Wikipedia talk:Arbitration/Requests/Case/Betacommand 3/Evidence

Arbitrators active on this case
Active:
 * AGK
 * 1) Casliber
 * 2) Coren
 * 3) Courcelles
 * 4) David Fuchs
 * 5) Elen of the Roads
 * 6) Jclemens
 * 7) John Vandenberg
 * 8) Kirill Lokshin
 * 9) Mailer diablo
 * 10) Newyorkbrad
 * 11) PhilKnight
 * 12) Risker
 * 13) Roger Davies
 * 14) SilkTork
 * 15) SirFozzie

Inactive:
 * 1) Cool Hand Luke
 * 2) Xeno

Recused
 * 1) Hersfold

A question about diff count
I see that mine is 0. But the talk page and AN thread that I linked to in my evidence section do include diffs though. Should I copy them here? ASCIIn2Bme (talk) 11:15, 3 November 2011 (UTC)
 * Doubt it's a problem. I have to assume that it tells if it's a diff by reading for timestamps in the URL, but you're citing a whole archived (thus static) discussion, so there are not timestamps. The clerks or Arbs will inform you of any problems, if you don't hear from them, you're fine.  S ven M anguard   Wha?  11:35, 3 November 2011 (UTC)
 * Okay, thank you. ASCIIn2Bme (talk) 11:37, 3 November 2011 (UTC)

Questions to the parties
In addition to the notes left on talkpages of editors who are either parties to the case, or have presented evidence, this serves notice that drafting arbitrator User:Kirill Lokshin has posted some questions to the parties on the workshop page. --Alexandr Dmitri (talk) 13:54, 9 November 2011 (UTC)

Replies by Fram to evidence

 * Reply to Toshio Yamaguchi: I am basing my opinion on many more than these diffs alone, obviously. In that section, I list 6 types of unnecessary changes, and 6 types of errors. In a followup post, I listed more problems. In other discussions, I have listed yet other examples, and also examples of where his attempted correction of these errors didn't result in an actual correction. For the first diff; how is duplicating the same info at the same spot not making an article worse? What does the added "Droban-Apherna | Facebook" actually add to the article or the link? It's a script-based copy of the link title, not a thoughtful edit. For the second one: in Delta's version, a link leads you here, while the pre-delta version leads you here, to the exact page that is used to source the information. How is it better to go to the front page of a bok than to the exact page we want? It's only one click from the correct page to the front page if you want that anyway, but it's a lot more complicated to get from the front page to the exact page. Any opinion on the other errors I pointed out?
 * By the way, fixing someone else's errors is only one possible way of working collaboratively. Detailed explaining of where there "manual edits" go wrong, with many diffs, is also a form of collaboration. It's not as if I went to his talk page and said "you make many errors" without any elaboration. That would indeed be an unhelpful comment, even though it was a correct one, and in that case I could correctly be said to refuse to work cooperatively.
 * Furthermore, I have explained this to you before, after which you apologized, but apparently I need to do this again: indicating errors and problems, whether I am right or wrong in seeing them as such, is not an assumption of bad faith or a violation of AGF. I have not indicated any opinion on why Delta makes these errors and edits, and have not assumed any bad faith in those diffs you present. I have every right to propose or support restrictions on the editing of any editor, if I believe that these edits are problematic, for whatever reason. That is not an AGF violation at all, no matter if you believe that I am right or wrong. Fram (talk) 11:51, 15 November 2011 (UTC)


 * Further reply to Toshio Yamaguchi: Citing sources. "When a book is available online through a site such as Internet Archive (pre-1923) or Google Books (post-1923), it may be useful to provide links to the relevant pages, using external link syntax, so that clicking on the book title will take the reader to the page in question." Fram (talk) 12:37, 15 November 2011 (UTC)


 * Reply to : I don't think it is useful to continue this discussion if that is the kind of argument you need to defend his actions. He is removing links which a guideline explicitly considers to be potentially useful. Can you perhaps explain how his removal of these direct links can be considered improvements? Fram (talk) 13:35, 15 November 2011 (UTC)


 * Reply to Dirk Beetstra: as indicated in the lengthy workshop discussion about this, I don't believe that the reason he made these edits is relevant here. Furthermore, the reason Delta did give eventually does not match the facts. My reply, which you so charitable describe as "Fram shows lack of understanding", indicates this clearly. Finally, he ran a bot script, had to revert it, and ran it again. What has presenting this to do with "failure to assume good faith"? He may well have ran it in good faith, believing that for some reason, running a bot script in such a way that it had the appearance of a bot edit would not be a violation of his editing restriction. It does raise questions about his ability to continue editing, when he doesn't even follow his restrictions while under ArbCom investigation, but I have made no statements about why he did it, and have not assumed bad faith. Pointing out errors or violations is not a lack of good faith anymore than e.g. correcting an article edit is showing a lack of good faith. Fram (talk) 15:19, 15 November 2011 (UTC)


 * Are you also here implying that $\delta$ is lying about the reason for doing the edit? (answering here, after all this is a talkpage)  --Dirk Beetstra T  C 15:22, 15 November 2011 (UTC)


 * Hurrah, I have just answered your bad faith question at the Workshop page, to find the loaded version of it here ("also"?). This should be sufficient. Fram (talk) 15:26, 15 November 2011 (UTC)


 * @Beetstra: I am not sure which edit you are referring to. But I also don't think that the reasoning behind an edit is germane to whether the edit violated his restrictions. The restrictions only refer to actual edits, not to reasons behind them nor to what the edits are intended to accomplish. &mdash; Carl (CBM · talk) 15:30, 15 November 2011 (UTC)


 * @Carl. Is it possible that that edit of  this morning (my time) to Wikipedia was fully intentionally, supervised to check a problem that existed, using a manually (as opposed to fully automated) run script, within the levels of his civility restrictions and speed restrictions to solve a problem?  --Dirk Beetstra T  C 15:39, 15 November 2011 (UTC)
 * I have no doubt it was intentional. It was not supervised, or at least not adequately supervised, because he had to revert the edits. As there were only four edits to enwiki, it was within his speed restrictions. The edits certainly had the appearance of being made by an automated script, as they were in fact made by the interwiki.py script.


