Wikipedia talk:Wikipedia Signpost/2013-06-19/Arbitration report

← Back to Arbitration report

Discuss this story

• Another prolific and valuable editor is biting the dust because Wikipedia insiders don't particularly like him. Rich is a martyr for the cause of content builders. Wer900 • talk 21:35, 20 June 2013 (UTC)
  • Actually he is generally well liked, the issue is one editor who reports him for any minor point and an administrative process that cannot apply perspective. Regards, Sun Creator^(talk) 00:07, 21 June 2013 (UTC)

• It's nice to see in depth material like this in the arbitration report. (It may have sometimes been like this before, but I don't remember it recently.) --Demiurge1000 (talk) 22:09, 20 June 2013 (UTC)

• "In depth" is nice, but the article would be better if it were less one-sided in Rich's favor. Fact is, he's blocked for a year because he's violated his sanctions multiple times, and been caught doing so despite his efforts to conceal his misdeeds. The original sanction would have been long since commuted if Rich had done the one thing requested: Put away all automation and type edits in by hand, like the vast majority of first time users do. The idea was to provide Rich an opportunity to have empathy with those whom his automation mistakes impacted. Instead, he has challenged the legitimacy of the sanctions, violated them willfully and repeatedly, and now seeks to be re-tried in the court of public opinion. The person who is solely and totally responsible for Rich Farmbrough being blocked for such an extended period of time is Rich Farmbrough--every process, at every turn, gave him ample direction on what to do, and ample opportunity to comply. It's a quite regrettable outcome, but one of his own making. Jclemens (talk) 23:35, 20 June 2013 (UTC)

• Took the words out of my mouth. Though you're the ex-arb in the room, so I guess it's for the best you're the one who said them anyways. :P I'll never understand how some people's response to a sanction can be to try every possible action except to willfully comply with it until they've proven it can be lifted (which, history has shown, is actually a lot easier than it sounds). This article, while interesting, fails to give due consideration to the reasons that ArbCom repeatedly denied RF's requests to lift the sanction. A two-sided examination of the events here would have referenced the unanimous dismissal of the "omnibus appeal", the fact that several arbitrators felt Rich's behavior could soon merit an indefinite ban, etc. I'm not saying Rich has no valid complaints, but let's remember that this is someone who has, among other things, attempted to get arbitration findings struck as NPA/BLP violations. I must say, I'm saddened to see material like this appearing in this section of the Signpost. Normally, at the end (well, now, the middle) of a dramatic week, you can always settle down with the Signpost and read a nuanced and balanced summary of whatever's unfolded. The Abritration Report, especially, has always impressed me with its neutrality, and this edition is an unfortunate departure from that tradition. This should have been published as an Op-Ed. — PinkAmpers&^{(Je vous invite à me parler)} 00:01, 21 June 2013 (UTC)

  I have to say, being unfamiliar with the case, that it's not at all clear to me why Rich was sanctioned in the first place, let alone blocked currently. Powers ^T 01:29, 21 June 2013 (UTC)

  And by that do you mean that you agree the article's biased, or that you agree with the viewpoint advanced in the article? — PinkAmpers&^{(Je vous invite à me parler)} 02:44, 21 June 2013 (UTC)

  I don't have enough information to say. It feels like the article is incomplete, which some may read as biased. Powers ^T 23:49, 22 June 2013 (UTC)

  Incomplete, yes. When I first started the whole interview process, I had hoped for things to be wrapped up in a neat package. Apparently WP isn't like that. I would defend the neutrality of the report though: both Mr. Farmbrough and the committee had opportunity to articulate their views. I am well aware that this interview type of format takes the report out of its comfort zone, and that some readers may miss the usual formula. As for nuance, I hope it is still there, but that is up to the reader to judge. Neotarf (talk) 08:39, 2 July 2013 (UTC)

• Well said, better than I could have done. Anomie⚔ 02:29, 21 June 2013 (UTC)

• I note one glaring error: AnomieBOT contains neither code nor data from any of Rich's bots, and the implication is frankly absurd. Since this piece calls out tag dating in particular, let's look at that. AnomieBOT's tag dating came about because Rich couldn't manage to keep his bot unblocked to do the task, and no one else wanted to do it long-term.

  Code first. It seems clear to me that even the first version of AnomieBOT's TagDater.pm is in a distinctly AnomieBOT style, but direct code comparison is of course impossible since Rich hadn't and hasn't to my knowledge released his code publicly.

