rich, on that one fact you are correct, you do indeed make some effort to find other examples and fix them." Not run off at some other bait-and-switch tangent. Rich Farmbrough, 00:43, 25 April 2012 (UTC).

It's hardly another tangent when you started of with "you could have told me of the problem". So you did some haphazard checks and corrections in this case, and not no checks as I erroneously thought. It was hardly a thorough check though, missing even one I gave as an example, and a number of others, and for some reason it only corrected this one problem and none of the others. And now that you have lectured me on what I should say, perhaps you can take your own advice and correct your own earlier error that "you could have told me of the problem", when I clearly did? Then again, I have requested you on this page already to follow your own advice, but I haven't seen any inclination to do so there either. Fram (talk) 07:39, 25 April 2012 (UTC)

"So you did some haphazard checks" really, can you not make one civil comment? I am of Rich Farmbrough, 12:20, 27 April 2012 (UTC).

Let's clarify this "bot to do thousands more" with a direct quote from the document:

Estimated number of pages affected: 3,000 - 4,000 over three - four years

This is a rate of maybe 3 per day, in fact slightly less. To call this a "request for a bot" is ludicrous on the face of it. When the document also says explicitly

Automatic or Manual: Manual

it is clearly a bad faith lack of AGF. Rich Farmbrough, 01:50, 21 April 2012 (UTC).

This quote from Charles is something Fram could take to heart. If you want to engage constructively with the people on a WikiProject, it is not the best thing simply to say that their attitude is lacking. Charles is far more diplomatic than I, but if neither his kind words, nor my blunt ones get the message across I'm not sure what can. Rich Farmbrough, 04:01, 20 April 2012 (UTC).

Comment by others:

Whether manual or automatic, this looks like a mass creation of rather deficient articles. Importing 1000's of them for others to bring up to Wikipedia standards IMHO is "collaborative editing" in about the same way that 52 Pickup is a multi-player card game. The heavy lifting of getting this sort of article in shape is researching and summarizing present-day scholarship on the article subject, not merely touching up the initial import. Rich should be willing to participate in that work if he's going to be the one creating the articles. It otherwise doesn't come across as an attractive division of labour for those of us who are not Rich. 64.160.39.210 (talk) 07:48, 23 April 2012 (UTC)

That is what the WikiProject is doing in part. I have indeed done several of these myself, including research into contemporary records. But that is to miss the point - it is like throwing the starter motor out of your car because it does nothing for the vast majority of the journey. Or castigating the stub-sorters, or typo fixers, or copy-editors for not "doing research". Stub sorters are the most wonderful people in the world, because they do in seconds what would take most of the rest of us at least several minutes. And there's more to importing these articles than meets the eye, like most things in this life. The idea is to provide a working base and - bearing in mind that we could, theoretically have already had all these articles for several years at least - something for a reader to work off. In many cases we have a much inferior article based on the 1911 EB, which is a summary of the DNB article. The longer we leave this work, the less benefit we get from the effort already put in. I am of Rich Farmbrough, 01:28, 27 April 2012 (UTC).

I'm certainly not castigating typo fixers. It's more like castigating high-volume editors who persistently don't fix their own typos because they expect other people to do it. Anyway, DNB typos should be fixed on Wikisource before any import takes place, per the WP DNB page. If you then import a DNB article and bring into shape for Wikipedia before going on to the next one, that's great. Or if you bring in one or two that you think are interesting and then don't get around to working on them, that's no big deal. Volunteers don't always follow through with their plans, and your specific selections at least reflect some kind of intelligent judgment, that is a contribution to the project in its own right. It's vacuuming them into mainspace indiscriminately with a bot that I have problems with, since (like 52-pickup) it mainly just creates crap for other people to fix. Obviously that is an editorial and philosophical viewpoint that isn't be held by everyone, but it's already been discussed by the community, and the result is WP:MASSCREATION, which should be taken seriously. If you're systematically creating articles that would be rejected at WP:AFC if submitted by unregistered editors, you're doing something wrong. 71.141.89.142 (talk) 01:07, 29 April 2012 (UTC)

Template

Comment by Arbitrators:

Comment by parties:

Comment by others:

General discussion

Query: Has anything good come out of these so called editing restrictions? Rich Farmbrough, 08:51, 18 April 2012 (UTC).