 * The deeper question that we need to start thinking about is how these things reflect on 's judgment, temporarily putting the edit restriction aside.
 * could have made these fixes by hand rather than by using interwiki.py. It just requires making a list of the correct interwikis and then editing the articles one at a time to put this list in. In other words, if the goal is to fix the problems with just one article across several languages, there is no need to allow the script to save any edits at all, it can be done with copy and paste. If the edits re being manually reviewed it is not particularly slower to do them with copy and paste.
 * could have reviewed the edits (more carefully) to ensure that, if they were saved, they would not need to be reverted.
 * could have waited until after the arbcom case to make these changes.
 * It is perfectly legitimate to consider the message the he sends by making a sequence of broken edits using interwiki.py while this arbitration case is open.
 * &mdash; Carl (CBM · talk) 15:54, 15 November 2011 (UTC)


 * Regarding the options:
 * Hmm .. but that would not have shown the errors made by the script
 * Maybe, I don't know enough of the inner workings of interwiki.py whether that would, again, have shown the errors made by the script
 * Maybe, but then the errors made by the script would have stayed .. unless someone else would have fixed them.
 * Another option: or the edit was fully supervised, intentionally to see where the mistake was, and maybe the error was not obvious (actually, I had to look twice to know why the first edit is actually wrong ..) - it was then reverted (which suggests that it was supervised, as the error was found). It is all again on the edges of the restrictions where the edges are too vague - edits may not look like they were made fully automated (with the vagueness that I can make automated edits which don't look like that, and I can make manual edits which look so much the same that they could be automated, what, maybe should be automated), edits should be fully supervised (still, that they are utterly wrong does not mean that they were not fully supervised - is a broken edit which is then repaired not supervised).
 * And calling this a 'sequence of broken edits' .. 1 broken, one to fix and then 2 proper edits. --Dirk Beetstra T  C 16:19, 15 November 2011 (UTC)
 * The error in is pretty clear: why is "FARC" there?  If he is not looking for that sort of thing, he is not supervising his edits in any meaningful sense.


 * It seems to me you are now claiming that was actually trying to debug the interwiki.py script. As a bot developer I can say that the approach  used is not the approach we actually use to find bugs in a script - if we suspect bugs we don't allow the script to save, we just look at the diffs to see what the error is. But I don't believe  is a developer for Pywikipedia, in fact he claimed he was unable to even change the edit summary. And I don't believe there are bugs in the script, the problem  alluded to was incorrect interwiki links on the articles that cause articles on different subjects to be identified.


 * The "sequence of edits" is that he did the same thing on 20 other languages, to the corresponding articles, as you can check on . For example the edits to eswiki, etwiki, and euwiki are separated by only one second (19:55:24 UTC to 19:55:25 UTC to 19:55:26 UTC).


 * You have not responded to what I think is the more important point: what reasonable editor would think that now is the right time for to do this sort of thing? &mdash; Carl (CBM · talk) 16:32, 15 November 2011 (UTC)
 * IMHO, the error is "why does interwiki.py put FARC there". Why did interwiki.py add FARC to en.wikipedia where it obviously is not a correct interwiki.  Where is interwiki.py getting the idea to put FARC there.  Thát is the error that needed to be evaluated.  Is there an error in interwiki.py?  Is there somewhere else an error?  For that is
 * I agree on that part, and as I said, I don't know how the logging and debugging in python works, where does it tell you where it found what and if it did that. I would expect that this is the type of error you might be able to figure out from looking at all the pre-save diffs (note that I don't believe that you can get everything from that, I've seen XLinkBot and COIBot give errors where the problem is not in the 'to save' page, but happens upon saving (e.g. encoding problems)).  The whole point is, that it is blatantly obvious that assumptions were made first, and only in hindsight you know.
 * This ArbCom has no jurisdiction outside of en.wikipedia. What this case regards, there is only one broken edit -  can break whatever he wants outside of en.wikipedia, what this case regards those edits don't even exist.
 * Well, you seem to suggest that I am unreasonable. I do not think it is unreasonable for any python/wikipedia expert to try and help out another editor.  The edits may be close to breaking restrictions, but they are not unquestionably a breach of restrictions.  --Dirk Beetstra T  C 10:52, 16 November 2011 (UTC)
 * The reason that interwiki.py put FARC on there is that some other language's page had an interwiki to FARC. In this case, before 's edits, at least ruwiki did - the arwiki inerwiki from the that link goes to FARC, as you can see from its dawiki link. This is a common problem with interwiki data, it is not a problem with interwiki.py.


 * claimed he was not able to even change the edit summary in interwiki.py. That goes against the idea that he is a python expert. I think you are stretching to extremely unlikely explanations to try to excuse 's behavior here. WP:AGF does not mean that we need to eliminate every other possibility (perhaps it was a computer virus?). We know that he ran the script, he has even said so; we do not need to worry about why he ran it in order to see that it didn't work properly.


 * Even if the edit was barely within the restrictions, there is still the problem that that decided that now was the right time to do it. That speaks to a lack of judgment. It's the same lack of judgment that  has been demonstrating for years, and the fact that he continues to do it during an arbitration case supports my point that he will not be able to improve on his own. &mdash; Carl (CBM · talk) 12:42, 16 November 2011 (UTC)


 * It is interesting, here, that, as it suits your argument, you go to the "reason for the edit", whereas elsewhere you say it is not germane. Rich Farmbrough, 13:28, 18 November 2011 (UTC).


 * It is not germane, I am just explaining to Beetstra what actually happened, because not everyone is familiar with how the interwiki system works and in particular how interwiki.py chooses which links to add. Beetstra was trying to argue that was actually debugging interwiki.py, but that is at odds with all the evidence, which includes the fact that interwiki.py was acting correctly. &mdash; Carl (CBM · talk) 13:36, 18 November 2011 (UTC)


 * Yes, and we can stretch this on and on an on. But we do get back to basic assumptions and that type of questions, and we are not going to convince each other or other users (not the aim here anyway).  Moreover, I see around also the same discussion lines coming over and over as well.  --Dirk Beetstra T  C 13:48, 16 November 2011 (UTC)

