Wikipedia talk:Requests for adminship/Archive 251

Reduce verbiage in vote sections
Simple proposal. Ask people to use no more than 25 words maximum to give their "rationales" of any sort on the main page, and have "questions" about any such votes be restricted to the Talk page in the first place, and also restricted to twenty-five words or less. Keeping to "gists only" should reduce the tendency for extended colloquy. I note that very few people change their positions upon being confronted by those questioning their rationales. Net result should be much shorter "discussions" ab initio. Collect (talk) 18:23, 3 August 2018 (UTC)
 * Alternatively, oppose !voters could refrain from taking responses to their !vote personally unless there is a good reason to do. Lepricavark (talk) 18:59, 3 August 2018 (UTC)
 * Nah, this would make it impossible to provide a good analysis, pro or con and would really reduce RfA to but a popularity-contest vote, instead of a consensus discussion marred by having aspects of a popularity-contest vote.  On the other hand, such a limit on  people's initial support/oppose/neutral rationales might be reasonable.  — SMcCandlish ☏ ¢ 😼  19:09, 3 August 2018 (UTC)
 * IOW, allow such a "first response" on the main page - and then go to the Talk page.   I would suggest that such "first responses" be strictly verboten to raise "personalities", though.  Seems sensible from here. Collect (talk) 19:14, 3 August 2018 (UTC)
 * Again, Advice for RfA voters is the answer. It's all in there.  Makes yet another essay redundant. Kudpung กุดผึ้ง (talk) 19:54, 3 August 2018 (UTC)
 * Essays are not very binding. An actual stated rule would be better considering recent RfA (heck, even old RfA) history. Collect (talk) 20:18, 3 August 2018 (UTC)
 * Nobody said essays are binding,, and that essay is far from prescriptive. It  says what it says and is something  for the less experienced or disingenuous voters to  reflect upon . Perhaps if you were to  read it... Kudpung กุดผึ้ง (talk) 04:34, 4 August 2018 (UTC)
 * OTOH, perhaps if you were to read my eighteen essays, you might get a better feel for my positions, I would hope. My position is that actual stated rules are far more useful than any essay can be. Is that more clear? Collect (talk) 13:37, 4 August 2018 (UTC)
 * 18 essays??, I'd rather read your 15 mainspace creations (stubs) and see which are the better contributions to building an encyclopedia. Kudpung กุดผึ้ง (talk) 14:23, 5 August 2018 (UTC)
 * Try my GA - one of which is unique, as far as I can tell, because it took an unwieldy mess and made it into a good article (Joseph Widney, one where a lot of collaborative work made a GA out of Christian Science, got a DYK for creating Charles S. Strong, and so on, not to mention the fact that I have worked on multiple other Wiki-projects as well.  I do suggest you read the Widney one though - it is easier to write something way too long than to pare it down to a useful size at times. Collect (talk) 15:07, 5 August 2018 (UTC)
 * Per SMc, this would make RfA one step away from a vote. I don't think that's good, RfA is already too close to a vote. ~  Amory  (u • t • c) 20:38, 3 August 2018 (UTC)
 * I agree with Amory; this is not a good idea. "No more than 25 words" would eliminate the whole idea that this is supposed to be a discussion about the person, not a simple vote. And it would give the 'crats nothing much to go on in close cases. And it would eliminate the ability to share information gleaned from looking into the person's background or contributions. --MelanieN (talk) 02:48, 4 August 2018 (UTC)
 * Read the essay . There are examples of text length. I don't  think  anyone is complaining  about  25 words, but  when someone posts an essay-length diatribe for a vote, that  becomes inappropriate -  especially  when 50  preceding  oppose votes have already  said what  can be told. Kudpung กุดผึ้ง (talk) 04:34, 4 August 2018 (UTC)
 * Most long oppose votes can be summarized into a couple of links and a short description, so this would be a good idea in theory, but in practice !voters are very hard to control. Esquivalience (talk) 14:59, 4 August 2018 (UTC)


 * I do sympathize with the reason for this proposal, and I think the suggestions about length at Advice for RfA voters are just about right, but I would be reluctant to support a policy-based limit unless excessively long answers become a persistent problem. I certainly oppose anything as short as 25 words, as there are definitely cases where a clear explanation of a relevant issue can not possibly be explained that concisely. Boing! said Zebedee (talk) 07:09, 5 August 2018 (UTC)
 * I originally used "50" but them counted a lot of oppose rationales and came down to 25, but 50 would allow a great deal of leeway, for sure. The "25" is not a "sacred number" at all. It does tend to reduce adjectives. Collect (talk) 15:15, 5 August 2018 (UTC)
 * "Make your point, then stop". If someone does not understand, they can ask. Cheers, &middot; &middot; &middot; Peter (Southwood) (talk): 16:33, 5 August 2018 (UTC)
 * Limiting the length of !votes does not make sense to me. It could see an increase in "per my question" opposes, and indirectly in questions. Worse it discourages people from properly assessing the candidate and then setting out their findings, and even worse, it would make the site snarkier as the first thing people would cut would be comments such as "despite the candidates excellent contributions in " ..... and "the candidate is heading in the right direction and I would be happy to reassess in x months".  Ϣere Spiel  Chequers  11:42, 6 August 2018 (UTC)
 * Indeed, and "change their positions" is not really the objective of discussions following a vote, it is more to make sure that facts and insinuations made in those votes are clarified for future voters, both by the original voter and by others at the discussion. Imposing arbitrary restrictions on how much voters or responders can say doesn't help anyone. &mdash; Amakuru (talk) 12:16, 6 August 2018 (UTC)


 * We whose views should be limited to 25 or fewer words should read 18 essays? I could sputter for a bit but I'll just say No.  (1 word over the limit, tsk.)  – Athaenara  ✉  16:14, 6 August 2018 (UTC)
 * Dang, just realized the removal of one preposition ("for") would get it to 25. Should have spent more time editing that!  – Athaenara  ✉  18:14, 6 August 2018 (UTC)
 * I think '25 words' was just an unreflected number thrown down with a flick of a wrist. At least I hope so.  But anyone who writes more essays than articles is confusing Wikipedia with Wordpress.  Makes me wonder if we have made any progress at all with voters since WP:RFA2011 and 's Dec 2015 reforms.  (This post, excluding signature and this note, is 55 words) . Kudpung กุดผึ้ง (talk) 21:09, 8 August 2018 (UTC)
 * Oppose twitterification of RFA. An "essay-length diatribe" is inappropriate because it's a diatribe, not because of the length. Nuance is (often) helpful. Yes, sometimes all it takes is 10 words. But sometimes it's important to put a situation in context, to explain why an example raises concern about a pattern (oppose), or to explain why concerns raised are not persuasive (support). And censorship is bad. Martinp (talk) 10:20, 9 August 2018 (UTC)
 * By the way, while indeed few !voters change their vote as a result of discussion, I for one sometimes find such discussion of prior !votes helpful to read as I make up my own mind. Martinp (talk) 10:23, 9 August 2018 (UTC)
 * Yeah this doesn't sound like a good idea, per above. I read Kudpung's essay. Most of it is common sense, but I wasn't aware of the 200 words guideline. I think I'm the only one who's posted anything resembling an "essay-length diatribe" lately, and it wasn't in the main voting section anyway. ♫ ekips39 (talk) ❀ 23:00, 10 August 2018 (UTC)
 * Just so you may not have a wrong notion; the 200 words limit is just an advice, and not a guideline (although I support the premise wholeheartedly). Lourdes  07:09, 11 August 2018 (UTC)
 * Yes, I realized my choice of wording could give the wrong impression, but I couldn't think of anything better. ♫ ekips39 (talk) ❀ 16:32, 11 August 2018 (UTC)

Proposal: five year statute of limitations on asserted negative matters
Not that this is currently a problem, but I recall seeing on occasion opposes premised on things like "this editor was involved in an edit war in 2009" or "I had a very abrasive encounter with this editor in 2007". Maybe three years would be better, or maybe ten years would, but I think that there should be some express limitation on how far back you can go to bring up negative information with respect to an adminship bid. People grow and change over time, and if an editor has done nothing comparable for the last several years, discussions should not be muddied with old grudges. I would have exceptions for circumstances where there is more recent misbehavior, and the older behavior is part of a continuous or repeating pattern, or where the editor has had minimal participation for the last several years, so the misbehavior is still within their last few hundred edits. bd2412 T 15:37, 3 August 2018 (UTC)
 * Why limit this to negatives? Argento Surfer (talk) 20:27, 3 August 2018 (UTC)
 * It's hard to envision a nominee supporting their nomination by reference to good edits made half a decade ago or more. I suppose it could come up if a nominee is asked what are the best articles they have contributed, and the nominee is particularly fond of an old work. Even so, examples of old positive contributions have far less potential to be unfairly disruptive to the process. bd2412  T 02:02, 4 August 2018 (UTC)
 * On the few occasions I have seen minor very old stuff like this brought up it seems to get dismissed by others as too old. Not many people can remember or find the diffs that far back. By the time you have the exceptions in, this would be very fiddly to draft, probably with a good deal of room to argue over. Just now people are bringing up stuff barely 3 months old, which seems more typical. Johnbod (talk) 02:12, 4 August 2018 (UTC)
 * Would it be more sensible with a different limitations period? I think that it is generally deleterious to the process for excessively old stuff to be brought up at all, even if it is quickly dismissed. I grant that someone with a grudge may find other reasons to oppose a nominee, but if all they can come up with is some very old issue, I would rather they leave it alone altogether. bd2412  T 02:17, 4 August 2018 (UTC)


 * I think this would have to depend on what the candidate did five years ago. Bringing up an edit war which took place five years ago would be excessive, but bringing up a siteban from five years ago would not be, IMO.  Hut 8.5  19:23, 4 August 2018 (UTC)
 * I think it would, if the candidate is currently a productive editor. Would it make a difference if it was a siteban from ten years ago, or twelve years ago? If not, then we might as well tell editors, once you've done something that someone considers to be bad enough, you can never, ever be trusted with the admin tools no matter how long you work on this project, or what you do here. bd2412  T 19:59, 4 August 2018 (UTC)
 * Something very serious like that would have to be addressed in the RFA, however old - wise candidates mention such things up front. I think we are mainly dealing with rather more trivial stuff like the examples given at the start. Johnbod (talk) 20:34, 4 August 2018 (UTC)
 * Suppose an editor was de-sysoped by Arbcom for blatant and serious misuse of the admin tools, 6 years ago. Just over 5 years ago, they had a name change, meaning there wouldn't even be an indication this was a 2nd RFA based on page name. Would we censor any reference to the past Arbcom case? To the fact that they were ever an admin? For editors that were blocked for an extended period, and subsequently unblocked, does the time run from the block or the unblock? (ie, unblocked a year ago, after 5 years of being blocked, is the block off limits?) At the end of the day, we shouldn't try to censor truthful statements about the past, and instead should trust other editors to consider the passage of time when deciding what weight to give them. Monty  845  04:12, 6 August 2018 (UTC)
 * This can't be applied as a strict rule, but people should take into account the fact that editors get better over time. ~ Rob 13 Talk 04:22, 6 August 2018 (UTC)
 * There are some editors that I would oppose based on ARBCOM cases that happened over 5 years ago. I suppose I could phrase the oppose as based on "a continuing pattern of behavior" for most of those.  power~enwiki ( π,  ν ) 04:38, 6 August 2018 (UTC)
 * So how about not doing this for anything which involved an ARBCOM decision? I can see the benefit of it otherwise for restricting personal grudges. Andy Dingley (talk) 10:22, 6 August 2018 (UTC)


 * Oppose in favour of full disclosure, including "who was involved" and "when". eg.  He, Bob and I had an edit war in 2009.  Transparency is the cure of most evils.  Perhaps also "Do not use the passive voice when making allegations".  Innuendo is easily couched in the passive voice, "There is innuendo in Bob's statements", but it is harder to assert an untrue "Bob conceals innuendo in his statements".   --SmokeyJoe (talk) 04:50, 6 August 2018 (UTC)
 * Transparency needs detail, preparation and time to read it. It's the headlines which stick. Andy Dingley (talk) 10:25, 6 August 2018 (UTC)