Comment by Arbitrators:

Comment by parties:

Yes, the abundance of problems before and in October 2010 has been reduced to more occasional but still too regular lapses, and the editing restrictions make it clear for people less familiar with the situation that when new problems are reported, they aren't a first time occurrence but part of a pattern. Fram (talk) 09:01, 18 April 2012 (UTC)

That's how you see it. From my perspective I voluntarily changed the way my bot worked, and the ER was imposed some considerable time later. Endless fun for you, of course, and other wiki-lawyers. But no actual benefit, in fact dis-benefit, since even more of my time is wasted on you, and the blocks you inspire, than before. Rich Farmbrough, 17:48, 18 April 2012 (UTC).

Diffs? Links? All I see is that the restrictions were swiftly imposed after you repeatedly made problematic edits, not "some considerable time later". Wikipedia:Administrators' noticeboard/Incidents/Rich Farmbrough/October 2010 started on October 20, 2010, and the ER was placed on October 26, 2010[65], after 16 people supported and 10 opposed a much harsher sanction. The second restriction was placed on 13 January 2011[66], the same day that Wikipedia:Administrators' noticeboard/Incidents/Rich Farmbrough/January 2011 was started. In both cases, the final straw that led to the restrictions was not your bots, but actions done with your own account. That you "voluntarily" changed how your bot worked obviously didn't really solve the problems you made on your own account.

And no, trawling through your thousands upon thousands of edits (user and bots), doing the work you should have done in the first place (checking your edits, correcting your errors), and often getting abuse instead of thanks as a reply for my diligence and my attempts to improve the way you work, is definitely not what I consider to be "fun". Starting AN and ANI discussions, or participating in ArbCom cases, is even less fun. But because you are an admin and bot operator, and because you make an enormous amount of edits, there are few people looking at your edits, never mind looking at patterns in them and repeated problems. Most people never see your edits, many others accept them in good faith (since your are an admin or a bot), never questioning whether your improvements actually improve anything. Some people notice a problem with one of your edits, and bring it to your talk page. You correct that one error, they are happy, and the other similar errors get by unseen and uncorrected. Only a few people take the time to actually make sure that you don't go off again with hundreds or thousands of problematic edits, and since you don't like this, you want these people to stop doing this, instead of changing your own behaviour and adhering to community standards and guidelines. I don't do this for fun, I do this beacuse it is necessary and few other people are willing to do it. Who can blame them? Fram (talk) 08:32, 19 April 2012 (UTC)

Wrong I welcome constructive checks by any editor other that you. I thought I made that clear. Rich Farmbrough, 01:42, 20 April 2012 (UTC).

And by the way it is downright misleading to say " You correct that one error, they are happy, and the other similar errors get by unseen and uncorrected." You know full well (or you should do, you have been stalking me for long enough) that I also look for similar problems and fix them, including those created by others. Again you throw away a line of horrendous implication in the middle of a random comment, and spread your misinformation. Rich Farmbrough, 03:42, 20 April 2012 (UTC).

This claim is actually present in my "evidence" section and supported by the facts. The "horrendous implication"s of it are your restrictions and block log. Fram (talk) 08:31, 20 April 2012 (UTC)

And strangely you claim not to care about the sort of change covered in the (first) ER, and only to care about "real errors". Yet you do your best to draw attention to "violations" repeatedly, and with some vigour. Rich Farmbrough, 18:54, 23 April 2012 (UTC).