Replies by Toshio Yamaguchi regarding replies by Fram to my evidence
Reply to : First, it would make things easier if you wouldn't move your comments around when I try to reply to them. Second, you say "For the first diff; how is duplicating the same info at the same spot not making an article worse?" As I see it he changes an unnamed link to a named link. You say this makes the article worse, I disagree. I do not see how that is worse than an unnamed link. Regarding your second example, which policy or guideline says we must link to the exact page of a source? The change he made might be controversial and I personally would rather like to have a link to the exact page than to the title page, but it is not prohibited by any policy or guideline to make this change and book sources are not required to be online anyway, so the assumption that this edit makes the article worse is absolutely subjective and not based on any actual policies or guidelines. This is a reason to create a guideline that says editors should link to the actual book page if it is available online rather than to the title page. Can you show me such a guideline exists? I am not aware of it. Toshio Yamaguchi (talk) 12:26, 15 November 2011 (UTC)

Reply to : "It may be useful to provide links to the relevant pages." is as good as saying "It may not be useful to provide links to the relevant pages." This just says nothing. Toshio Yamaguchi (talk) 13:20, 15 November 2011 (UTC)

Reply to : Incorrect. Book sources are not required to be accessible online at all, so changing the links to the online accessible versions of the sources does neither improve nor detoriorate the article. Toshio Yamaguchi (talk) 13:46, 15 November 2011 (UTC)

I changed the last word in my reply above from 'source' to 'article' here, as this is what I meant. Changing links obviously has no effect on the source itself and I accidently used the wrong word. Toshio Yamaguchi (talk) 12:21, 16 November 2011 (UTC)

Sven Manguard's reply to evidence presented by 69.149.249.38
The IP's evidence, showing a list of times that Delta went over the edit throttle restriction (he is able to make 40 edits in a 10 minute period of time), is telling of one of the core problems with the current sanctions. Yes, edit counts in the 80s, 90s, and 100s are, if they were not authorized ahead of time, clearly problematic. However, I would note that of the 35 items listed, only a small minority were clear cut cases of the restriction being ignored. The IP, however, chose to list times where he edited 42 or 43 times in 10 minutes, an indication that he or she is completely ignoring the spirit of the restriction in order to hammer Delta on the letter of the law. Wikipedia does not work like that, IAR exists for precisely this reason. I count 14 times that Delta went over 50 edits, at which point Delta is clearly pushing it harder than he should (and which would be the point at which I'd warn him to slow down). I'll also note that most of the violations happened on the same day. Three of the largest counts on that evidence list all happened in a two hour period on 2011-03-19. Seven of the bulleted items happened on 2011-05-11, including six of the fourteen major violations. The IP hasn't found a pattern of abuse, he's found, at most, a dozen or so bad days over the course of a year.  S ven M anguard  Wha?  07:48, 16 November 2011 (UTC)

P.S. Far, far to much has already happened in the Delta affair for me to believe, even the slightest bit, that this IP isn't an established party in the case posting in anonymity. I have no idea which one though, and I won't pursue the matter further because of that.


 * IAR does not apply to edit restrictions - that would be bizarre. The spirit of the restrictions was that should never have gotten close to making 40 edits in 10 minutes, but he has ignored the spirit of the restrictions, as I document in my evidence. There is nothing in the spirit of the restrictions that would allow  to break them as long as he only does it 12 days a year. In the end it was up to him to follow the restrictions, it's not up to us to make excuses for him when he doesn't. (By the way, even pre=authorized edits are still subject to the speed restriction.) &mdash; Carl (CBM · talk) 12:34, 16 November 2011 (UTC)


 * IAR does apply to if the community chooses to enforce the restriction; even if it is an objective measure such as this edit rate doesn't mean a block is automatic. (This is the point I believe Sven's making on that). Obviously it has not happened, but if edited under the rate for a full year except for one period where he went over to 41 edits / 10 minutes, and this wasn't discovered until a year later, someone would want a block placed on him.  That may be extreme, but some of his blocks in this year were for far-removed edit rate breaks.
 * Now, granted, were I in 's shoes, knowing that I'm going to be watched like a hawk, I'd be editing at a purposely slower rate until I'm sure my timing's down right, and even then take caution to stay a few ticks away from 40. But to come back around, he is trying to abide by the restrictions with some mistakes (this reflecting one of the clarifying questions asked: is  willfully ignoring the restrictions? ) --M ASEM  (t) 14:07, 16 November 2011 (UTC)
 * My 2p is that I don't see any sign he is trying to stay within the restrictions. I could agree to that interpretation if he violated them once, was warned, and stopped violating them. But the evidence shows that he has been warned over and over, and blocked over and over, and he continues the same things. That pattern is not really compatible with an editor who is making an effort to improve. In other words, if this is what happens when tries to follow the rules, then it is not clear whether he can ever be a productive editor. But I think it is more likely that he simply was not trying to follow the rules. &mdash; Carl (CBM · talk) 14:23, 16 November 2011 (UTC)
 * IAR is bizarre, and it applies to everything. The Wikipedia rules system is not a formal logic, or a legal framework, but a codification of practice. As such it is constantly changing (and, unfortunately, growing) incomplete and self-contradictory.
 * IAR is not needed here for most of the items, I agree with Sven that this is really "one day, one task" Δ went over the top, while the rest could have been legitimately brought to his attention this one is more extreme. The edits in question on the "extreme" day were so trivial (at least those I have looked at) to accomplish that expecting Δ to spend more than a fraction of  a second verifying each of them is absurd. Note that there were no issues brought to his talk page over these edits, the exercise seems an unmitigated win for the encyclopaedia. Rich Farmbrough, 12:01, 18 November 2011 (UTC).


 * It would be bizarre for an editor who has been sanctioned to be able to say "I can ignore my sanction when I see fit" just as it would bizarre for a banned editor to be able to appeal to IAR and edit anyway. The consequence of a community sanction is that the editor in question is required to follow it. When the sanctions are established, they establish a consensus that the forbidden types of edits are not an improvement to the project, and in fact are sufficiently unwelcome that the editor is forbidden from doing them. &mdash; Carl (CBM · talk) 12:42, 18 November 2011 (UTC)