 * Oppose Things like EEML can never be forgotten. 78.28.54.8 (talk) 18:06, 13 August 2018 (UTC)

Editors with ECP from ~12 months ago
This is a list of editors who got the ECP permission in June, July, or August 2017; either from making their 500th edit then, or making their first edit since ECP was introduced. (I've filtered out blocked users, users with <2000 edits, or users who registered before January 2010)

It would be impolite and excessive to litigate each of these individually, but there should be at least one of them who's a reasonable admin candidate now, right? Or will be in the near future? If not, maybe you should lower your standards. The project functioned for a very long time making people with this much (or less) experience admins, the solution to "why aren't there more new admins" is obvious. power~enwiki ( π, ν ) 18:31, 23 August 2018 (UTC)


 * User:*angys*: registered 2016-12-28T10:49:52Z, editcount 4998
 * User:Acorona619: registered 2016-11-09T04:16:05Z, editcount 2736
 * User:Alanna the Brave: registered 2017-05-04T23:23:21Z, editcount 3975
 * User:Alsoriano97: registered 2016-04-14T17:26:09Z, editcount 3653
 * User:American Military History: registered 2017-07-23T16:04:54Z, editcount 2090
 * User:Anarcho-authoritarian: registered 2016-10-16T20:40:04Z, editcount 2089
 * User:Anbans 585: registered 2014-11-06T20:29:18Z, editcount 5023
 * User:Anon9088: registered 2017-05-03T00:54:04Z, editcount 2190
 * User:AntonyRomanos2017: registered 2016-12-31T08:10:04Z, editcount 2025
 * User:Ariel Cetrone (WMDC): registered 2017-05-01T10:37:19Z, editcount 3143
 * User:Arllaw: registered 2016-02-11T23:49:51Z, editcount 2288
 * User:Arutai: registered 2017-05-21T21:37:25Z, editcount 4262
 * User:Aspening: registered 2017-05-21T19:54:53Z, editcount 9956
 * User:AvalerionV: registered 2017-04-24T18:21:22Z, editcount 10860
 * User:AyodeleA1: registered 2017-03-26T23:21:14Z, editcount 7060
 * User:Benawu2: registered 2012-06-17T22:00:56Z, editcount 2994
 * User:Bigcheddar: registered 2014-05-12T01:23:24Z, editcount 2836
 * User:BlueFire10: registered 2011-07-26T13:23:09Z, editcount 3871
 * User:C1MM: registered 2017-05-23T22:12:06Z, editcount 3127
 * User:CASSIOPEIA: registered 2017-05-10T01:53:06Z, editcount 36225
 * User:CPA-5: registered 2016-08-27T23:31:50Z, editcount 4019
 * User:Cageyfella: registered 2015-07-30T21:26:19Z, editcount 3776
 * User:Candido: registered 2015-08-30T17:52:12Z, editcount 2366
 * User:CataracticPlanets: registered 2017-03-16T02:56:28Z, editcount 4053
 * User:ChandraHelsinky: registered 2015-03-04T12:18:59Z, editcount 2129
 * User:Charlie Foxtrot66: registered 2017-04-29T13:40:53Z, editcount 2034
 * User:Chetsford: registered 2017-04-30T22:09:02Z, editcount 16719
 * User:Cky2250: registered 2011-10-29T16:36:31Z, editcount 3240
 * User:Crook1: registered 2017-06-18T22:07:37Z, editcount 5041
 * User:D4R1U5: registered 2014-01-25T09:37:51Z, editcount 9616
 * User:DSmurf: registered 2017-05-11T22:54:25Z, editcount 4029
 * User:Daiyusha: registered 2015-10-23T08:11:44Z, editcount 4241
 * User:Damian Vo: registered 2012-12-04T14:52:54Z, editcount 7990
 * User:DerDFB: registered 2017-07-19T11:48:38Z, editcount 6281
 * User:DilletantiAnonymous: registered 2017-06-02T14:33:15Z, editcount 14285
 * User:Dlthewave: registered 2014-11-06T01:30:36Z, editcount 4041
 * User:Doctorg: registered 2016-08-06T12:04:34Z, editcount 3707
 * User:Donald Trung: registered 2017-05-28T04:10:08Z, editcount 7253
 * User:DovahDuck: registered 2016-10-11T02:30:23Z, editcount 6106
 * User:DragoMynaa: registered 2017-06-23T23:37:23Z, editcount 3865
 * User:Drali1954: registered 2013-02-16T07:08:02Z, editcount 2629
 * User:Ecangola: registered 2017-07-03T13:34:00Z, editcount 13281
 * User:Ecpiandy: registered 2011-11-10T00:49:42Z, editcount 29703
 * User:ElectionsGeek: registered 2017-04-28T10:26:41Z, editcount 4989
 * User:Emiya1980: registered 2017-04-30T19:15:50Z, editcount 3695
 * User:Espngeek: registered 2016-02-22T12:26:07Z, editcount 5362
 * User:FR30799386: registered 2017-04-08T03:50:41Z, editcount 3576
 * User:Fucherastonmeym87: registered 2017-04-15T23:31:19Z, editcount 6557
 * User:Furicorn: registered 2013-05-02T02:27:30Z, editcount 2685
 * User:Gaioa: registered 2010-10-25T09:32:26Z, editcount 2043
 * User:Galactikapedia: registered 2017-04-25T03:48:44Z, editcount 5787
 * User:Gandhi (BYU): registered 2017-05-08T19:02:22Z, editcount 2551
 * User:Gapfall: registered 2017-05-26T08:33:56Z, editcount 3713
 * User:Gooseflesh12: registered 2017-01-01T05:52:26Z, editcount 4262
 * User:GrafVonTirol: registered 2017-05-23T02:29:06Z, editcount 3424
 * User:GrammarFascist: registered 2011-09-28T02:25:01Z, editcount 2964
 * User:Grimes2: registered 2016-05-14T11:21:52Z, editcount 4182
 * User:Guysela25: registered 2014-05-27T06:00:51Z, editcount 4644
 * User:Gz33: registered 2010-05-26T13:05:05Z, editcount 4471
 * User:Hameltion: registered 2017-05-20T22:59:47Z, editcount 3447
 * User:Harambe Walks: registered 2016-08-27T01:26:26Z, editcount 4221
 * User:Hardinggratz: registered 2011-02-25T14:32:37Z, editcount 2008
 * User:Hearmesir: registered 2017-06-03T01:45:59Z, editcount 2898
 * User:Hiwilms: registered 2014-05-17T07:10:30Z, editcount 2078
 * User:Hrodvarsson: registered 2017-05-13T22:49:50Z, editcount 9139
 * User:I am aman goyal: registered 2017-03-09T04:02:27Z, editcount 2912
 * User:IOnlyKnowFiveWords: registered 2016-11-26T10:28:22Z, editcount 2011
 * User:IiKkEe: registered 2014-04-06T20:50:08Z, editcount 8129
 * User:Ivar the Boneful: registered 2017-07-13T06:57:49Z, editcount 7532
 * User:Jackmegill: registered 2016-10-29T13:32:44Z, editcount 2121
 * User:JahlilMA: registered 2017-05-14T05:15:52Z, editcount 8295
 * User:Jarould: registered 2014-06-10T03:02:11Z, editcount 2787
 * User:Javert2113: registered 2010-03-05T01:29:59Z, editcount 5737
 * User:JelgavaLV: registered 2013-02-01T03:09:40Z, editcount 2262
 * User:Jenhawk777: registered 2017-05-15T04:19:15Z, editcount 8928
 * User:Joaofranciscoaugusto: registered 2017-04-11T17:51:26Z, editcount 2084
 * User:JohnWickTwo: registered 2017-02-11T01:56:26Z, editcount 2055
 * User:Kai445: registered 2010-09-20T16:09:17Z, editcount 2621
 * User:Katchilla: registered 2013-08-28T03:53:06Z, editcount 2094
 * User:Kb.au: registered 2014-08-03T09:01:47Z, editcount 4534
 * User:Kiewii: registered 2011-08-28T15:48:30Z, editcount 5148
 * User:KingdomHearts25: registered 2010-01-09T08:15:19Z, editcount 2717
 * User:Klilidiplomus: registered 2011-11-21T10:52:44Z, editcount 12152
 * User:Kshitij.agr: registered 2017-02-18T08:53:21Z, editcount 2259
 * User:L Kensington: registered 2010-07-02T15:52:05Z, editcount 48871
 * User:L1amw90: registered 2017-02-09T04:45:17Z, editcount 3009
 * User:Lachie23: registered 2016-10-15T14:15:29Z, editcount 2587
 * User:LeftAire: registered 2011-10-12T01:39:10Z, editcount 3073
 * User:Legalskeptic: registered 2010-01-27T15:26:16Z, editcount 9071
 * User:Levellend: registered 2010-06-14T22:26:48Z, editcount 10934
 * User:Lofoten1994: registered 2017-07-08T14:23:20Z, editcount 3122
 * User:Lusotitan: registered 2014-04-12T17:31:24Z, editcount 2770
 * User:M9154: registered 2017-05-27T07:50:42Z, editcount 2907
 * User:Malayy: registered 2017-03-28T04:25:07Z, editcount 3118
 * User:Mariacricket: registered 2017-04-23T01:51:22Z, editcount 4731
 * User:Mathphysman: registered 2013-09-27T18:53:47Z, editcount 2796
 * User:Menaechmi: registered 2017-05-08T19:33:15Z, editcount 2346
 * User:Meryam90: registered 2011-06-10T13:32:15Z, editcount 5666
 * User:MetalSword: registered 2017-05-20T22:55:56Z, editcount 2883
 * User:Mompati Dikunwane: registered 2013-01-23T10:38:18Z, editcount 5159
 * User:Myk Streja: registered 2017-05-17T02:12:52Z, editcount 3249
 * User:NZFC: registered 2017-04-04T06:15:10Z, editcount 7514
 * User:Nanerz: registered 2017-05-28T03:15:38Z, editcount 2041
 * User:Neils51: registered 2012-03-04T21:00:09Z, editcount 7923
 * User:Nihlus: registered 2016-09-09T01:49:13Z, editcount 13072
 * User:Nixinova: registered 2016-11-22T03:25:02Z, editcount 4032
 * User:OrganoMetallurgy: registered 2017-05-24T14:08:18Z, editcount 6332
 * User:Oronsay: registered 2014-10-27T18:15:46Z, editcount 6324
 * User:OttawaAC: registered 2011-08-08T23:49:57Z, editcount 7043
 * User:Phoenix God: registered 2016-12-09T12:32:09Z, editcount 3009
 * User:Polyporales: registered 2016-09-04T20:04:36Z, editcount 2800
 * User:Pozzi.c: registered 2017-05-24T23:38:23Z, editcount 5773
 * User:Prince-of-Canton: registered 2017-04-30T23:54:49Z, editcount 2025
 * User:Quite A Character: registered 2017-05-31T18:30:30Z, editcount 26018
 * User:RafaelS1979: registered 2015-09-26T20:34:13Z, editcount 5278
 * User:Ralbegen: registered 2016-05-01T14:45:35Z, editcount 2786
 * User:ReaperSox: registered 2017-04-13T17:01:35Z, editcount 10125
 * User:Rhadow: registered 2016-02-17T09:34:28Z, editcount 5923
 * User:Robertgombos: registered 2012-07-29T11:49:28Z, editcount 3842
 * User:Ryangrxy: registered 2015-02-07T10:10:25Z, editcount 3390
 * User:Sakralamn: registered 2016-07-13T14:48:42Z, editcount 3766
 * User:Saksapoiss: registered 2016-09-20T20:41:39Z, editcount 3467
 * User:Samantha Ireland: registered 2015-01-20T18:28:02Z, editcount 3269
 * User:Sb008: registered 2017-04-09T17:21:11Z, editcount 2863
 * User:Sb2001: registered 2012-06-16T20:11:16Z, editcount 3257
 * User:Scrabble Scribble: registered 2017-04-21T15:02:21Z, editcount 6849
 * User:ShotgunMavericks: registered 2017-02-10T13:01:26Z, editcount 2043
 * User:SiberianCat: registered 2013-09-23T18:20:04Z, editcount 2060
 * User:Sivagopalakrishnan: registered 2016-08-12T20:31:03Z, editcount 2748
 * User:Speedthief1: registered 2014-01-05T07:49:55Z, editcount 5653
 * User:SportsFan007: registered 2012-02-21T18:04:38Z, editcount 5748
 * User:Stalin990: registered 2015-02-01T21:17:51Z, editcount 3111
 * User:Standard2211: registered 2011-11-05T01:43:33Z, editcount 2529
 * User:Starspotter: registered 2017-03-25T12:47:04Z, editcount 2440
 * User:Stormy clouds: registered 2017-04-20T18:34:55Z, editcount 3487
 * User:Stricer1990: registered 2016-07-06T19:14:51Z, editcount 2066
 * User:Supasun: registered 2011-05-12T13:16:14Z, editcount 2407
 * User:Sørhaug: registered 2016-04-23T18:04:18Z, editcount 2167
 * User:T81487: registered 2017-01-30T20:21:38Z, editcount 3284
 * User:TheDisneyGamer: registered 2015-04-30T20:09:34Z, editcount 3184
 * User:TheDoctorWho: registered 2017-07-20T21:37:01Z, editcount 4587
 * User:TheLastBrunnenG: registered 2017-05-17T20:12:18Z, editcount 2085
 * User:TheMovieGuy: registered 2017-03-02T00:48:41Z, editcount 8764
 * User:Theinstantmatrix: registered 2017-07-22T22:56:10Z, editcount 22750
 * User:Theophilus92: registered 2017-04-26T09:43:24Z, editcount 2384
 * User:Toreightyone: registered 2017-05-11T22:31:36Z, editcount 2567
 * User:Tornado chaser: registered 2017-05-14T00:34:32Z, editcount 15019
 * User:TrevelyanLittle: registered 2016-11-27T11:15:56Z, editcount 2017
 * User:Usb10: registered 2010-07-21T15:25:08Z, editcount 5089
 * User:Vanischenu: registered 2011-11-17T16:34:03Z, editcount 4782
 * User:Volunteer1234: registered 2017-04-14T02:13:24Z, editcount 3988
 * User:Volvlogia: registered 2015-10-28T23:09:57Z, editcount 2122
 * User:Vorbee: registered 2016-04-20T20:13:08Z, editcount 2174
 * User:Whispyhistory: registered 2017-04-11T10:03:48Z, editcount 6837
 * User:Wiki Feed Bot: registered 2017-01-02T21:10:48Z, editcount 5649
 * User:Wiki-wuki: registered 2016-10-14T17:51:26Z, editcount 6561
 * User:Wikiims: registered 2012-02-22T11:57:44Z, editcount 2283
 * User:Wkg321: registered 2017-06-14T17:34:42Z, editcount 8793
 * User:WomenArtistUpdates: registered 2017-03-14T16:03:23Z, editcount 6378
 * User:Wumbolo: registered 2016-04-23T14:31:43Z, editcount 7803
 * User:XOR'easter: registered 2017-04-01T16:34:07Z, editcount 4677
 * User:Yellow2pig: registered 2017-04-29T22:51:28Z, editcount 2485
 * User:ZUKAGRAD: registered 2016-05-30T13:13:38Z, editcount 2087
 * User:ז62: registered 2016-10-28T17:05:53Z, editcount 2072
 * User:カビル: registered 2017-05-08T11:10:46Z, editcount 4403