I have only drawn outside attention to your violations when a) they caused problems (either making existing articles worse, or creating new problematic articles, templates or categories) and b) you didn't correct the problems when I pointed them out to you on your talk page. B) has been skipped two times recently, since it usually had limited to no effect with serious problems. A) is the central issue. I haven't tried to draw any outside attention to your violations when they didn't cause any problems, I have better things to do with everybody's time than that. The restrictions were not a goal on their own, they were implemented to prevent further problems; so I brought things only to AN or ANI when you violated your restrictions and these violations again created problems and you were not cooperative in solving them (e.g. by simply continuing with the task and creating the same problems over and over again, even after I had contacted you and given examples of the specific problems the task caused). Fram (talk) 20:06, 23 April 2012 (UTC)

But if someone was causing problems and was not solving them, then further measures would need to be taken anyway. So the ER makes no difference. I am of Rich Farmbrough, 01:14, 27 April 2012 (UTC).

Comment by others:

No, nothing good has come out of it. Because you're not abiding by them, Rich. Whenaxis (contribs) ^{DR goes to Wikimania!} 21:08, 18 April 2012 (UTC)

Actually by and large I am complying with the first. You haven't realised, clearly that it is an ongoing campaign by two editors (or two campaigns by two editors, if you prefer) to rigorously prevent me from doing anything that they can by any means they can, that is abusing them. But to the extent that things trivially break it, it seems to me, and to many others, that this is not something that is a benefit to consistently attack someone over. Just get on with the work, and forget all this stupidity. Even Fram claims now not to care about the ER, but "only the substantive mistakes". Would that it had always been that way. Mistakes we can deal with. Aggression and vexatious complaints not so much. Rich Farmbrough, 03:38, 20 April 2012 (UTC).

Maybe I should remove my evidence (and my analysis of your evidence) since it's just a cat fight between several editors that just got ugly without being resolved properly. Whenaxis (contribs) ^{DR goes to Wikimania!} 18:24, 20 April 2012 (UTC)

The reality is CBM and Fram are the main two editors who have the skill and patience to figure out and report Rich's missteps, other than Rich's fellow bot enthusiasts who are basically enablers of the problems and who don't help. CBM and Fram should be thanked for this tedious and unpleasant work. If they stopped, that just means a lot more errors would go undetected. We'd be worse off for that. I've had some time the past couple evenings to look at old incident reports. I had seen most of them before, but had forgotten how bizarre they were (including recent ones). I am depressed. 67.117.130.107 (talk) 03:25, 27 April 2012 (UTC)

You are calling Xeno, Anomie, Hersfold and Headbomb "enablers"? Nice. I am of Rich Farmbrough, 12:17, 27 April 2012 (UTC).

I just wanted to note that although CBM and FRAM and both capable editors they are far from impartial in this case. As I have said before I don't really care about the changes Rich is making to minor things and I find arguing about these edits to be a waste of time because whatever they do or do not do they don't hurt anything either. Both Fram and CBM seem to take the complete opposite view as I do and have become almost obsessive in hounding Rich about any edit he makes. I have seen them both leave messages on Rich's talk page and stop his bots for the most insignificant of issues which has in the past caused more harm than good in many cases. One of which led to Anomie having to add logic to his bots that would date maintenance templates because Rich's bot was stopped so often it was causing problems elsewhere. So although I agree that CBM and FRAM are both good editors in their own right I do not IMO think they need to be congratulated for running Rich down.Kumioko (talk) 14:17, 27 April 2012 (UTC)

I also find it a little dubious that this case has attracted a couple of IP editors with such insights into the inner workings of WP and this case. I don't wanna seem like I am accusing anyone of anything here but it seems a little strange to me. Kumioko (talk) 14:17, 27 April 2012 (UTC)

Ehh, ok, maybe I went overboard about "enablers". An obvious example though is someone like Dirk B., who also contributed heavily in Betacommand 3. While I'm sure Dirk is a nice person who means well, his desire for WP to endlessly let ill-advised bots run rampant is IMHO ridiculous. I'm mostly an admirer of Xeno, who is pretty good about this stuff, and who has been quite critical of Rich's activities in multiple ANI discussions. Hersfold obviously is the one who brought this case against Rich, also showing wisdom, though I think Hersfold's proposed remedy (desysop Rich) doesn't exactly fit the problem.