Replies by Toshio Yamaguchi regarding evidence presented by 69.149.249.38
Reply to 69.149.249.38: The restriction clearly says "Betacommand must not average more than four edits per minute in any ten minute period of time." Unless it can be shown that the Average of the number of edits is over the limit, this evidence does not support the assertion that he has violated the speed restriction. Toshio Yamaguchi (talk) 09:23, 16 November 2011 (UTC)
 * Assuming that the evidence presented is correct, it actually does support that assertion: it always indicate the ma number of edits in a ten minutes period, and in every case that max is above 40, so the average is above 4. Not by much in most cases, and most of them are not very recent, but this is not some vague restriction with grey areas, this is a clear one, which he violated on these occasions. Whether these (mostly) minor infractions are a problem or not is a different discussion, but I don't believe there can be any doubt that if the evidence presented is correct, he clearly violated the fairly precise letter of that restriction. Fram (talk) 09:35, 16 November 2011 (UTC)
 * The restriction does not prohibit him from making more than four edits in some ten minute period of time, as long as the average does not go over the limit. Unless it can be shown that the average was over the limit (which the evidence presented by the IP does not), this is not a proof that Δ has violated the restriction. Your reading that the average is the total number of edits in some single 10 minute period of time is just your personal interpretation of the text of the restriction which is not evident from the text of the restriction. Toshio Yamaguchi (talk) 09:45, 16 November 2011 (UTC)
 * I'll let some uninvolved people explain this to you, perhaps you'll believe them, but your interpretation of this restriction is very hard to follow, unlike the actual restriction. Fram (talk) 10:06, 16 November 2011 (UTC)
 * Toshio Yamaguchi, I'm just as critical of this evidence as you are, but in all of those cases, Delta did average more than 4 edits over a 10 minute period of time (i.e. he had more than 40 edits in 10 minutes). Mind you, I make the point above that counting 42s and 43s against him seems, well, more vindictive/hostile than it is logical/constructive, but they are, by strict letter of the law violations.
 * As for Fram's comment about the restrictions being easy to follow, I challenge you to try those restrictions out yourself for a few days. Unless you're constantly looking at a stopwatch, getting into the 45s and 50s without meaning to is easy.  S ven M anguard   Wha?  10:45, 16 November 2011 (UTC)
 * I mean that the text of the restriction is easy to follow, i.e. easy to understand: stay below 40 edits in a ten minute period. Things like a "patetrn" or "appear to be automated" are more prone to discussion and have more of a grey area. The 40 edit restriction may be easy to abide by or easy to neglect / violate / pass as well, but my (in hindsight) ambiguous statement was not intended in that way. Fram (talk) 10:51, 16 November 2011 (UTC)
 * Ah, yes, it is one of the more clearly worded restrictions, although strangely that might be part of the problem. Too clearly worded and it can be abused by people that are willing to throw the book at someone for 42 edits. Too loosely worded and you get a whole different type of mess.   S ven M anguard   Wha?  11:27, 16 November 2011 (UTC)