Anonymous voting
I recently proposed (on this page) that the RFA process be fundamentally changed to deal with the problem of distracting, bad-faith oppose !votes that lead otherwise productive discussions astray. The reception to this proposal was generally negative, largely, it seems, because other editors do not think this is nearly a big enough issue to warrant such major changes. Instead of re-proposing the same ideas to address the same problem, I want to resurrect one specific idea I proposed that I still think would improve the RFA process: having people !vote anonymously. That is, I think voting for RFAs should be like voting for ArbCom candidates, so that you just check either "Yes" or "No" on an online, secure form for someone's RFA, and the votes are tallied off-wiki. This of course would mean that "support"/"oppose"/"neutral" sections on RFAs would no longer exist and no one would know who voted which way, or why they voted that way. I think this would make RFAs much less hostile and thus more appealing to potential, worthy candidates. But I recall that in the previous discussion I started, which I mentioned at the beginning of this post, some people did not think this would be a good idea. I want to get more feedback from other editors as to whether they like the idea of anonymous voting and why. IntoThinAir (formerly Everymorning) talk  17:12, 30 August 2018 (UTC)
 * I oppose this change. I find great value in seeing why people !voted as they did, and have occasionally switched my !vote when good cases have been made. -- SarekOfVulcan (talk) 17:15, 30 August 2018 (UTC)
 * I agree with Sarek. RfA should be a discussion, not a vote. Quite a few years ago, we had RfA-style !voting for ArbCom, and there were very few opposers. Then when we switched to anonymous voting for that, the voting patterns changed dramatically, because no one could see who opposed. If we were to do that here, with RfA, the amount of opposition would increase a lot, not decrease as you would hope. --Tryptofish (talk) 19:14, 30 August 2018 (UTC)
 * Imma raise a relevant point made by on WT:FAC Just as a point of order, back when Arbcom votes were publicly logged there was no apparent shortage of opposers, while even high-profile people with a reputation as "insiders" or of having a clique of followers could attract significant opposition without any apparent fears of reprisals among the opposers. The introduction of secret ballots was the result of a somewhat questionable supervote by the closer as part of this RFC (73 in favour of publishing the names and votes of voters either during or after the election, 71 in favour of secret ballot), not by some kind of mass acclaim. &#8209; Iridescent 08:21, 22 April 2018 (UTC) It seems like making votes private does not really change the amount of support/opposition that people get. Jo-Jo Eumerus (talk, contributions) 10:22, 2 September 2018 (UTC)
 * I wish the discussion and the voting were separate. (No democratic vote I know of allows campaigning on the ballot paper). I am not sure anonymity in the voting would help, though. —Kusma (t·c) 05:58, 31 August 2018 (UTC)
 * Surely it would severely limit the 'Crats ability to weigh up the pro's and cons in a discretionary range case? Leaky  Caldron  09:47, 31 August 2018 (UTC)
 * I actually like the idea of comments being left in an anonymous way, so they could be seen but the person making them wouldn't be known. But I don't know how that could be done. Aiken D 10:13, 31 August 2018 (UTC)
 * I strongly oppose having annonymous voting/commenting. We don't have this system anywhere else on wikipedia, why would we for a discussion? Annonymising comments generally lead towards ostrasizing, and harsher language, as it hides people to consequences. I'm not condoning being harsh on people because of a viewpoint (far from it), but if a silly comment is made, it's important to let the user know why the rationale is flawed. As an aside, how would this proposal work against SPIs? On a smaller scale, if this were a thing for say, afD, sockpuppets and SPIs would be rampant to get a page deleted/kept, why would this be more acceptable when deciding if someone has rights to the strongest powers on Wikipedia?  Lee Vilenski (talk • contribs) 10:39, 31 August 2018 (UTC)


 * Oppose as above. Not needed and ripe for abuse. GiantSnowman 11:02, 31 August 2018 (UTC)
 * It might make the RfAs less hostile, or it might not. It's really hard to know. Discussion wouldn't stop. Yes, ArbCom elections are anonymous voting style. But, discussion still happens. In fact, it takes a long time to work through all of it during an ArbCom election cycle. Consider; a candidate has a statement (think of it as a self nom statement), an area for questions for the candidate, and an area to discuss the candidate. If you implemented anonymous voting, I think it very likely we would implement a similar construct to how ArbCom elections are done. Thus, there would still be plenty of discussion, question and answers, etc. So, from a bureaucracy standpoint, I think the bureaucracy to become an RfA would increase. From an hostility standpoint, it might not have any effect at all. --Hammersoft (talk) 14:19, 31 August 2018 (UTC)
 * RFA is intended to strongly include discussion as the principle deciding factor on whether someone is granted the tools. It is not a vote but rather !vote/participation in a consensus building process. Mkdw  talk 17:32, 31 August 2018 (UTC)
 * I can recall RFAs which have tanked after someone pointed out a major flaw with the candidate, I can also remember RFAs where people have switched their vote in the opposite direction after having things pointed out to them. Such as a recent deletion of a candidate's article being a recent deletion of an ancient article. That sort of dialogue is one of the saving graces of RFA and ths proposal would lose it with no obvious benefit.  Ϣere Spiel  Chequers  20:44, 31 August 2018 (UTC)
 * As others have said above, the discussion element of RfA is important. It is useful to read the rationales provided by other !voters. Also, the anonymous voting system would be ripe for abuse. The process does need to be reformed, but this is not the way to do it. And even if somebody came up with a good proposal for RfA reform, it most likely would not pass. Lepricavark (talk) 03:54, 1 September 2018 (UTC)
 * Oppose. RFA is (supposed to be) a discussion, not a vote. Besides, you idea is technically infeasible unless we use xtools. L293D (☎ • ✎) 02:51, 2 September 2018 (UTC)
 * Oppose My concern is that this could potentially lead to on/off-wiki canvassing by candidates/opposers as there would be less opportunity to raise the issue with the closing bureaucrat. Number   5  7  09:03, 2 September 2018 (UTC)
 * Oppose. Unlike ArbCom elections, RFAs are surveys which include rationales. People running for ArbCom have already run through a gauntlet of a battery of questions and responses in the run-up to the election; plus they are generally already experienced administrators themselves (no non-admin has come close to being elected an arbitrator). Having rationales for support, neutral, and oppose !votes, and discussing them, and knowing who is making which comment and which !vote, is absolutely intrinsic to the RFA process. If someone is afraid to stand up and have their RFA !vote(s) be seen and counted, well then they need to ask themselves why. Softlavender (talk) 09:19, 2 September 2018 (UTC)
 * This has been suggested a number of times before, but has never gained any ground. As SarekOfVulcan and others say, it's supposed to be a discussion not a vote, and we do need that discussion. A candidate's record here is usually sufficiently extensive that few individuals can review it in sufficient depth to make a proper appraisal, and we do depend on each other's reviews and discoveries. I have often changed from my initial feelings (sometimes even after first !voting) based on the keen analysis of others, and we should not lose that. Also, crat chats could not work for borderline cases if all they could see was a vote. The only way something like this might work is if RFA was split into two phases, as has been suggested, a discussion and then an anonymous vote. But even if that gained support, we'd have the technical workload of managing an Arbcom-style secret vote, with the required scrutiny period and scrutineers, for every RFA rather than just once a year - and that's surely a non-starter. (I know the difference between "every RFA" and "once a year" is becoming ever smaller, but that's a different issue). Boing! said Zebedee (talk) 09:49, 2 September 2018 (UTC)
 * Oppose for two reasons. First, I fear that anonymity would lead to more, not less, trolling and negativity, since editors could write mean or disrespectful comments without fear of damaging their wiki-reputations. Also, from a personal perspective, if I had been criticized strongly by one or more editors during my RFA and didn't know who, I would be nervous when interacting with anyone on talk pages afterwards.  I'd rather know up front who has issues with my editing so that I can choose my words accordingly.&mdash;Anne Delong (talk) 11:59, 2 September 2018 (UTC)
 * Oppose — RfA is somewhere between a vote and a discussion, and both elements should not be radically reformed in this way. —  python coder    (talk &#124; contribs) 01:27, 3 September 2018 (UTC)
 * I feel like I should clarify a point about what I am proposing: I am proposing that voting in RFAs happen entirely behind closed doors rather than on a publicly viewable page, the same way ArbCom elections are done, for example. This means I am not proposing keeping the system of public voting and commenting but just removing the names from people who comment or explain their vote, which appears to be how my proposal has been accidentally misinterpreted by some above. Instead, I want the entire system of publicly voting and commenting on RFAs to be completely eliminated. Obviously this is a drastic proposal that, also obviously, many are opposed to, but I think that voters are more than capable of researching a candidate adequately to decide whether to support or oppose their RFA without help from other voters. IntoThinAir (formerly Everymorning) talk  17:00, 5 September 2018 (UTC)
 * It's a non-starter and you are wasting your time. Unfortunately you are also wasting a lot other people's time, those who cannot resist the temptation of commenting. I would say that you have missed the point of both RfA and !votes. Leaky  Caldron  17:07, 5 September 2018 (UTC)