  As for data, AnomieBOT's tag dating has two major sources: WP:AWB/DT for the list of templates, and the category hierarchy under Category:Wikipedia maintenance categories sorted by month to find pages with templates needing dating. Regarding the former, I note that, despite running his tag-dating bot, Rich appears to have made only two trivial edits[1][2] to WP:AWB/DT. Nor did he create Category:Wikipedia maintenance categories sorted by month, although he does seem to have played a part in some of the complex templates currently used in that hierarchy. I don't see that contributing to the templates used in an existing category hierarchy rises to the level implied here.

  As for the many lesser inaccuracies and omissions in this piece, I think I'll refrain from specific comment. There are enough people who idolize Rich for various reasons (the saddest being, IMO, those who hold him as some sort of mascot for people "oppressed" by "Wikipedia's [ArbCom/admin] oligarchy") that I don't want to get into a drawn-out argument with. Anomie⚔ 02:29, 21 June 2013 (UTC)

  • If this makes any difference, the exact wording of Farmbrough's statement to the Signpost was, "All AWB bots and AnomieBot effectively use some of the code/data from my bots which was turned into Wiki Pages." Neotarf (talk) 18:22, 21 June 2013 (UTC)

Since there are some remarks here about neutrality, let me make it clear that none of the opinions expressed in the piece are my own. I will try to find time to say more about this later, but for now let me just say that if someone has a question about who or what is being cited, I would be happy to provide more detailed diffs. Neotarf (talk) 09:26, 21 June 2013 (UTC)

None of the opinions, perhaps. But the selection of opinion to present leading to the POV? Anomie⚔ 12:58, 21 June 2013 (UTC)

If there is some opinion or discussion somewhere that has been left out, anyone is welcome to add it. Neotarf (talk) 08:39, 2 July 2013 (UTC)

This is probably the worst-ever article I've read in the Signpost. If it was an essay by Rich about why he reckons he was unfairly treated that would be one thing, but it's basically an opinion article which cherry picks Rich's interpretation of events while leaving everything else out. Where's the coverage of his multiple attempts to push against his restrictions? (which is what contributed to the block). Some of the prose would be deleted from a stub article - for instance "It has been suggested that this will have a chilling effect on other bot operators, that they will be afraid of making mistakes and getting banned. Says one talk page commenter, "A lot of bot ops and potential botops think twice before starting a bot. I have talked with several editors who want too but are afraid if they make mistakes that the zero defect mentality will get them banned." - who has suggested this, and who is the mysterious "one talk page commenter"? Given the number of chances Rich was given there doesn't seem to be any imminent danger of them being "banned"(!) - are bot owners really so silly that this represents their consensus opinion? Nick-D (talk) 10:00, 21 June 2013 (UTC)

Nick-D, if the links to the discussions weren't clear enough, this particular discussion took place on the blocking admin's talk page. I was not covering the arbitration committee at the time of any of Farmbrough's previous interactions with it, however, (if that question was not merely rhetorical) any available information about them would be found in the Signpost archives. It was not our goal to rehash material that has already been reported, but to provide the briefest of backgrounds for readers who are new to the situation. Neotarf (talk) 18:09, 21 June 2013 (UTC)

If that was your intention, then I'm afraid that I don't think that it was successfully executed. This is a woefully one-sided view of things, which basically misleads readers by presenting cherry picked material focused on Rich's justifications for his actions. Nick-D (talk) 01:18, 22 June 2013 (UTC)

The arbitration committee was given a chance to respond before publication, and they did so. If there is material somewhere that has not been represented, anyone is welcome to point it out. Neotarf (talk) 08:39, 2 July 2013 (UTC)

• I am one of the editors who concluded that Rich Farmbrough was using automation in violation of his agreement not to do so. The spirit of the restriction was to not change a bunch of things with one mouse click or key press. Rich Farmbrough had a history of doing that and making the exact same minor error in many places. Rich made it clear in previous discussions that he realized that doing that was the behavior that he was not to repeat. One could argue that doing a search and replace and hitting the OK button after examining each replacement was allowed, but doing a search and replace and letting it make multiple changes all at once would be an unambiguous violation, and I am convinced that this is what happened.

Consider this edit. It may help to look at the before and after pages as well as the diff. Clearly he was attempting to fix the places where a word like Madhubala was surrounded by ‘ and ’ and also in italics using the usual '' wikimarkup. In other words, change ''‘Madhubala’'' to ''Madhubala'' by removing the ‘ and ’.