From User talk:Sven Manguard Sven, it would lead to far to continue this discussion on the Arb Case page, but if I were to have such a "speed limit" restriction, I would no longer make any AWB edits, and you would be very hard pressed to find many other times when I edited more than 40 times in 10 minutes. The most I could find for November was 22 edits in ten minutes, and for October 24 edits in ten minutes. So yes, in reality it would be fairly easy for me to follow that restriction; just stop using AWB. Fram (talk) 11:12, 16 November 2011 (UTC) End cut and paste
 * Without checking, I probably have quite a few pockets of rapid edits, in the 40s or 50s over a ten minute period of time. It's a function of gnome work; when you hit a pocket of 'somethings' that need fixing, and it's a simple change, you can end up with a very high rate of edits over a short period of time, without the use of scripts.  S ven M anguard   Wha?  11:24, 16 November 2011 (UTC)
 * Right, but if you were under the restriction you would just avoid doing those sorts of things. I disagree when you say "abused by people that are willing to throw the book at someone". The blocks for violating the speed restriction were properly using it to prevent from causing problems. We may grumble when we get a speeding ticket for driving slightly too fast, but it's still a valid ticket. &mdash; Carl (CBM · talk) 12:38, 16 November 2011 (UTC)
 * My point, to pick up your analogy, is that police don't give speeding tickets for driving 38 in a 35 MPH zone, they give tickets for driving 48 in a 35 MPH zone. I'm not saying that we should overlook it when he makes 60 or 80 or 100 edits in 10 minutes, but I am saying that when he does 42 or 43 in ten minutes, we shouldn't get so uppity about it.
 * One of the great ironies of the Delta situation is that Delta has been dragged out into the courtyard for a beating so many times over very minor offenses that when the major offenses do come along, a) the beatings for minor offenses are pointed to in order to illustrate a pattern of harassment, b) large portions of the community will write it off just by seeing Delta's name in the thread, simply because the last five times his name was in a thread it ammounted to nothing, and c) the people who are left, and haven't burned out too much to get involved, are the most entrenched, partisan users for and against Delta.
 * If a judge keeps getting cases where she has to adjudicate over tickets for people driving 38 in a 35 MPH zone, she's going to decide that the police officer, not the people with the tickets, is the bigger problem. The same applies here. Stop sounding the alarm for the small stuff and a) Delta will be able to demonstrate that he can be constructive when he's not being chased, and b) if Delta does something major, people might actually pay attention.  S ven M anguard   Wha?  13:41, 16 November 2011 (UTC)
 * If cannot be constructive when he knows his edits are being scrutinized, I don't think we can expect better when they aren't. The question we should be asking is "why did he even get close to 40 in 10 minutes?", not "is 42 enough to block, or is 45?". He could have stayed at 20 edits in 10 minutes and made the problem disappear, he has been warned many times, but he has continued to push the limits. That is the key point we need to address in this case, I think. &mdash; Carl (CBM · talk) 13:39, 16 November 2011 (UTC)
 * I think all this reflects that this specific CR does not best serve the goals of the CR, to get Delta to edit like a human and taking responsibilities for those edits. It is a prevention against "disruption", yet if all, say, 41 edits in 10 minutes are in consensus, there will be more disruption at ANI reporting that block and the subsequent discussion than the actual edits themselves.  Granted, edit rate is a measure that we can use to determine if s doing thing automatically or humanly, and flagging his edits when they go over a specific rate makes complete sense. We should be looking at the end result before considering the edit rate: did all the edits in said period have consensus, have proper edit summaries, didn't break anything else, and the like?  If there's a "no" among those, we can then examine the edit rate and make a determination easily that  wasn't editing like a human at that time.  I think there's a better way to place a cap on 's activities where in case he is otherwise editing at a slow manner but introducing problems/performing non-consensus actions, the damage is minimal, but it's not an edit rate, at least, as defined by CR#3. --M ASEM  (t) 14:19, 16 November 2011 (UTC)
 * I agree - the CR has not been effective. Why? I think because it is too loose. We assumed would make an effort, so we only tried to limit things that he was expected to avoid doing anyway. But  viewed the restriction as a limit that he could go right up to (and beyond) instead of a line he should stay well inside of. I think the solution is to write a tighter restriction that is designed to be treated that way. This could include a specific ban on all NFCC work along with tight enough editing restrictions to force  to abandon the gnoming edits that consistently lead him into trouble. &mdash; Carl (CBM · talk) 14:33, 16 November 2011 (UTC)
 * You are missing the point. We need restrictions that allow  to show that he can edit like a human, communicating in a civil manner, taking responsibilities for his edits, and making sure that when he does something at a larger scale that it has consensus first, and that he does these without disruption. Tightening the existing ones down even further do nothing towards that, in part that they aren't serving either the community well (when we spend more time arguing over the results than the actual actions  did) nor giving  the opportunity to reform.  If we are asking him to edit like a human, we have to recognize that humans can both edit faster than the rate given, and are prone to mistake.
 * Ignoring the NFCC aspect and just focusing on wikignoming, if you called for a block on him doing this, this would effectively force him to work on main space articles, which isn't necessarily bad, of course. But what then suddenly can happen is if, say, adds a reference and then in a second edit immediately afterwards, adds a name to that reference, or adds a missing title, or any other activity that would otherwise fall into the wikignome category.  Someone will try to bust him on that.  That's how much animosity there is for him.  This is why these current restrictions, and even what you suggest, will fail - because they aren't aimed at the larger behavior of his editing but the minutiae.  It's easier to make evaluations on the minutiae towards blocking, but that only serves to make it a game to both sides.  We should be asking how we can encourage  to participate in a more human manner, and how to measure that at a broad scope, not in fine details like edit rates or the like.
 * At the same time, however, if we were to device a broader approach to restrictions, I would favor a one-strike remedy on this: first community-affirmed violation would be a block of, say, a month or so, and second time would be it, banned from the project. But again, this requires that the restrictions are not based on minitiae that can be gamed.  For an example, if the community/ArbCom decided that Delta shouldn't even be using an offline script (much less direct automated tools), and we find he's editing at sustained fast rate for an hour or more, we know something's up, and the community can decide if that really is an automated tool.   (Note, this is an example, I am not suggesting this is necessarily a finding).   It's looking at the larger-term patterns to determine if  is in compliance with expectation, not what happens minute to minute. --M ASEM  (t) 14:55, 16 November 2011 (UTC)
 * I think that the proposal for a limited number of distinct pages edited per day (like 10), but no limit on the number of type or number of edits to each page, would be objective and would be less subject to gaming either way. That, combined with a tight NFCC topic ban, seems like a reasonable way forward. &mdash; Carl (CBM · talk) 16:22, 16 November 2011 (UTC)
 * There is something to be said along the lines of unlimited edits on a limited number of article pages as a fair balance, with the understanding that can request raising that limit after some period (3 months? 6 months?) to the community and/or ArbCom depending on that resolution. I would still think that if he's doing wikignoming there should still be something like the first CR, but more explicit based on the Hammersoft VPR thread. (eg: still gain approval for specific wikignoming tasks at VPR, but also make sure these are documented with discussion links on a user subpage, and provide clear editsummaries possibly pointing back to that subpage to explain what has been done).  Also, I would make sure there are allowances that if he realizes or someone points out that he made a non-obvious error some days later, he can revert his changes on old articles without counting towards that article count.  There's a lot of small specifics to work on this idea of unlimited edits on limited articles, but I do agree that's a fairer way to approach what the community wants out of . --M ASEM  (t) 16:48, 16 November 2011 (UTC)
 * "If Δ cannot be constructive when he knows his edits are being scrutinized" The point is that, if anything, Δ refuses to stop being constructive, just because a bunch of rules lawyers are watching him. I'm not sure if this is the case, or if he simply does not give arbitrary restrictions the importance of improving the project and hence sometimes transgresses.  Regardless if his editing is not causing problems we should not care about "rules".  If it is causing problems we need to look at the problems and only if enforcing those rules would improve the encyclopaedia should we enforce them. Rich Farmbrough, 12:13, 18 November 2011 (UTC).


 * The community consensus that established the sanctions established that the forbidden edits are not constructive, and are in fact so undesirable that in fact he is forbidden to do them. Hence cannot use the claim that the edits are constructive (even if he happens to think they are) to justify violating the restriction. In this case have a decision by the community already about these edits, which is not true in most cases, because most editors are not under editing restrictions. &mdash; Carl (CBM · talk) 12:50, 18 November 2011 (UTC)
 * No it didn't. It merely established that at the time that there was consensus to restrict his editing in a specific way among the then denizens of whichever noticeboard was responsible. It is a gigantic leap from that to condemning the entirety of his editing, from which, it is argued, some (possibly vanishingly) small proportion may be in technical breach of an ill thought out restriction. Rich Farmbrough, 12:31, 21 November 2011 (UTC).

Sven, re "Without checking, I probably have quite a few pockets of rapid edits, in the 40s or 50s over a ten minute period of time."−I checked, and you have never made 40 mainspace edits in 10 minutes under your current account (something Δ has done many times). There were two sessions when you made 40/10 counting all namespaces, one around 2011-05-01 03:18Z and one around 2011-07-10 04:57Z. It doesn't look like careful editing to me. For example, here you left a level 2 vandalism warning after reverting what looks like a perfectly good edit. Here you left vandal1 when the edit you reverted was obviously test1. Here you reverted with the huggle boilerplate "unexplained removal of content" when the person's edit summary contained the explanation, "This does not belong at the top of the article". I quit examining at that point when it became obvious that I'd keep finding more of the same. I'd urge you to slow down with your own editing, rather than trying to defend Δ's speed on the grounds that yours is comparable. You are both way too careless. 66.127.55.52 (talk) 06:11, 4 December 2011 (UTC)

Q for arbitrators
The evidence target date is today, the 17th. Is there any other evidence you would like someone to dig up? &mdash; Carl (CBM · talk) 03:04, 18 November 2011 (UTC)


 * I anticipate asking questions of individual parties over the next week or two; but, at this point, I would suggest not presenting additional evidence unless specifically requested. Kirill [talk] [prof] 17:58, 19 November 2011 (UTC)

Reply to errors (sorry, "evidence") by Rich Farmbrough
For some reason Rich Farmbrough has a dislike of adding diffs to evidence, which is not really helpful of course and makes it necessary to take his claims with a grain of salt. Anyway, he dismisses all my evidence based, accordingto him, on one "random" example, which he believes is wrong. He claims that the book "War and Revolution in Yugoslavia 1941-1945: The Chetniks" is from 1975, just like Delta claimed all along, and that my claim that it is from 2001 is incorrect. He is right that the book is from 1975, see e.g. here.