Fewer admins per edit
Comparing today's figure of 522 active editors to 12th Sept 2014 when we had 610 (Sept 2014 was close to the low point of the 2007-2014 decline in editing). Our number of active admins has dropped by over 13% since then, and with a replacement rate of barely 2% this would only be sustainable if our admins averaged 50 years adminship post RFA, and that won't happen unless we find ways to extend their real lives. Editing volumes have picked up since the 2014 trough, in the second half of 2014 we were taking over ten weeks per ten million edits. Editing levels now are up more than 10% compared to late 2014, with circa 9 weeks per ten million edits. 2018 editing levels ares more comparable to 2012 two years before the 2014 minima, but with fewer admins available per million edits taking place. If RFA was healthy you would expect that as the community stabilises we would have a growing proportion of admins and relatively few newbies. Instead we have a dwindling number of admins and an editing level that is stable or possibly growing and is currently more than ten percent above the 2014 lull.  Ϣere Spiel  Chequers  11:01, 12 September 2018 (UTC)
 * True, but how many people do we have with unbundled privs as compared to 2014? Maybe there's more work happening that makes admin work less in demand than 2014. That's hard to know of course. More abstractly, the pattern you highlight will very likely continue. It's the natural course of things. As the project ages, the nature of the editing base will change with it. Actual article creation will slowly decline, and more gnomish behavior will increase, for example. As we asymptotically approach completing the project, there will be less and less large work to do. The Foundation has done an extremely poor job of anticipating the changing nature of its projects as they age. Instead, they anticipate perennial growth, and are astonished when it doesn't happen. Your excellent RFA by month does wonders to illustrate how we have changed. The heyday of Wikipedia is 10 years in he past, and it's not coming back. --Hammersoft (talk) 12:42, 12 September 2018 (UTC)
 * There is also more automation, be it bots or scripts, that reduces the workload. In fact, many backlogs (such as GAN) are from lack of editors and not lack of admins. Regards So  Why  14:01, 12 September 2018 (UTC)
 * The big unbundlings, Rollback, File Mover and Template editor all came long before 2014, I'm not sure how they effect this, other of course in that some ancient Admins may have got the bit to get one of those rights and never much used the other bits.  Ϣere Spiel  Chequers  20:00, 12 September 2018 (UTC)
 * Are there statistics on delete, block, and protect - the three permissions that are the most quintessentially admin? Deryck C. 13:06, 12 September 2018 (UTC)
 * Related; ~2 years ago I asked for a bot to keep track of certain backlogs of things needing admin attention. wrote a bot to answer that request, but they've since gone semi-retired and the bot no longer functions. The source for it can be found here. Its output can be seen here. It was somewhat of a proxy to see how much load we have compared to admin availability. --Hammersoft (talk) 17:10, 12 September 2018 (UTC)

Off-wiki links
I noted this at the current RfA, but I’ll post here as well: I think both Jbh’s RfA and the Kevin’s RfA both had opposes based on off-wiki behavior, and twice it was linked to, both times by individuals who are either a current admin or eligible for the bit (courtesy ping to ) WP:DOX covers this quite clearly: I know that complained about this on his talk page, and while  is gracious enough to try to explain it, it doesn’t change the fact that outside of RfA, similar material has been oversighted in the past, and really shouldn’t be allowed at RfA simply because it would be the definition of the Streisand effect to try to get rid of it. TonyBallioni (talk) 21:24, 27 August 2018 (UTC)
 * Why would it be allowed at RFA? AFAIK WP:OUTING is taken very seriously in every corner of Wikipedia and it should always be oversighted as soon as it is detected, and the perpetrators warned. &mdash; Amakuru (talk) 21:56, 27 August 2018 (UTC)
 * It’s not. The issue here is the candidate has already addressed it and acknowledged it, so redacting would generally cause more disruption, and in the case of Jbh RfA, redacting would have drawn significantly more attention than just letting it be. The high profile nature of RfA basically creates a scenario where linking to off-wiki internet accounts is very difficult to respond to after the fact, and candidates are put in a situation of having to deal with opposition based on information that is arguably oversightable, which is unfair to them if the information shouldn’t have been posted to begin with. TonyBallioni (talk) 22:05, 27 August 2018 (UTC)
 * This sort of behavior should absolutely not be tolerated at RFA (or anywhere else on the site). As you say, at this point however, now we can't go back. SQL Query me!  22:27, 27 August 2018 (UTC)
 * Yes, I see what you mean about redacting causing more harm than the original issue, Tony. We should probably at the very least add a prominent warning about outing policy in the box at the top of the RFA page. And if a question or vote is posted that breaks the policy, and hasn't yet been responded to, to remove it immediately, informing the person who added the text why you have done so. I guess if someone spots something off-wiki that makes them personally want to oppose the candidate, we can't really stop them making an oppose, but they would be forbidden to allude in any way to the off-wiki reasons for their oppose. &mdash; Amakuru (talk) 22:34, 27 August 2018 (UTC)


 * More or less copying a comment I made at the RfA. Well, hadn't realized this would turn out this way. My apologies for the same; if Kevin wishes, I'll strike both the question and my !vote at this RfA. While I'm not sure whether that would help, like I said, did not realize this would be perceived in such a way. Lourdes  03:17, 28 August 2018 (UTC)
 * I agree that outing is not desirable and not a sensible action on the part of experienced contributors (although current warnings about it are not prominent). Having said that, in this case the person "outed" has confirmed that they did something they shouldn't have done and subsequently (albeit innocently) tried to conceal it. I don't consider that either person needs disciplining; they just need to avoid doing such things again and to help alert others to the dangers of doing so. We all make mistakes - yes, even administrators. Deb (talk) 11:58, 28 August 2018 (UTC)
 * Such 'mistakes' seem to be increasing in frequency though and they place a candidate in an unfair and untenable position – they can ignore it, and have their RfA suffer for it; they can complain about it, and have their RfA suffer for it; they can address it off-wiki, and have their RfA suffer for it; or they can be open about it, out themselves, and have their RfA suffer for it. All in all, since a candidate does not have reasonable recourse to the normal processes one has when faced with outing or off-wiki harassment there needs to be, from this point forward, a clear statement that such will not be tolerated and will result in immediate blocks – prevention by example if you will. Candidates are vulnerable enough and the community must step in when things start to go off the rails, which linking off-wiki is by definition. A candidate at RfA should be expected to demonstrate the ability to deal with 'bad behavior'. Becoming a punching bag for people who do not want their poor behavior pointed out is not dealing and at some point the community expectation of 'restraint' becomes a community expectation that one participate in one's own abuse. This is toxic in itself and, in my rather unwavering opinion, contributes greatly to the perception that RfA is a grueling process rather than an examination of one's ability to be trusted as an administrator.  Jbh  Talk  13:00, 28 August 2018 (UTC)
 * , I agree that no discipline is needed. My point in raising this here is that in 2 of the last 3 RfAs, we've had breaches of the outing policy by people who should have known better. That is unacceptable and needs to stop: we can't allow this to be the new norm.Re: the person acknowledging it: the system at RfA encourages people giving in to outing because it moves so fast that if they ignored it and tried to get it oversighted, because they may feel that it would create significantly more controversy and have a negative impact on the RfA. I think Amakuru's suggestion of noting it in the edit notice could be helpful: something like Off-wiki information should not be presented unless a candidate has made an explicit link to it on-wiki previously. could work. TonyBallioni (talk) 13:12, 28 August 2018 (UTC)
 * +1. Innisfree987 (talk) 17:40, 28 August 2018 (UTC)
 * +1. SarekOfVulcan (talk)  17:55, 28 August 2018 (UTC)
 * This is better than nothing but it seems like a violation of OUTING in such a high profile forum deserves a stronger rather than weaker response. Otherwise when there are bad intentions (which I don't think there was on this RFA) it creates an incentive to do it anyway. I think this not becoming the new norm is incredibly important. Best, Barkeep49 (talk) 18:27, 28 August 2018 (UTC)
 * I totally understand that. However, it doesn't look like the current RfA is suffering as a result of the outing. I don't really understand what was going on behind the scenes with the previous no-consensus RfA, but I note that the candidate had at least as good an edit history as the current candidate and yet was criticised for lack of content creation. Deb (talk) 14:02, 28 August 2018 (UTC)
 * I agree that neither of these issues really affected their respective RfAs, but permitting "innocuous" violations of our OUTing policy because of good intentions and negligible effects is a terrible precedent to set. OUTing normally results in a block; make it clear that linking off-site is outing and apply a block the next time this happens. Oversightable information gets emailed to the ArbCom, and I see no reason this should be different. If the problem is serious (e.g, L235's supposed COI), ArbCom should be made aware of it anyway; if it's not serious, it probably doesn't matter. 's outing was clearly motivated by a vendetta and should have resulted in a permanent block.
 * At the end of the day, you can always say "Oppose because of off-site conduct with on-site implications" and leave it at that. —Compassionate727 (T·C) 16:08, 28 August 2018 (UTC)
 * Considering I was unwilling to allow innuendo and misquotation to magnify the problem and acknowledged the off-site posts there is an argument to be made that it was not technically outing. It was off site harassment which deserves an indef but I would have to start up an ArbCom case and draw in the several editors who have been subjected similar tactics (albeit only on-wiki to my knowledge) which would just turn into a drama circus for no sure outcome. All-in-all not worth my effort to start – I have more pleasant and productive things to do like pulling all the thistles and brambles from my garden bare handed.  Jbh  Talk  18:55, 28 August 2018 (UTC)
 * Your argument that it isn't technically outing is not cogent: if someone outs you, and you acknowledge the content of said outing in a reply, you have not retroactively un-outed yourself. You're addressing the fallout, not endorsing the use of the nuclear bomb. And opening such a loophole makes about as much sense as the argument itself. As for what requires ArbCom, I'm not sure and try to stay away from that, but I agree the level is generally "too much". This should be a case where a singular admin blocks, then the blocked user either 1) throws a tantrum and leaves ANI unamused or 2) agrees to not do it again and the block goes away. Crikey, even the nastiest ANI threads waste less community time than the simplest ArbCom cases (which are somehow never simple). —Compassionate727 (T·C) 19:21, 28 August 2018 (UTC)
 * I do not really disagree. I was only addressing the counter-argument which had come up elsewhere. As I remember the sequence was something like an editor made statements about me saying 'x' "off-site", where "x" was false and, after another editor commented that 'x' was a 'bad thing' I said 'no, they are talking about WO and what is being said is false'. So, I think, I was the first to mention the thread was on WO and thereby 'outed' myself. Later, an admin quoted something I said out of context and it spiraled from there. I probably broke some rule by copying the thread over to WP but by that point I'd rather be hanged for something I did vs something untrue. Anyway, the sequence of events can, in my opinion, support harassment but since I think I was the first to actually name the site rather than imply it, it would be very hard to get a consensus that what occurred was OUTING ie linking an off-site account to a WP user without their consent. It is also why I would strenuously object to allowing votes of "per behavior off-site..." because it leaves one wide open to a similar strategy ie one which is impossible to refute without self-outing or, even to votes which are bare-faced lies where there is no 'off-site' eg (oppose per off-site stuff -&gt; I never posted off-site -&gt; yes you did -&gt; prove it -&gt; can't that would be OUTING -&gt; but... -&gt; result=RfA disrupted.) So I strongly feel any reference to off-site anything needs to be bright line prohibited. There is simply no acceptable reasonable way to deal with such accusations.  Jbh  Talk  22:06, 28 August 2018 (UTC)
 * I think severity has to weigh in somewhere. Suppose, to use an extreme hypothetical, a candidate is a poster on a neo-nazi website where they have bragged about how they are able to push their POV on Wikipedia and solicited other racists to assist them.  That seems like important information for the RFA to have. --B (talk) 18:05, 28 August 2018 (UTC)
 * Then send it to arbcom. Natureium (talk) 18:12, 28 August 2018 (UTC)
 * • One can always find an edge case to justify doing nothing, it seems to have become a regular art form on Wikipedia. Well past time to stop with such arguments. Rather do something and use IAR to sort out the edge cases if they pop up – that is what it is for. Jbh  Talk  18:41, 28 August 2018 (UTC)
 * The statement that "Jbhunley's outing was clearly motivated by a vendetta" appears unproven; maybe it's fortunate that I'm not familiar with the background. However, I'm certainly not suggesting that we should fail to take action when there is repeated failure to abide by guidelines and policies, especially when that failure appears intentional. In recent months I've felt obliged to keep tabs on a number of people whose edits varied from the inconsiderate to the downright dangerous. Yet I fail to see what would have been gained by blocking either of the parties involved in the current "outing" "incident", since it seems highly unlikely that either of them would make the same mistake again. Deb (talk) 19:11, 28 August 2018 (UTC)
 * Once people started misquoting and complaining the thread had been moved into a non-public area I posted a transcript here. I can prove, with diffs, that every statement was false or misleading and made with the evident purpose of calumny by false light. Per wiki-civility norms saying those statements were the result of a vendetta would probably not be proper – saying they have the appearance of being the result of a vendetta though is, I believe, quite sustainable. Saying they were off-site harassment is, also, I believe, quite sustainable.  Jbh  Talk  19:31, 28 August 2018 (UTC)
 * FWIW, policy does not preclude us from identifying a duck, although we can spend all day arguing about whether that identification would be correct, and for that reason alone I'll concede I was out of line. I'm irritated about other misconduct that (maybe) happened during 's RfA that was more consequential, not that it excuses anything, nor does the fact that I'm upset about it matter. —Compassionate727 (T·C) 20:33, 28 August 2018 (UTC)
 * It feels like there's an opportunity to do something to uphold an important policy on a highly visible page even if that something is not blocking a good faith editor (as it seems to have clearly been in this case). But I think it's important that we agree that action is required going forward so we don't open ourselves up for abuse by a bad faith editor and we don't put a nominee in a bad position even if done from a good faith editor. Best, Barkeep49 (talk) 20:17, 28 August 2018 (UTC)
 * Having read that transcript, I'm unimpressed with the conduct of any of the participants - but most of us have been there. Maybe there's a vendetta, maybe there isn't; I think it's unprovable. Moreover, antisemitism, when kept to oneself, isn't a reason to discipline an individual; nor is it a reason to deny them adminship. There are far worse racist and politically-inspired contributors who do allow their personal prejudices to rule their edits (if you want to know who I mean, you only need to look at Talk:Jörg Kastendiek. Raising the issue on the NPOV noticeboard generally achieves nothing. This is the kind of thing, as admin, that would have had to deal with, and there isn't always an easy answer. Deb (talk) 08:16, 29 August 2018 (UTC)
 * Seems you missed the point: "antisemitism, when kept to oneself, isn't a reason to discipline an individual; nor is it a reason to deny them adminship" - the claims were flat out lies made for the express purpose of shit-stirring. Random trolls are one thing but a prolific Wikipedians with such an untethered relationship with the truth is a problem. Regardless, considering your and other's responses, the RfA quite clearly had the right result - I am temperamentaly unfit to be an admin if it means not responding off-wiki using whatever strategy I feel is appropriate to manage off-wiki attacks. Maybe I'm polite, maybe I try to defuse, maybe I'm mean or even slip and call someone an asshole. I did all of those things but there is no call for someone to bring that on-wiki period. Especially not with passive aggressive 'ooo he said something naughty off-wiki' or out of context quotation like was occurring before I posted the whole thing - yeah that was a questionable move but if people are going to judge me for off-wiki behavior I prefer that have the whole context - probably evidence of a character flaw on my part but... The more I consider it the more annoyed I get with the apologists, or even the 'fair-minded', who think Collect's behavior was anything but contemptable(Linking it this time ), it was a malicious attack fueled by a grudge bitterly held for more than three years - and even then was based on imagined slights, flat out misrepresentations and a couple lies dearly held until they seem to have become real in his mind. Again, that is not my interpretation - there are not "two sides" here. If you think I exaggerate or want me to prove my assertions (so as to avoid accusations of PA) please feel free to email me or ask on my talk page. I will bow out now so as not to further derail the general question with my specific complaint, as well as to avoid intemperatly giving my further opinion on the matter. There are absolutes and some things are just wrong. -- Damn, I guess I actually am still pissed off about the whole thing... go figure... I think I'll find something else to do for a bit.  Jbh  Talk  12:18, 29 August 2018 (UTC)
 * After reading that through I want to make clear my vehemence is not directed at you. It's just an upwelling which, for my internal equilibrium at least, needed to be said. Replying to your post just opened the spigots, so to speak. I just wanted to make that expressly clear if it was not. Jbh  Talk  12:29, 29 August 2018 (UTC)
 * It was clear that the injustice of it was what upset you. However, I supported your candidacy and perhaps I should make it clear that I would have done the same if I had been aware of the accusations because (a) there was no substance to them and (b) if there had been any substance to them, there was no evidence that you had allowed your personal beliefs to influence your editing. Deb (talk) 13:53, 29 August 2018 (UTC)
 * For what I hope are obvious reasons, I've never gone to look at whatever happened offsite in relation to these two RfAs, so I don't personally know or want to know the extent to which anyone was actually outed. It seems to me that what happened to JBH was deliberately malicious and should have been dealt with harshly. What happened to Kevin may perhaps have been a careless mistake – I certainly don't see any evidence of malicious intent. The community has a clear consensus that outing, as it is carefully defined at WP:Outing, should be subject to zero tolerance. But there is also a gray area where an editor's offsite activities might not reveal "personal information" as the outing policy defines it – and we need to be thoughtful about not taking an indiscriminate zero tolerance for everything that kind of looks like outing but isn't. Again, although I think that I can assess intent in these two cases, I'm deliberately ignorant about the nature of the information that might or might not have been revealed. --Tryptofish (talk) 21:31, 28 August 2018 (UTC)