This is a standard automated-text-repair operation, and I have written small scripts to do things like it many times, usually on a batch of a few thousand documents at once. The ‘ is easy: just replace all instances of ‘ with nothing. Alas, if you then try to do the same with ’, you accidentally turn Bob’s hat into Bobs hat and actress’ family into actress family. Now of course if you are editing manually, you just remove the ’ from ''‘Madhubala’'' and not from Bob’s hat or the actress’ family. But how to do it with automation?

Here is where Rich Farmbrough appears to have made a programming error. After batch-replacing ‘ with nothing, he batch-replaced ’'' with ''. Alas, sometimes the second ’ doesn't have a '' after it, and so the programming error turns ‘Madhubala’ into Madhubala’ and ‘I got it’ into I got it’.

We know this was automated, because a human who is moving through text manually removing quotation marks simple does not remove multiple ‘ characters without removing the matching ’ character, but only when the ’ is not followed by two '' characters. There are mistakes that humans make and there are mistakes that only automated editing tools make, and this was the latter.

Also, missing the giant red error on the page indicates that he probably never read anything but the wikimarkup, and the edit summary ("Italics for book titles, not bold.") which confuses ''‘ with ''' tends to support this theory. --Guy Macon (talk) 10:58, 21 June 2013 (UTC)

I'm trying to follow this explanation in terms of the rest of the text, particularly the "Eureka" portion at the end. "And the way he had twisted the title ‘Eurek(h)a’ of the book" becomes "And the way he had twisted the title Eurek(h)a’ of the book", and "shamelessly borrowing the joyous quote ‘Eureka’ which meant" becomes "shamelessly borrowing the joyous quote Eureka’ which meant". But in the same section, "Mohan Deep got 'Eurekha!' vetted" and "Rekha was to talk about 'Eurekha' 4 years later" do not have the same error introduced.

For anyone trying to follow the "red error" issue, a red "cite error" in the main text is first introduced here [3] (not by Farmbrough) on 06:31, 22 March 2013, and another smaller red "unknown parameter" in the references section here [4] at 05:03, 22 March 2013. There are eleven intervening edits before Farmbrough makes the disputed edit. Ten edits after Farmbrough's edit, the article is moved from "Articles for creation" to article space [5] (again, not by Farmbrough) on 11:05, 23 March 2013‎ with both red errors still in it. The first red error (in the text) is eliminated here [6] at 11:09, 23 March 2013, about 13 edits after Farmbrough touched it, the second red error is finally removed, three weeks later, here [7] on 06:06, 15 April 2013, when the page number is added to the citation. Farmbrough made some 13 edits to the article before the error was removed, there were 85 edits by all editors altogether that left a red error still in the text. Neotarf (talk) 13:38, 2 July 2013 (UTC)

• I will keep my comment brief but as I have stated in other venues in the past, the blocking of Rich and his bots over minor edits was a decision of monumentally poor judgement. Even more so given his eventual banning from the project was due to the extrmely poor wording of his sanction. Sepcifically the decidedly poor use and interpretation of the terms "broadly construed" which allow an administrator unlimited discretion. An editor being banned over little more than a difference of opinion about the types of edits that should be done is nothing short of just plain dumb. Every month Rich and his bots aren't editing is directly equivelant to tens of thousands of useful edits not being done, setting the project back in an unmeasurable amount. Kumioko (talk) 13:55, 24 June 2013 (UTC)
  • "Broadly construed" points both ways--it's easy to see why people who like to Wikilawyer hate it, because it's essentially IAR applied to sanctions: in any sanction that includes "broadly construed" it is very explicitly stating that the spirit counts, not just the letter. ArbCom tends to use a lot of "broadly construed", because people who are able to find consensus at lower levels of dispute resolution don't make it to ArbCom cases. Jclemens (talk) 02:44, 28 June 2013 (UTC)
    • But your assumptions are that the wording assumes that the editors is guilty and that every one of the 1400+ admins who could block that user, would do so fairly or has the same understanding or interpretation of the rules. Unfortunately that is often not the case. It doesn't have to do with wikilawyering and fankly I only see that as an excuse by those who support a system of making it easy for admins to eliminate editors without due process. Broadly construed measn that any interpretation of the judgement is justified. It has nothing to do with wikilawyering and everything to do with fairness and common sense. Kumioko (talk) 17:05, 2 July 2013 (UTC)

Postscript. Several comments above have raised questions about neutrality in terms of the format of this arbitration report. As I said there, and will say again, comment was invited from both Mr. Farmbrough and the arbitrators. I also promised, somewhere upthread, to try to find time to comment about this at greater length, so I will say something more now about the writing of this report, and the departure from the usual arbitration report format.