However, and this is a very big however: the article where Delta made the error (or not, according to Rich Farmbrough), is Maribor, and the error is in this diff. The book in question is not the one discussed above, but "War and revolution in Yugoslavia, 1941-1945: occupation and collaboration" (note the different subtitle): as you can see here, this book is from 2001, not from 1975. This is not a reprint, but a different book, for example it has 788 pages instead of 464.

(Addendum) Note that the actual quote for which this is the ref, can indeed be found on page 85 of the 2001 book, but not in the 1975 book. (end of addendum)

Considering that the one factual example he gives is wrong, I suggest that his section of evidence and his not only uncivil but baseless conclusions like "For this reason, which is pretty much representative of Fram, all his assertions, even on a factual level, have to be taken with a liberal dose of salt." should be disregarded by the ArbCom members. Fram (talk) 12:14, 18 November 2011 (UTC)
 * Actually it is a two-part work, for some reason the original cite used the ISBN from part 1. Had you fixed the ref, which you seem to find beneath you, no confusion would have arisen.  You are of course correct that volume 2 was published in 2001, and I will rewrite my evidence appropriately. Rich Farmbrough, 12:32, 18 November 2011 (UTC).


 * "which you seem to find beneath you, ". Thanks for the needless dig. The error was pointed out to Delta, he didn't correct it. Because he finds it beneath him as well? Or do you make such judgments only about people you disagree with? Fram (talk) 12:39, 18 November 2011 (UTC)
 * Δ is probably inured to your diffs, since they usually highlight trivialities you are abusing to cause drama rather than significant problems. If you stuck to identifying real problem, assuming that you can distinguish between the two, you could perform a valuable service to editors you haven't already alienated.Rich Farmbrough, 13:51, 18 November 2011 (UTC).

Rich Farmbrough corrected his evidence here. Despite the complete error of his evidence, he didn't feel the need to change his conclusions, which make one wonder whether they are actually based on any real evidence. The "corrected" evidence forgets to point out that the ISBN is the original error in that ref, while book title, page and year were the correct ones for the sentence referenced. After Delta's "correction", both the ISBN and the year were wrong, the title and the page were still correct. Apparently, pointing this out leads Rich Farmbrough to the conclusion that "For this reason, which is pretty much representative of Fram, all his assertions, even on a factual level, have to be taken with a liberal dose of salt." While I am glad that Rich Farmbrough considers cases where errors I find are indeed actual errors as "pretty much representative of Fram", the rest of his sentence no longers follows logically from that compliment. As before, I presume that ArbCom members will have no problem distinguishing Rich Farmbrough's empty rhetoric from the facts, which quite clearly disagree with his conclusions. Fram (talk) 12:48, 18 November 2011 (UTC)
 * Since you regularly post half correct screeds on my talk page and other discussion locations, I am familiar with your operating style, more than most. It is one of the reasons that your comments are generally a source of displeasure, as I have indicated, pretty robustly. Anyone with any sense would see that since I am referring primarily to personal knowledge and in particular to the evidence where diffs have already been given. The value in repeating these is virtually zero, in fact leaving them out has proved a benfit.  Leaving them out provided (not intentionally I might add) additional scope for you demonstrate here your combative and negative approach.  This will help anyone trying to evaluate your contribution to the discussion to understand that you have thrown objectivity out of the window some time ago. Not that the talk page constitutes part of the case.Rich Farmbrough, 13:40, 18 November 2011 (UTC).


 * Still, despite all your personal knowledge, it would be a good idea to present a diff of where I called these errors "egregious errors": in your again revised evidence, you claim that I made the asserion that this was an "egregious" error, which doesn't strike me as a comment I would have made. On rechecking the page User talk:Δ/20110901, I can't find it either. So, yes, please provide evidence in the forms of diffs, or don't call it "evidence" but "subjective rant" or something equally applicable.
 * I also like it how you come here, start with calling me a "usual suspect", refer to my "usual style", "which is pretty much representative of Fram", "poor reasoning is compounded with poor data", "a bunch of assuming bad faith, and considering rules more important than the encyclopaedia" (never mind your comments in follow up posts), and then pretend that I have demonstrated a "combative and negative approach". Did you really expect that your neutral and factual comments (i.e. you present one fact, but it turns out to be incorrect) would get a very welcoming reply? I have provided diffs for my claims, making it possible for everyone who is neutral in this to check whether I am objective and correct in my assertions; you, on the other hand, make unsubstantiated, incorrect negative claims, spending more time here in attacking the contributor than in looking at the actual edits. If you are not willing or able to support your statements with diffs, I'll no longer waste my time on your evidence and go back to improving the encyclopedia. Fram (talk) 14:14, 18 November 2011 (UTC)
 * That is the first sensible thing you have said for some time. Rich Farmbrough, 14:37, 18 November 2011 (UTC).