I do think Headbomb is something of an enabler (see my discussion with him further up, where he wants to let Rich keep running bots while blocked). Kumioko, like Rich, in my opinion has repeatedly shown poor judgment about bots and caused disruption with them. As for my editing from IP addresses, look at the bright side. Without an account, I can't get editing tokens from the API, which makes me significantly less likely to run any bots myself. 71.141.89.142 (talk) 00:43, 29 April 2012 (UTC)

Just to clarify the only thing I truly did wrong with my bot is 2 edits to 2 talk pages to let them know I couldn't respond because someone felt the need to block me in the middle of a dozen discussions and WP isn't smart enough to allow specific things to be edited while blocked. Its all or nothing. Other than that my bot did fine except for some editors with ownership issues not liking my selection of articles. I'm still sore about that whole ordeal rather than rub salt in the wounds lets get back on topic. I too have a lot of respect for Xeno and Hersfold and have mostly positive interactions with them myself. Everyone though, as my own history shows, is entitled to bad judgement from time to time. To follow suit with your comments though I too think that de-sysopping Rich would be pointless (especially since it wasn't his admin decisions that are in question). Kumioko (talk) 00:59, 29 April 2012 (UTC)

I'm not trying to rehash your past dramas but I think your description of your own history is in error. If you're exercising better judgment now, that's great. (If you really want to discuss specifics, we can do so on your talk page). The level of bad judgment acceptable from an editor depends on what kind of edit it is. For normal article editing, quite a lot of bad judgment is ok. For example, outright vandalism is pure bad judgment, but we revert it and shrug it off with {{test1}} unless the vandal is persistent. Other stuff like sysop actions are potentially disruptive enough that we limit access to them to people with demonstrated skill, and expect them to make relatively few errors. Running bots that edit large numbers of pages is up there with potentially disruptive sysop actions in terms of limited error tolerance. Bot ops don't have to be sysops, but to avoid errors, there are principles (WP:BOTPOL mostly) they are supposed to follow closely, chief among them that bot tasks require BRFA's. The errors of Rich, yourself, Betacommand, and various others, generally revolve around gaming and evading BOTPOL. The community does too little about this in part because the technical nature of the problems isn't so comprehensible to most general editors. The community has also left BOTPOL mostly up to the bot community, which has in turn been way too enabling (that word again) of gaming. I may be doing a lousy job of it, but what I'm trying to do by commenting in these cases is to help raise general understanding of the issues, in the hope of seeing better community intervention against future bot-related drama than what we currently get. 71.141.89.142 (talk) 02:16, 29 April 2012 (UTC)

71.141.89.142, just to be clear, I am against having bots going rampant (I have no clue how you come to that assumption), and I will block bots and editors (even admins) who go rampant with their bots or scripts. I have done so (yes, I did block an admin for running a script on their main account, and I will do so again if needed). --Dirk Beetstra ^{T C} 05:25, 29 April 2012 (UTC)

There is a basic misunderstanding present about BOTPOL. I rather think that had I run, let us say, 6 independent and unlinked accounts, one as botmeister, one for large manual jobs, one for heavy content editing, one for policy and discussions, one for admin actions, and one for technical template and category work, they would all be (relatively) unmolested. As it is people make all sorts of assumptions, and are not shy to state their conclusions without checking the basis for them. Out of context statements like CBM's "BOTPOL applies to ... use of scripts..." don't help - while BOTPOL does apply to use of scripts, etc. it is only the specified section, and really this says little to nothing not already covered by other policies (which is why I oppose it - these matters can be dealt with as well by the general rules as by specific rules, which are therefore a form of WP:CREEP). Rich Farmbrough, 20:04, 29 April 2012 (UTC).