When we first started discussing the possibility of doing an interview for the arbitration report, I knew immediately that I wanted to interview Farmbrough. When a disruptive user is indeffed, there is often a sigh of relief that goes up from the community. But with Farmbrough, that wasn't what happened. Instead, there was a whole dialogue that started up about process--the process surrounding arbitration enforcement and bot development, and people stated bringing up points about the structure of WP. In fact, the working title of this report was "The Farmbrough amendment request - a closer look at automation and arbitration enforcement".

While I was putting the report together, I was also trying to decide whether Farmbrough was a hero or a villain. But you would probably have to edit in the same area with someone to really answer that question, and as far as I know, I have never had any interaction with Farmbrough on WP at all. And I suspect the answer to that question is not that simple. I also didn't look at any of the diffs for any of his cases until after I had submitted the report, so whatever I wrote came directly from discussion about the current case, and not from trying to judge for myself what had happened.

My primary interest in writing this was to find out something about how WP breathes and grows, and sometimes hiccups, in the context of real users, not in the abstract. But if the comments here are any indication, most readers are not interested in the meta-type issues I tried to bring out in the report. They are interested in Farmbrough himself; and they are not finished with talking about what happened.

Neotarf (talk) 11:59, 2 July 2013 (UTC)

Simply inviting comment isn't really enough for neutrality. In addition you've cherry-picked a few heavily biased "anonymous" comments, but didn't invite comment from any of the other users mentioned or anyone involved in the case beyond two of the arbs.

Regarding "When a disruptive user is indeffed, there is often a sigh of relief that goes up from the community. But with Farmbrough, that wasn't what happened.", that's not really accurate. Usually a sigh of relief goes up from part of the community while the disruptive user's supporters (their "fan club", although that term sometimes has a derogatory connotation) complain about every aspect of the situation. Sometimes there are not many supporters or they drop the issue quickly. In this case there are a fair number of them and the list includes some extremely vocal and tenacious members. In the end it comes down to trying to draw the line for how much disruption can be excused by good editing; look at some other cases where a disruptive editor also does good work, both where the user wound up banned and where they didn't, and you'll find the same thing.

Regarding "most readers are not interested in the meta-type issues I tried to bring out in the report", I for one don't see any meta-type issues in your editorial. When the whole piece is focused on the one editor and to a lesser extent his arb case, it should come as no surprise that people discuss that editor and that arb case. In a good case study, the meta issues are still clearly presented and are the real focus of the work. Anomie⚔ 13:19, 2 July 2013 (UTC)

But this isn't a case study, and I am not presenting issues, or a point of view. Since my role here is to be a neutral observer, or at least to be equally detested by all parties, I can only point out issues that have already been raised by others. In a perfect world, PR specialists would give me pithy and quotable sound bites that were all understandable, and spelled correctly. But this is Wikipedia, in all of its chaotic glory, and even if I would like to make something up, I am stuck with what people have actually said. For example, I was particularly puzzled by several comments that Wikipedia had become a "personal fiefdom"; it was not clear how this might apply to the automation group. Also, I never did receive an answer to where someone could go for assistance if they had a problem with a bot. As far as I can tell, they take it to Farmbrough's talk page, where he still dispenses technical advice on occasion. And if you check again, I think you will find that arbitrator T. Canens did adequately explain the arbs' decisions (and the issues you brought up), even if he wasn't involved in the original case.

If you really don't see the procedural issues, I would suggest you go to the two talk pages linked in the report, and read the comments yourself. For those who automatically discount anything that isn't written by an admin, there are comments there by admins as well. There are plenty of editors there, who have taken the lead in trying to define and resolve some of the issues, instead of just complaining that no one else has resolved anything to their satisfaction.

I sympathize with anyone who has ever been irritated by someone else's editing, but clearly this is a complex situation. And it is likely to come up again, quite possibly with the same issues still unresolved. I too wish I had been able to stumble across some easy answers, wrapped up in a neat package, but this is Wikipedia: what you see is what you get. Neotarf (talk) 15:40, 2 July 2013 (UTC)