The author of this book, Jozo Tomasevich, appears to have died in 1984, so I suspect the 2001 edition is some kind of republication that combined two earlier volumes into one. 67.117.144.140 (talk) 16:13, 18 November 2011 (UTC)
 * "died in 1984" According to Worldcat, he seems to have died in 1994 (see "Alternative Names on the right) . According to his personal library donation, he seems to have died in 2000 though (?) . More importantly here, Tomasevich wrote two volumes in his "War and revolution in Yugoslavia" series:
 * Vol 1., subtitle "The Chetniks" on the Chetniks was published in 12 editions/reprints or so, with the first one in 1975.
 * Vol 2., subtitle "Occupation and Collaboration" on the Ustase and other pro-Nazis. First published posthumously in 2001, shortly after his death. This volume does not include anything on the Chetniks; it's not a reprint or 2nd edition of the first volume.
 * His 2nd volume was edited by his daughter. For proof see . ASCIIn2Bme (talk) 06:30, 19 November 2011 (UTC)
 * You're right about 1994, I must have mis-read it as 1984 (my library's catalog page displays it in a tiny little font). Thanks. 67.117.144.140 (talk) 22:13, 19 November 2011 (UTC)
 * Added: Book review. More info.    He died on 15 October 1994 according to obituary published in 1995 .   There's quite a lot of secondary coverage about his books, so I'm surprised we don't have an article about him.  More relevantly, it looks like Fram is right about the citation, and a correct gnomish edit would have been to fix the ISBN instead of making the date wrong.  The edit also removed a non-working image link File:Maribor-Kalvarija.jpg captioned "A view from hill Kalvarija" from a gallery, but I think image:Maribor from Kalvarija hill.jpg is what was intended.  Overall the diff looks like a bot edit rather than a gnome edit.  They are not the same thing.  67.117.144.140 (talk) 23:35, 19 November 2011 (UTC)


 * Finding a date based on an ISBN (or on a book-title) is a gnomish edit, fixing the ISBN because it is blatantly wrong is making it a regular edit. But this again shows clearly how evidence is used here: the one person's wrong is the other person's correct, in the end what is the error is completely grey.  And that is constantly the problem with judging 'mistakes' by editors - often there is some input wrong, and that propagates.  Of course, you can present that as 'you made an error here, the date is 2001, not 1975' (plainly assuming that the editor you are talking to made a mistake), and you can present it as 'it looks like something goes wrong here, I think the date must be 2001, not 1975 - do you know what happened?' (leaving any assumption open).
 * Regarding the image, you may want to review that, 67.117.144.140 - that image did exist 4 days before the gnome edit - it was deleted, and therefore the reference was removed from the gallery. --Dirk Beetstra T  C 14:16, 21 November 2011 (UTC)
 * It was an error. This doesn't mean that it was a deliberate one, far from it, or that there can't be a reason why that error was made. At the time I presented it, I had checked that the book existed, was from 2001, and contaiend references to events from later than 1975 (so it wasn't just a 2001 reprint of a 1975 book). I didn't "assume" that it was a mistake, I "knew" it was an error, so I didn't need to present it in your second manner. On the other hand, Rich Farmbrough wrote in his "evidence": "Very simply, Fram is wrong and Δ is correct (or Amazon is wrong too). The book was published 1 March 1975. " Perhaps you can direct your lessons in how to present "facts" and "evidence" to him instead of to me? Or is this similar to your evidence, where you only notice supposed ABF problems with CBM and me, but fail to notice the actual ABF (and civility) problems by e.g. Rich Farmbrough, or the total lack of evidence he and Delta have for their "evidence" and "workshop" contributions? I really don't need any more of your assumptions and patronizing edits, you have been shown to violate your own AGF norms in them often enough. Fram (talk) 14:27, 21 November 2011 (UTC)


 * Exactly the point. --Dirk Beetstra T  C 14:31, 21 November 2011 (UTC)
 * Re: "Finding a date based on an ISBN (or on a book-title) is a gnomish edit, fixing the ISBN because it is blatantly wrong is making it a regular edit." So gnoming means someone who just uses a script to fetch info from a database and doesn't stop to think if doing that is the right thing in a given context? Discrepancies between title/isbn/date of a book in a citation can mean a number of things, and the fix is not obvious. Deciding that the RightWikignoming&reg; thing is to always assume the ISBN is the correct piece of info is laughable. ASCIIn2Bme (talk) 07:40, 22 November 2011 (UTC)


 * I don't think I said that, and it was certainly not what I meant there. --Dirk Beetstra T  C 12:28, 22 November 2011 (UTC)


 * Dirk, you are right that File:Maribor-Kalvarija.jpg was deleted. I had checked the en.wp deletion log for that file and found it empty (suggesting that the file never existed), but it turns out that it was actually on Commons, and was deleted a few days before Δ's script operation for (cough) NFCC reasons.  I don't think that helps Δ's case any, especially since a replacement image was available that I linked above. 67.117.144.140 (talk) 04:38, 27 November 2011 (UTC)

Replies by Toshio Yamaguchi regarding evidence presented by WJBscribe
"A number of users have mentioned that interaction with Δ is rendered needlessly complicated by his choice of username."

While the first bullet point of your evidence in fact confirms there was opposition to the renaming, I don't think the claim that interaction with Δ is rendered needlessly complicated by his choice of username is valid. For example, in this arbitration case, his username Δ has been used by most of the participants apparently without major problems. And even if a user couldn't manage to find the character inside the special characters dropdown menu or paste it from somewhere else, they could still simply refer to him as "Delta" or even "Triangle" and I believe the majority of editors knew which editor is referred to with this. Furthermore the evidence you provide does not support your claim. You provided no diffs that confirm that such complications actually occurred. Toshio Yamaguchi (talk) 20:02, 6 December 2011 (UTC)
 * I see multiple edits by SporkBot to the case pages, switching one Δ template to another. That's already mildly disruptive.  Plus it refers to some TfD discussion (that I couldn't find) implying that more community energy was burned in deciding whether to make the switch.  That said, this is a minor issue compared to the rest of the case. 66.127.55.52 (talk) 08:24, 8 December 2011 (UTC)
 * I guess you are referring to edits such as this one. In which way is that edit disruptive? Can you base that claim on any policy or guideline? Furthermore these edits are performed by a bot, and one that is not run by Δ. If you think the edits of that bot are disruptive, please discuss this with that bots owner. I do not see the relevance for this case. Toshio Yamaguchi (talk) 09:03, 8 December 2011 (UTC)
 * I concur that it's disruptive. BCD (Betacommand) is how he's known, even if he's changed his name.  The documentation says it can be substituted, so the change from BCD to DELTA is not encouraged; the change to Δ is encouraged.  — Arthur Rubin  (talk) 15:25, 8 December 2011 (UTC)
 * And what does this have to do with this arbitration case? Δ is not the one who made the change to that template, so you should bring this up at the template's talk page or list the template at WP:TFD, thereby proposing to rename it. Toshio Yamaguchi (talk) 16:31, 8 December 2011 (UTC)
 * That there are multiple other editors finding themselves maintaining special purpose infrastructure just to deal with Δ's antics is not a good sign. 66.127.55.52 (talk) 08:00, 10 December 2011 (UTC)
 * This "special purpose infrastructure just to deal with Δ's antics" is just a single template generating a symbol that can also be inserted from the special characters dropdown menu. Anybody is free to refer to him with something other than this template, such as Delta. This is just a solution looking for a problem. Toshio Yamaguchi (talk) 10:04, 11 December 2011 (UTC)