Survey
We seem to have consensus to do something, so let's start simple and before everyone forgets about this thread. et al. suggested we add Off-wiki information should not be presented unless a candidate has made an explicit link to it on-wiki previously. to the RfA edit notice. Formal discussion and vote? —Compassionate727 (T·C) 20:33, 28 August 2018 (UTC)
 * ALT1: It is forbidden to post off-wiki information unless a candidate has already made an explicit link to it on-wiki. Enterprisey (talk!) 20:42, 28 August 2018 (UTC)
 * Oppose per WP:CREEP. It is quite unclear about the scope of such restrictions.  For example, the current RfA candidate seems to be active on IRC in Wikipedia channels.  Is that in scope or not?  In the recent RfAs, !voters say things like "I've met Philafrenzy at the pub before, and he seemed ..."; "I have known Philafrenzy for several years, mainly through..."; "I have had lots of conversations with this candidate at "; " I met Kevin at Wikimania in Montreal..."; "I know L235 from my ArbCom days, where..."; "I have also enjoyed meeting Kevin in person at Wikimania 2017. His participation in offline Wikimedia activities...".  When a candidate is engaged with other editors in multiple ways and across multiple projects, it is quite hard to separate the information gained into such a narrow frame.  Everyone is very connected nowadays through many channels – Facebook, Twitter, email, IRC, in-the-flesh, &c.   !Voters should not be punished for saying what they know and how they know it.  The alternative would be worse - a back-channel whispering campaign with any accusations being kept from the candidate. Andrew D. (talk) 21:04, 28 August 2018 (UTC)
 * "!Voters should not be punished for saying what they know and how they know it." - Um, what? If they violate DOX, then absolutely they should be. Whether it's "accidentally" or they deliberately perpetrate a vendetta-based smear-campaign, like the one we just witnessed at RfA this month, then they certainly should be "corrected" (since we dont call it "punishment") as per the block/ban policy.
 * "The alternative would be worse - a back-channel whispering campaign with any accusations being kept from the candidate." "Worse" how? A couple of trolls blading someone via email, or on another site that can't be mentioned here anyway, can't have the same effect as some of the blatant, (and unchecked btw) bell-can't-be-unrung, Streisand-effect, RfA-tanking, worthy-candidate-sinking nonsense that has taken place here lately. So let's control what we can control. Post a stern warning for violations, and any unique situations that "CREEP" in can be dealt with on a case-by-case basis. - wolf 14:23, 30 August 2018 (UTC)
 * Support either original or ALT1, as a reasonable restatement of existing policy that would forbid nothing new despite what Andrew says (also, IRC logs are already suppressable, fwiw.) TonyBallioni (talk) 21:21, 28 August 2018 (UTC)
 * I'm fine with ALT2 as well, but ALT0 would be my first choice as it is how the policy linked in ALT2 is usually applied (i.e. no linking to off-wiki profiles.) I wouldn't want it read as an excuse for people to say that policy allows for off-wiki linking, which in the overwhelming majority of cases it simply doesn't. TonyBallioni (talk) 23:50, 28 August 2018 (UTC)
 * I very much appreciate the intention here, but I think such a change would be a bad idea. First of all, simply referring to or linking to something offsite is not, de facto, outing. Outing occurs when an editor's real-life identity or personal information ("personal information" is defined at WP:Outing) is revealed against the editor's will, or when there is an attempt to do so. And the fact is that WP:Harassment covers what policy needs to cover. The community should have been quicker to seek removal of some posts in the two most recent RfAs, but the solution to that is to be quicker. There are times when raising an issue of offsite conduct, raised without violating policy, is a valid part of an RfA comment. --Tryptofish (talk) 21:22, 28 August 2018 (UTC)
 * For clarity, that means that I oppose both ALT0 and ALT1. --Tryptofish (talk) 22:20, 28 August 2018 (UTC)
 * If an editor has linked to an off-wiki identity on-wiki, then it isn't covered by outing, but we traditionally have considered linking to other internet accounts to be prohibited by the policy as well unless the editor has made the connection themselves. As I mentioned at the current RfA, ArbCom viewed it this way on at least one recent case. The issue is that some people solely think the harassment policy applies if I say something like Look, is John Doe who lives in Montreal and is faculty at XYZ University., which is what I think led to the linking at the recent RfAs. The way one oversighter described it to me was "If I can't tell it from looking at someone's user page, it's likely covered." The issue here is that if there are off-wiki postings of concern in the majority of cases, ArbCom is best to handle it. TonyBallioni (talk) 22:10, 28 August 2018 (UTC)
 * If there were some other site at which I posted as Tryptofish without self-identifying in any way (there isn't), linking to it would not be outing (but could be foolishly giving in to a false flag). It could however be regarded as harassment, depending on the details. And handling it through functionaries is, as you correctly say, a preferable (and if it constitutes some form of harassment, required) approach. On the other hand, the example you just gave most certainly is an outing violation, whether or not I actually am that person. In any case, the bottom line as I see it is that WP:Outing, as it exists, should be applied at RfA, but this doesn't require creating any new policy. I would be fine with simply making it clear that it applies here, per what Natureium says just below. --Tryptofish (talk) 22:20, 28 August 2018 (UTC)
 * The same bit of the harassment policy that deals with outing deals with this type of conduct (I quoted it above), and from a basic perspective, we usually handle it the same way: alert the oversight team and/or ArbCom and let them decide the best way to handle it in a given circumstance. I get the distinction you are trying to make, but I don't think its a very practical one since it invites people to test the limits of the harassment policy when our response in most cases is the same.In the example you just gave, regardless of whether or not you had personal information on the Tryptofish profile on another site, someone should not link to it on-wiki, and if the information is pertinent, they should email functionaries or ArbCom, not directly link to it in an RfA (or other on-wiki venue) and let people scramble to figure out how to deal with it after the fact.Re: creating a new policy: I don't see anything here doing that as we currently apply policy. I prefer either of these versions to spell it out for people, but would be fine emphasizing that the harassment policy applies at RfA if neither of those gain consensus. TonyBallioni (talk) 22:31, 28 August 2018 (UTC)
 * At this point, I think we are splitting hairs about what happens with a link to another site that contains no "personal information" as defined by the policy. In any case: don't post it, send it to a functionary instead. But if anyone misled you into thinking that it can be outing without any "personal information" or something that looks like it, that's not what the policy says. If you don't believe me, go read it. It's all about "personal information". There's also the issue of "opposition research" described there, which is why any such linking is a bad idea. And the more we discuss this, the more strongly I oppose the proposals here. There is a real danger in criminalizing all mention of offsite material. Editors can weaponize an innocent mention of something offsite to get an advantage over the editor who mentioned it. Believe me, I've had experience with this. Let's say I saw an RfA candidate whom I had previously observed posting at Wikipediocracy in an effort to get others to dox me – would you really want to say that I couldn't mention this (not necessarily linking to it) as a reason to oppose that RfA? I agree with the editors who see instruction creep in these proposals: what happens when there is a subtle difference in wording from the existing policy? Better to use the existing policy, as it exists. No one can validly claim that RfA is a policy-free zone where the harassment policy cannot be enforced. --Tryptofish (talk) 22:55, 28 August 2018 (UTC)