This is a terrible form of evidence. It is scary that a person could be sanctioned by a comment that could be truthfully said of every Wikipedia editor. Under such guise, Anything can be said about any editor at anytime. Just today, a number of Wikipedia administrators have asked ArbCom to open an emergency case against WJBscribe asking that he be desysoped. In fact that number happens to be zero. So tell us what is the number who have mentioned the things stated as evidence here? My76Strat (talk) 01:26, 15 January 2012 (UTC)
 * Calling the first bullet point at Arbitration/Requests/Case/Betacommand 3/Evidence evidence supporting the claim that "...interaction with Δ is rendered needlessly complicated by his choice of username." is like saying "Millions of people have believed earth is flat for centuries, thus earth is flat". Toshio Yamaguchi (talk) 22:14, 20 January 2012 (UTC)

Moved from My76Strat's evidence
With regard to 1RR, I'd like to see a diff where Δ engaged in edit waring to force his views upon any article. Furthermore, I'd like to see a diff where Δ was anything less than collegial in any subsequent discussion. Producing such a diff would be contrary to the Δ I know and have collaborated with to such an extent, I anticipate no such diffs will be presented here. Moved - My76Strat (talk) 00:54, 15 January 2012 (UTC)

Replies by Toshio Yamaguchi to Newyorkbrad
Reply to : You agree here "However, given that interpretation and implementation of those sanctions has led to ongoing disputes, the community sanctions are superseded by the more straightforward remedies provided for in this decision." and so does the majority of arbitrators. So how is my evidence here a misreading? Toshio Yamaguchi (talk) 22:46, 7 February 2012 (UTC)
 * If I read your evidence correctly, you are arguing that because the Committee voted down the proposal to confirm the existing community sanctions, that means we found the sanctions were invalid all along, and therefore couldn't be the basis for any remedy.
 * What actually happened was that I proposed a clearer wording, emphasizing that the community sanctions were valid when imposed and had to be obeyed, although we are now choosing to replace them with something more straightforward. This does not in any way say we found that the community sanctions were invalid or that Delta didn't need to abide by them. It says exactly the opposite.
 * As a more general statement, "ArbCom incompetence" is an unnecessary and inflammatory comment to make in any context, and it certainly isn't helpful in the least on an evidence page. Newyorkbrad (talk) 22:50, 7 February 2012 (UTC)
 * Unclear wording of the sanctions seems to be one of the main issues this case is all about. How could the sanctions have been valid when nobody can unambiguously evaluate whether Δs edits violate the sanctions or not? And where is the evidence for the claim that "Delta didn't need to abide by them". Where is the evidence that Δ didn't try to follow the restrictions? Toshio Yamaguchi (talk) 23:05, 7 February 2012 (UTC)
 * Evidence that he violated the sanctions is laid out in detail in my section of the evidence page and the section that lists when &Delta; violated the speed restriction. It is not at all true that nobody can determine whether they were violated. In any case the arbitration committee is free to determine for themselves if &Delta; was sufficiently dedicated to following the restrictions. &mdash; Carl (CBM · talk) 23:26, 7 February 2012 (UTC)
 * These issues have been discussed at some length on the workshop and the proposed decision talkpage. I don't propose to repeat all that discussion here; my main aim above was simply to address what appeared to be a misconception. Newyorkbrad (talk) 23:30, 7 February 2012 (UTC)
 * @CBM:At least three of the blocks for his alleged violations of the sanctions were later overturned as the alleged ground for the block was found to be incorrect. I think that is just a confirmation of my claim that an unambiguous evaluation of whether Δs edits are in compliance with the sanctions is impossible under those sanctions. If ArbCom can't come up with something better than ambiguously formulated sanctions, I would call that incompetence. Toshio Yamaguchi (talk) 23:41, 7 February 2012 (UTC)
 * They did come up with an unambiguous sanction in the form of a ban. It is worth remembering that the sanction were only put in place in 2008 to avoid a ban at that time, so it was somewhat in &Delta;'s interest for the sanctions to succeed. If there was intractable community disagreement about the extent to which &Delta; violated the sanctions since then, that is exactly the sort of situation when arbcom is intended to make a binding decision. &mdash; Carl (CBM · talk) 00:03, 8 February 2012 (UTC)
 * Correct you are. They did make a binding decision, one in which they bound themselves to Betacommand's progress as an editor by way of assigning and expecting to hear reports from mentors. They utterly failed in their self assigned tasks. ArbCom incompetence. As for (now) Delta, he was banned from NFCC enforcement this past summer. Has he failed? No. Where the blame lies seems pretty blatantly obvious to me. --Hammersoft (talk) 02:06, 8 February 2012 (UTC)
 * In the evidence I presented, I mention at one point that I explicitly did not include things related to NFCC. That still leaves a long list of unapproved editing tasks and violations of the editing speed restriction. In the end &Delta; is the one responsible for his own edits, not any mentors. But he apparently also did nothing to ensure that the mentoring arrangement went through, so the blame for that cannot lie entirely on the mentors, unless we view &Delta; as unable to fend for himself. &mdash; Carl (CBM · talk) 03:10, 8 February 2012 (UTC)
 * did everything that was asked of him in regards to mentoring. How was he to know that his mentors were not doing their job? They were required to inform him, and didn't. They were required to email ArbCom monthly and didn't. He kept the mentoring notice on the top of his page for the entire period of the mentoring. There is no failure on his part in regards to mentoring. ArbCom failed, pure and simple. There is no wiggle room on that. The fact that ArbCom refuses to acknowledge that and that failure's role in this case is extremely telling. --Hammersoft (talk) 03:35, 8 February 2012 (UTC)
 * Your second sentence has the answer to the question in your first sentence. But I agree that the mentoring apparently failed, which is why I argued that a new set of mentors would not be effective this time. &mdash; Carl (CBM · talk) 03:54, 8 February 2012 (UTC)
 * No, it doesn't. had no reason to believe in any respect that he was not being mentored. --Hammersoft (talk) 14:01, 8 February 2012 (UTC)