 * I think it should be emphasized that WP:OUTING applies at RfA the same as it applies everywhere. Natureium (talk) 21:28, 28 August 2018 (UTC)
 * No, because it's not black and white as others have said per WP:CREEP, and WP:AAAD. Those who can suppress already have policy and means to do so; and per AAAD (argument 1), off-wiki is an admin suitability criteria to consider, especially since off-wiki is an administrative concern for the wiki in multiple areas from COI to the most recent twitter case that resulted in Administrative Noticeboard discussion and Administrative sanction then later went up to Arbcom. Alanscottwalker (talk) 21:56, 28 August 2018 (UTC)
 * Support Alt1 but really support Tryptofish. The answer here is to have removed the content which was trouble quickly so that it didn't become a part of the RFA to the extent that removing would have potentially harmed the candidate. Best, Barkeep49 (talk) 21:54, 28 August 2018 (UTC)
 * Support original wording, ALT1 second choice. -- SarekOfVulcan (talk) 22:02, 28 August 2018 (UTC)
 * Support ALT1 or ALT2. If someone uncovers something particularly grievous off-wiki they can email the candidate and/or notify Arbcom, but posting it in the RFA and forcing the candidate to publicly confirm or deny a link is an unfair practice. Brad  v  22:11, 28 August 2018 (UTC) I originally supported ALT1, but was convinced by Tryptofish below to add my support to their version.  Brad  v  00:18, 29 August 2018 (UTC)
 * Comment The issue principally is only about Editors who have publicly tied their Wikipedia usernames to other online or offline activities.Pharaoh of the Wizards (talk) 22:36, 28 August 2018 (UTC)
 * Are you missing a not? Because this is about WP:OUTING, not about links to information that candidates have disclosed on-wiki. Brad  v  23:09, 28 August 2018 (UTC)
 * Strong support either alt but I do think it important we get this message on to the page. Linking of undisclosed off-wiki information is very definitely outing, even if the usernames are the same and even if the user has publicly admitted their real name and location. This note is not instruction creep because it simply reminds users of a policy that is already in place, and will serve to prevent recurrences of the examples already mentioned here, which were well-intentioned mistakes. &mdash; Amakuru (talk) 22:53, 28 August 2018 (UTC)
 * Just to add, I would also be happy with ALT2 if that's the only version we can get consensus for, and certainly linking to the policy page is a good idea. But like Tony I am concerned that restricting it to the term "personal information" may give the impression of wriggle room which doesn't exist. Tony's diffs above show that two sitting arbs have deemed the recent links to off-site forums during RFAs to be unacceptable. My angle on that is that although an editor may appear to be just anonymous on the off-wiki site as they are here, the internet is a heavily-linked sponge of information, and nuggets of information found on another site may lead someone to yet more websites, or other clues to identity which weren't available with on-wiki information. Plus the risk that the user with an identical handle on another site is not actually the same person, or worse still an impersonator. Far better not to go down that route at all. &mdash; Amakuru (talk) 14:14, 30 August 2018 (UTC)
 * What's important to recognize is that, in those two examples, the reason that the Arbs said that was precisely because there was (apparently, because I haven't looked myself) that kind of "bread crumbs to be followed" situation that could lead to personal information. And I think we all agree about that. I do. And that's covered by the existing outing policy – because it involves personal information. But there are absolutely some kinds of "information" that are not "personal information". Would anyone argue otherwise? So what I am advocating is something very simple: just use the same terminology as the existing policy. If the existing policy has wriggle room, then fix it there. But it seems to me to be very simple to use the same language as the policy here. And it puzzles me that editors would want to fight to use language that is different than the existing policy, a fight that strikes me as not worth having. --Tryptofish (talk) 19:27, 30 August 2018 (UTC)
 * ALT2: "It is forbidden to post off-wiki personal information unless a candidate has already made an explicit link to it on-wiki."
 * This is an alternative that I could support. The only difference from ALT1 is that I changed "information" to "personal information" and linked the phrase to WP:Outing, which defines the term. With this change, it really is restating existing policy. But ALT0 and ALT1 do not restate what the existing policy says: they extend it. --Tryptofish (talk) 23:05, 28 August 2018 (UTC)
 * Support ALT2 (first choice), support original wording (second choice), oppose ALT1. While the specific purpose of WP:OUTING is obviously to prevent personally-identifiable information from being involuntarily released, and accounts on other websites aren't necessarily considered to be "personally identifiable information", I think most of us would not want them involuntarily disclosed on-wiki. I appreciate that we should err on the side of privacy considerations, and adopting the originally-proposed text as a general rule seems reasonably within the spirit of the OUTING policy (though ALT2 is more in line with what policy actually says). I would fairly strongly oppose the ALT1 wording, because the theoretical possibility to make an exception should never be prohibited, and ALT1 seems to attempt to draw a hard line against even that. We will of course always continue to draw a hard line on actual doxxing either way, but as points out, in some extreme cases, it might be entirely reasonable point out, or at least question, a Wikipedia user's link to an off-wiki account, for the benefit of the project (e.g. "hey just wondering, are you this neo-Nazi that goes by the same username?"). We need to approach this with common sense and not overcorrect.  S warm   ♠  23:40, 28 August 2018 (UTC)
 * Just for the record, WP:OUTING also has an exception paragraph. (As to common sense in general, pretending something does not exist does not make it not exist). -- Alanscottwalker (talk) 23:51, 28 August 2018 (UTC)
 * No preference - which is to say I'm happy with any of the suggestions, but I don't think it resolves the issue if and when outing occurs. Deb (talk) 08:05, 29 August 2018 (UTC)
 * Support ALT2 and support procedural RfX renomination of any editor who has been violated by outing. w umbolo   ^^^  11:51, 29 August 2018 (UTC)
 * Support Alt2 My concern, though, is how will we handle it when this arises?  I mean, we're talking about an edit notice, right?  Does anyone actually read those?  Or alter their behaviour based on them?  I suspect not, and therefore another outing will occur.  What will we do?  End the RfA, blank it, oversight the outing, then renominate the candidate?  Turn the offender over to the tender mercies of ARBCOM?  Have another big discussion, here or at ANI or elsewhere, which draws even more attention to the issue?  I don't have an answer, but we need to be aware that this sort of nonsense is going to continue happening. Happy days, LindsayHello 14:27, 29 August 2018 (UTC)
 * I'd prefer to oppose based on WP:CREEP, but moreso because we already have the policies and guidelines to take care of this. It is that is lacking. Decide how to deal with the situations that do arise (blocks, bans, revdel or suppression, as appropriate), don't make more rules that really just repeat that we don't want people to do this. --Izno (talk) 14:38, 29 August 2018 (UTC)
 * Just to repeat, the proposal is not to create new rules, but to put a reminder of the existing rules in front of people at edit time, so that they are reminded of them at the time they may be about to infringe them. You say enforcement is lacking, and you're probably right, but this isn't about punishment, as the rules are not being broken intentionally. Much better not to have the problem not occur in the first place than have to enforce it later,and this is a simple proposal to help prevent it occurring. &mdash; Amakuru (talk) 14:52, 29 August 2018 (UTC)
 * Is it just reiteration? The proposals above do not even link to the policies in question. There is dispute below that the intent of the proposed additions match the intent of our policies. If it's  a reminder, what's the point? What about banner blindness? --Izno (talk) 15:02, 29 August 2018 (UTC)
 * Myself, I often ignore edit notices, I have to admit. --Tryptofish (talk) 21:14, 29 August 2018 (UTC)
 * Support anything that makes potential outing more difficult. Obviously, if an "outee" approves of someone linking to their off-wiki activities, that is another matter, but this appears to refer specifically to posts that are made against the subject's wishes.
 * (Full disclosure: I had personal information posted on-wiki multiple times against my will, and was harassed off-wiki for months, by a troll who is, unfortunately, apparently still active, and yet no one has been willing to do anything about it. This circumstance perhaps makes me kind of ... more of a stickler for enforcing this particular policy than I might otherwise be.)
 * Also, Andrew's comment about how everyone's social media profiles are interconnected is gibberish. Yeah, for some Wikipedians this is the case, but for the vast majority it is not. It also reads like a (perhaps inadvertent) defense of "This person mentioned their Wikipedia activity on an off-wiki forum! They outed themselves!", which if not unintentional is outrageous and offensive.
 * Hijiri 88 ( 聖やや ) 15:19, 29 August 2018 (UTC)


 * Oppose  This is "creep" at its most inexplicable.  For example, where a person can be reasonable asserted to be a "known person" such as   this proposal would cripple discussions completely. In addition, known IRC "handles" and the like (where the others on IRC treat the person as the person the handle indicates) should be forbidden if we wish this sort of "rule" to be established. Lastly, where a person asserts themselves to be an editor on-Wiki, then if the on-Wiki editor says he is not that person, then such comments should be stricken, sure.  But how many cases of that have occurred in the last decade or more?   Collect (talk) 16:01, 29 August 2018 (UTC)
 * Just as a point of clarification: IRC handles generally must have a host associated with them in order for any action to be taken i.r.t. that account, and the application for a host requires on-wiki confirmation: would be able to provide more information on that than I can. If they don't, they will be asked to confirm it through an action on-wiki. IRC logs are also not allowed to be posted on-wiki. TonyBallioni (talk) 18:24, 29 August 2018 (UTC)
 * I think ALT2 is okay (except for stuff like discovering somebody is a paid editor), but I also think we are approaching this the wrong way. This is not relevant to RFA, it is relevant to every single page on Wikipedia. —Kusma (t·c) 19:02, 29 August 2018 (UTC)
 * Re stuff like discovering somebody is a paid editor, that is explicitly not a valid exception to the outing policy. As indicated at WP:COI, suspicions of that nature should be aired through the proper channels, via oversighters and arbcom. Posting explicit information about COI editing would only be admissible at RFA if it is confirmed on-wiki. &mdash; Amakuru (talk) 14:59, 30 August 2018 (UTC)
 * Support - Prefer Alt1, but anything is better then nothing. - wolf 14:25, 30 August 2018 (UTC)
 * This is not necessary; we already have rules about linking to presumed off-site identities anywhere, so we don't need extra rules about doing it in RfAs. Formal sanctions for bad behavior on Wikipedia are unavoidably slow and patchy - for good reason, we all try to stick to the principle that blocks are preventative, and if a block isn't going to solve the problem then it's not really practical to maintain lists of demerits like a band of roving assistant vice principals. But in many ways the most effective consequences for bad behavior are social, not the result of formal action. The logical consequence of doing something this boneheaded is that people may think you're a bonehead. Violating the privacy of an RfA candidate (already a stressful position), especially over something as minor as the L235 and Jbhunley examples, and even more especially to ask a cheap "gotcha!" type question like the one posed to L235, is just plain piss-poor judgment. What happens when you do something like that is: people like me who hadn't closely followed your work but had generally positive impressions learn of this kind of behavior and think "geeeeez, that was a totally unnecessary asshole move". People are surprised at you, they're disappointed in you, and they lose confidence in your judgment, and even if you apologize and back off, you're still the person who in the very recent past thought that playing internet detective on someone's RfA was a smart idea. (I'm kinda digging in on ' example in particular because "hadn't realized this would turn out this way" is a pretty weaksauce apology from someone who passed an RfA themselves and damn well should have known better.) So in short - no, we don't need more rules; I'm not even sure we need more "enforcement"; what we need to do is really, robustly recognize that this kind of thing is Not Done not just because there's a rule about it but because it's a WP:JERK thing to do. Hopefully somebody reads all this mess and thinks "huh, yeah, I'd feel pretty bad if someone linked that dumb stuff I posted on reddit/twitter/whatever five years ago in the middle of my RfA, those things are stressful enough as it is" and then "you know what, I'd also feel like an ass if I linked that dumb stuff someone else posted on reddit five years ago and tried to embarrass them about it" and then we're one person closer to a culture where we don't have to write rules about how outing is bad specifically in RfAs because you'd feel like an ass for doing it and thus you don't do it. Got it? Opabinia regalis (talk) 18:53, 1 September 2018 (UTC)
 * Hello (and others who've said as much), might I reiterate my apology here? I do feel like a jenny linking that stuff and it's a lesson that'll remain well embossed. If you'd like me to release a clearer statement here or at any other page, I would surely do so. Warmly,  Lourdes  00:21, 2 September 2018 (UTC)
 * Well, I'm just some rando who has no involvement in the situation and just doesn't want this to lead to more instruction creep; it's not really what I'd like that matters. But as an opinion from some rando, I'm not concerned about where the apology is located, and I'm not much for prepared statements - but what was missing for me was thinking beforehand about what harm might have been caused (which can be hard to predict from an isolated link) or recognizing after the fact that it could have caused harm, even though it luckily didn't in this instance. Of course I have no idea if you've had those thoughts, or sorted it out off-wiki, or whatever, and I have no interest in being or encouraging the Apology Police, but what did get verbalized on-wiki has a bit of a "oops, but no harm, no foul" quality. Opabinia regalis (talk) 04:48, 3 September 2018 (UTC)
 * Understood. Lourdes  05:08, 3 September 2018 (UTC)
 * Support  Hawkeye7   (discuss)  10:51, 2 September 2018 (UTC)
 * Support making it explicit. There's no rule creep here, as posting off-wiki links to alleged accounts/publications/etc by a Wikipedia editor which they have not already disclosed is already forbidden by outing policy. Boing! said Zebedee (talk) 11:06, 2 September 2018 (UTC)
 * Support. Recent RfAs have proven that this is necessary. —  python coder    (talk &#124; contribs) 20:36, 6 September 2018 (UTC)

Procedural concern
It seems to me that an informal "survey" among editors who watch here is not sufficient to enable the proposed changes. Maybe to determine the best proposal to make formally via a subsequent RfC, but not to actually make a policy-related change to the RfA page. If any such proposal is to be implemented, I think there needs to be a formal RfC, open to the community. --Tryptofish (talk) 22:58, 28 August 2018 (UTC)
 * This is just a restatement of existing policy... TonyBallioni (talk) 23:00, 28 August 2018 (UTC)
 * No, it isn't. --Tryptofish (talk) 23:00, 28 August 2018 (UTC)
 * We aren't going to agree with this, but it 100% is. We do not allow linking to identities, real or perceived, unless someone has linked to it themselves. TonyBallioni (talk) 23:02, 28 August 2018 (UTC)
 * Tony, you know I like you, but you are wrong. Linking to identities is the key thing here, but ALT0 and ALT1 just refer to "information". That doesn't just mean "personal information". I'll repeat here something I said above: If someone posts at Wikipediocracy trying to get people there to dox me (really happened, but nobody followed through on it), and then they become an RfA candidate (person who did that was a candidate, unsuccessful, but before the Wikipediocracy thing happened), I want to be able to oppose on that basis. Under existing policy, I could. Under ALT0 and ALT1, I could not. --Tryptofish (talk) 23:10, 28 August 2018 (UTC)
 * How would you know for certain that the editor on Wikipediocracy is the same editor as the one on Wikipedia? Brad  v  23:15, 28 August 2018 (UTC)
 * That's an excellent question. Yes, there is certainly a possibility of false flags, but that is something an RfA candidate can clear up. If I were to oppose on that basis, that would be an entirely valid reason to reply to the oppose, in the way that opposes often attract a lot of replies. But it wouldn't be a valid reason to block me. --Tryptofish (talk) 23:21, 28 August 2018 (UTC)
 * As I see it, that's the main issue here. In the current RFA, links were provided to statements off-wiki, and the candidate was then forced to confirm or deny the connection. That's an unfair practice in the middle of an RFA — that confirmation should have happened privately, if at all. Brad  v  23:24, 28 August 2018 (UTC)
 * And I agree with that, with the proviso that the current RfAs have involved "personal information", not miscellaneous information. --Tryptofish (talk) 23:27, 28 August 2018 (UTC)
 * Then we may need to define what that "information" is. The definition you linked to only lists legal name, date of birth, identification numbers, home or workplace address, job title and work organisation, telephone number, email address, other contact information, or photograph, whether such information is accurate or not, which does not cover usernames on other websites. Brad  v  23:30, 28 August 2018 (UTC)
 * You see what you just said? What I linked to was the existing policy. If you want to create a new definition, then you would be creating new policy, and I insist on an RfC for that. --Tryptofish (talk) 23:33, 28 August 2018 (UTC)
 * Hoisted with my own petard. Okay, I'll support ALT2. Brad  v  23:36, 28 August 2018 (UTC)
 * Not so fast. See paragraph six of WP:OUTING. Except in narrow circumstances, linking to off-wiki accounts that haven't been publicly connected is also, separately, prohibited. ~ Rob 13 Talk 15:43, 30 August 2018 (UTC)
 * By my arithmetic, paragraph 6 is the paragraph that begins "Posting links to other accounts on other websites...". If I'm missing something, please set me straight. But I'm not seeing anything in that paragraph that actually says that if there is another website where, for example, I have been posting as Tryptofish (I haven't) without posting any sort of personal information, just the screen pseudonym Tryptofish, if someone links there, they would be outing me. It certainly does follow from the language that if the other site also contained something, anything, that could lead someone to something identifiable about me, then it becomes an entirely different matter, and I would agree with you completely that linking there or even saying that it exists without linking would be impermissible (subject to some very narrowly defined conditions, as you correctly say). So if linking to external sites involves personal information, I agree with you, but as for linking to external sites where personal information is in no way involved, I don't see any language in the policy that separately identifies that as a violation, although in some circumstances there could be issues of hounding there. --Tryptofish (talk) 18:51, 30 August 2018 (UTC)
 * To understand that paragraph, you probably need to understand the background of the policy. Previously, it stated that you couldn't link to other accounts except in exceptional circumstances. An RfC was held specifically to allow linking to sites related to paid editing, to allow investigations at COIN. That paragraph is the result, which now notes one approved instance in which you can link to external websites. The community's consensus on the change was to approve that one instance of posting to off-wiki accounts, so the approved instances are the previous wording (exceptional circumstances) plus one additional approved case (linking to job posting sites). ~ Rob 13 Talk 19:26, 30 August 2018 (UTC)
 * See here, especially after the break, and the follow-up RfC at this link. ~ Rob 13 Talk 19:31, 30 August 2018 (UTC)
 * I wonder if you remember who the editor was who initiated the discussions you just linked to. It was me! (Look at my comments in your first link, then go back to Archive 6, and follow the discussions from Archive 6 to Archive 9.) I even wrote most of the language that you are now referring to. So I'm pretty sure that I understand and remember the history about that. My entire argument here is that what we are discussing now should use the same language as the policy, which is "personal information", not all "information". Editors seem to think that I somehow believe it's OK to link to other accounts that contain or lead to personal information that has not been voluntarily posted onsite here. I never said that. --Tryptofish (talk) 19:47, 30 August 2018 (UTC)
 * I decided to go back and really check carefully what the history was of the sentence Posting links to other accounts on other websites is allowable in specific situations (but see also Linking to external harassment). It is not as describes it above – and I think that Arbs and other functionaries as well as administrators, as a group, really need to understand what the policy is, because they are responsible for enforcing it correctly. The RfC described above as establishing consensus for linking in certain COI discussions did nothing of the sort. It was started in June 2016 and led to what was essentially no consensus, being closed as "stale" without a consensus statement that December: link. As of that time, the sentence said "on a case-by-case basis", and was tagged as "under discussion": link, as it was for a long time. Here is the edit, made by me in April 2017, that implemented the language about "specific situations": . The subsequent edit that implemented the present-day language was made by me here: . The talk page discussion that established the consensus for that language had begun in January that year: link, in response to the decision by the WMF to change the Terms of Use regarding paid editing. Discussion of that point continued on and off from January (Wikipedia talk:Harassment/Archive 14) through April (bottom of Wikipedia talk:Harassment/Archive 15). It did not really play out as a change from never post links to other accounts, to never post links to other accounts, except for one specific case about COI. It was a change from a longstanding but undefined "on a case-by-case basis" to clearer language that reflected how to deal with the ToU for paid editing. "Case-by-case" had been on the policy page since 2015: diff. Prior to that, there was a consensus that "posting ANY other accounts on ANY other websites" was not outing: Wikipedia talk:Harassment/Archive 2. So: it is untrue that the community originally considered posting links to other accounts to always be outing, only to carve out an exception for COI. --Tryptofish (talk) 22:18, 31 August 2018 (UTC)
 * Nevertheless, linking to an off-wiki site with no evidence but similarity of usernames is not appropriate, and should be banned. If it takes an RFC to clarify that, then we should make that happen. Brad  v  23:39, 28 August 2018 (UTC)
 * You can make your own Oppose based on your suspicions, but you may not link to them. Even if no personal information is involved, It falls under the opposition research clause mentioned above. &mdash; Amakuru (talk) 23:18, 28 August 2018 (UTC)
 * I agree: make the oppose, but don't link. However, there has been a history of blocks for outing in which no link was made, just a comment that there is offsite personal information. So the key thing here is "personal information" as it is defined by the outing policy. If you make it any kind of information per ALT0 or ALT1, you are creating new policy. --Tryptofish (talk) 23:24, 28 August 2018 (UTC)
 * I know, I enjoy disagreeing with my friends the most . This would be a very tricky situation, and I definitely get where you are coming from, and I am not sure exactly how I would handle it. What I'm saying is that I know for a fact that ArbCom and the oversight team have interpreted linking to other social media/internet profiles in the past as oversightable, and that this was actually a somewhat significant issue at Arbitration/Requests/Case/German war effort, because of posts on an off-wiki site that were considered in evidence privately. I think how I would deal with the situation you mentioned would be to say something like Oppose, per off-site harassment I have experienced from this user, the proof of which I have mailed to ArbCom. I'll ping User:ArbitratorX to confirm they've seen it. or something of the sort. TonyBallioni (talk) 23:18, 28 August 2018 (UTC)
 * Something else occurs to me. In the specific context of RfA, an oppose worded that way would be very difficult for the rest of the community to evaluate. And let's just say that someone acting in bad faith posts exactly that oppose without specifying which Arb, and then a whole lot of other editors start to oppose based on that, but in fact nothing happened offsite. It can backfire that way, which is why we should not go beyond what WP:Outing requires. --Tryptofish (talk) 23:55, 28 August 2018 (UTC)
 * Good, then I hope you also enjoy being wrong. { But about handling it, I'll make it easy for you: just support ALT2, problem solved. Editors got a bit emotional about this situation (always happens with outing situations), and that led to hastily-written proposals that contain mistakes. Let's just get the wording right. --Tryptofish (talk) 23:31, 28 August 2018 (UTC)
 * I prefer ALT0, because in my view, it is how the policy in ALT2 is usually read, but I would not oppose ALT2. My concern is users using that wording and this discussion as an excuse to post material that is in the overwhelming majority of circumstances not allowed and is potentially oversightable. Like you mentioned above, we're splitting hairs here, but the takeaway is that people shouldn't be linking to profiles on other sites. TonyBallioni (talk) 23:46, 28 August 2018 (UTC)
 * How about ALT0 with the word "information" blue-linked to WP:Outing? --Tryptofish (talk) 23:50, 28 August 2018 (UTC)
 * Yes, I would support that. The point here is not to create new policy, but to emphasize that the existing harassment policy applies to RfA, and that if they have questions as to whether or not it is permitted, they should think before posting. TonyBallioni (talk) 23:54, 28 August 2018 (UTC)
 * Good, but please understand that, because the point here is indeed not to create new policy, it's important not to word it in a way that does. Even if the intention is good, and it just sounds like common sense that everyone would understand (yeah, right), the word "information" is so inclusive that it will eventually be gamed. --Tryptofish (talk) 00:01, 29 August 2018 (UTC)
 * I'm fine with that. I don't consider an edit notice to be creating policy, but that could be because I'm over involved with that side of Wikipedia that I don't always see how someone who isn't a project space denizen would read it. I think information while linking to the appropriate policy section splits the baby nicely. TonyBallioni (talk) 00:04, 29 August 2018 (UTC)
 * Oh, that poor baby! { --Tryptofish (talk) 00:10, 29 August 2018 (UTC)
 * I just noticed for the first time that the proposal is about the edit notice rather than about the page. Goes to show what happens when things are discussed in a hurry. It doesn't change any of my views, though. --Tryptofish (talk) 00:12, 29 August 2018 (UTC)

I do not think anyone should have been blocked here either. Both of the recent attempts were clearly in good faith, if misguided. I do think it should be clarified in the desysop policy that knowingly posting oversightable material is grounds for desysop (we've had 3 cases of people who have passed RfA posting material that could be suppressed in the last few months: two at RfA and the other being Andrevan: that is unacceptable), but this is more from the view that administrators should take special caution when doing anything that approaches OUTING.In terms of blocking for this at RfA: I think it would be appropriate if we made it clearer that this wasn't allowed and people continued to do it anyway. The Jbh RfA was a mess, and I'd forgive anyone who witnessed it for coming away with the impression that we allow linking to off-wiki forums and discussion of candidate behavior on other websites. Part of my reason for starting this discussion was that I wanted to raise the profile and issue that 2/3 RfAs this month have had issues with linking to candidate posts on other websites. It's more dramatic than I normally prefer, but I feel strongly enough about this issue that I think it is worth raising. We can't have situations where someone Google's someone's name, finds a post on [Scandalous forum here] by someone with the same name who isn't the Wikipedia user, and utterly ruins their reputation based on false assumptions. That is the real danger here, and the reason why we don't allow linking to other websites. TonyBallioni (talk) 19:23, 29 August 2018 (UTC)
 * , re: ALT2, would you be fine changing "personal" to "private" in the Wikilink? I actually think that would be the best of all three options proposed now (even better than your proposed compromise of Wikilinking on ALT0). TonyBallioni (talk) 14:55, 29 August 2018 (UTC)
 * No. The policy calls it "personal information", not private, and it makes it very clear that personal information needs to be kept private. So there's no advantage to changing it to private, and there is a significant disadvantage in that the phrase has not been officially defined by the community. --Tryptofish (talk) 18:14, 29 August 2018 (UTC)
 * That's fine. My concern is that it "personal" will be read too narrowly (as in multiple oversighters have confirmed  that off-site linking to profiles is covered under the oversight policy, and thus would be personal information), but I get your concern. I think we are approaching this from opposite angles: you don't want an edit notice to be weaponized ( a fair concern), and I don't want it to be used to excuse posting oversightable material by being read very narrowly, which is not how it is interpreted in practice. TonyBallioni (talk) 18:20, 29 August 2018 (UTC)
 * I just looked at those two diffs. You are exactly right in how you identify our differing concerns. But the diffs drew my attention to something more. Both of those Arbs said, correctly in my opinion, that it should not have been posted. But they both also, again correctly in my opinion, never said that the editor who posted them should have been blocked or otherwise sanctioned for having posted it. That latter aspect of it is something where personal experience has made me very wary of the weaponizing aspect of these things (not so subtle hint: take a look at my block log – one of the two Arbs you cite clearly has a more nuanced view today than they did back then). It's one thing to oversight something that should be oversighted, but another to treat good-faith mistakes as a hanging offense. I once held an RfC asking whether the community believes that "preventative not punitive" should apply, or not apply, to blocks for outing attempts. It was pretty much no-consensus, with remarkably strong emotions on all sides, but I was struck by how many editors felt so strongly that being punitive for even borderline cases was the right way to go, for reasons of punishment being preventative in the sense that if it's firm enough no one else will ever dare to do it. (In my personal and slightly dyspeptic opinion, that bears some resemblance to the medieval justifications for drawing and quartering.) --Tryptofish (talk) 18:55, 29 August 2018 (UTC)
 * Yeah, we really are not as far apart as it might appear. At this point, let's find out what everyone else thinks. --Tryptofish (talk) 21:11, 29 August 2018 (UTC)

What a complete and totally boring waste of time this entire talk page is. And this latest pettifoggery is a complete joke. Rfa, Rfa, Rfa....get a life! It's a pity that as much effort could not be put into improving articles and for admins., dealing with queues. Leaky Caldron  21:08, 29 August 2018 (UTC)
 * To reiterate: If the consensus here is ALT2, no problem as far as I'm concerned. But if anyone wants ALT0 or ALT1, I insist that this would require a community RfC. --Tryptofish (talk) 23:40, 28 August 2018 (UTC)
 * Of course, if the community would stop with their RfA fuckwittery, we could get some new admins through and helping with the queues. Just a thought, you know, a problem shared is a problem halved and all that cliched bullshit. Nick (talk) 21:17, 29 August 2018 (UTC)
 * Yes and if the community could stop with their BS linking to stuff off site during RfAs we wouldn't have to have this discussion either. Best, Barkeep49 (talk) 21:23, 29 August 2018 (UTC)

- "Yes, there is certainly a possibility of false flags, but that is something an RfA candidate can clear up." - wtf? Isn't this precisely one of the things we're trying to avoid? Some 'opposer' makes an accusation about something off-site, it turns out to be wrong, (whether it is was a legitimate mistake, or worse, deliberate), the damage is already done, and can't be undone. The candidate is stuck defending against this "false flag", some people will pop in, see the initial accusation, oppose and leave without ever finding out the truth of it. End result; the RfA is tanked. Blackstone's formulation can basically apply here; Better that a candidate, who may have once done something inappropriate off-site, but otherwise has a great résumé for admin, be given the tools, than a worthy candidate who has never done anything wrong, be denied them because of mistaken identity. Lastly, I agree with Amakuru; simply posting a reminder of existing policy is not creating new policy. - wolf 15:09, 30 August 2018 (UTC)
 * I do realize that the words I said in that quote were not really quite right, so I accept that. What I was trying to communicate is that there is a scenario in which the RfA candidate could promptly respond with something like "No, that's not me, it's an impersonation" and I'm very confident that the community would accept that on WP:AGF and completely disregard the allegation. I really do believe that's the way the community would react. And if there were a few opposes based on something that was refuted, I'm also confident that the Bureaucrats know how to discount those opposes. But of course, that's a situation in which everything works as it's supposed to, and I should not have implied that nothing can ever go wrong. And I don't think that your view about Blackstone's is really shared by the community: I think there are quite a few RfA regulars who will oppose on the basis of having once done something wrong offsite if they consider it bad enough that they would not trust the candidate with the tools. I realize that the topic of harassment characteristically brings out strong feelings (a la WTF), but we still need to be thoughtful about these things and not go overboard. What you said about posting a reminder – I don't think you've really listened to what I said. Posting a reminder of existing policy is indeed not creating new policy. Of course! I never said otherwise. But posting a "reminder" that modifies the meaning of existing policy, to mean something that the policy itself does not say, that's what I'm objecting to. --Tryptofish (talk) 19:09, 30 August 2018 (UTC)


 * I've referred to something earlier that perhaps has gotten lost in the wall-of-text, so I'll repeat it in the hope that it will make what I am concerned about here clearer. To begin with: of course no one should link to or describe offsite personal information in an RfA or anywhere else. Of course. But it actually happened that an editor who was angry at me over something once posted at Wikipediocracy to start a process of doxing me. They listed a whole lot of stuff about the pages I edit here and asked for other people to track down who I am. Nothing happened, because nobody there responded. Anyone think that's OK? It happens that the editor who did that had previously been an unsuccessful RfA candidate, so it's theoretically possible that they might run again sometime in the future. If they do, I'm going to oppose.
 * So does anyone think that I should not be able to oppose on that basis, so long as I don't point to any personal information about that other editor? Does anyone think that I should not be able to explain why I oppose? I wouldn't be opposing based on personal information about that other person, but it would be "information". And most importantly, does anyone think that if I did oppose, it would be me who was showing an interest in outing? --Tryptofish (talk) 20:02, 30 August 2018 (UTC)
 * How would you know for sure that the editor on Wikipediocracy is the same person as the editor here? If they have linked to their Wikipediocracy account here on Wikipedia you could mention it, but otherwise it would be WP:OUTING. You could oppose based on your suspicion, but without a confirmed connection it would be just that. Does that make sense? Brad  v  20:08, 30 August 2018 (UTC)
 * In this case, it happens that they have in fact linked here to their account there, so there is no likelihood of a false flag, and no, it would not be outing. In fact, they posted here that they were going to do that there. I realize that there are other hypotheticals where that would not be the case. For the very simple argument that I'm making, that we should change "information" → "personal information", let's recognize that I would technically be referring to offsite "information". I should be able to refer to that "information" in an RfA, and doing so in the way that I describe here would not violate the outing policy, so we should not word the edit notice in a way that would allow someone to weaponize such an oppose. --Tryptofish (talk) 20:40, 30 August 2018 (UTC)
 * In your initial scenario, and assuming no on-wiki connection, you can email that information to ArbCom. It would quite clearly be grounds for a ban. And I think it would be OUTing, because the only exception I see for that type of linking is posting on pages relevant to addressing whatever it is you're referring to (the example being linking offsite to a declared COI and the COI noticeboard), and we don't have such things for harassment. Notably, harassment is prohibited, so ArbCom will either sanction the editor or drop the incorrect complaint; there is no ban on having a COI. —Compassionate727 (T·C) 20:42, 30 August 2018 (UTC)
 * I know. But for the purpose of this discussion here, I want to make it clear that I should be able to oppose in an RfA when there is such an on-wiki connection. And it's not just about me: I'm simply providing an example. I'm concerned about the difference between "information" and "personal information", and how it's just not worth it to fight for "information". --Tryptofish (talk) 20:52, 30 August 2018 (UTC)
 * So then I don't see how the distinction is a problem. As it is, the suggestions note exemptions for material the user in question has linked to: that seems to me to cover the situations you are concerned about. Am I missing something? —Compassionate727 (T·C) 00:58, 31 August 2018 (UTC)
 * With ALT0 and ALT1, we would be saying that it is forbidden to post or present off-wiki "information". In the example that I just gave, if I were to discuss how I oppose a candidate because they attempted to organize an effort to out me, I would be presenting "information" about something that occurred off-wiki. As a result, that person could weaponize the language in the edit notice to, paradoxically, accuse me of outing them, even if I had presented absolutely nothing in the way of "personal information" about them. And nothing in my oppose would actually have violated the outing policy as it now exists. As far as I'm concerned, that would be an outrage. And you better be sure that words matter here, because no matter how much common sense went into the drafting of the language, bad faith editors will seek out ways to game it to their advantage. The outing policy uses the language "personal information". It seems reasonable to me to restate it using the same language as the policy, rather than using different language. In the earlier stages of this discussion, some editors were acting as if it would be of overwhelming importance that we say "information" and not say "personal information" as per ALT2. Based on how the discussion has gotten quieter, I hope that they are now reconsidering that after further reflection. In my opinion, it is unreasonable to insist that we cannot say "personal information". And that's all I've been arguing for. Just go with ALT2, and there is no problem, because we genuinely would be restating the existing policy. --Tryptofish (talk) 15:49, 31 August 2018 (UTC)
 * I disagree (I think fundamentally) with your interpretation of what constitutes "personal information," but you are correct that the policy consistently qualifies information as "personal", which I find sufficiently cogent. (Although when I scrutinize the exemplification of personal information provided by policy, I unfortunately doubt that either of the incidents that caused it actually constitute a violation, so this whole discussion may have accomplished nothing.) But while we're being deliberate about word choice, may I append or link to to your phrase forbidden to post? I think we want to say that linking to where someone can find the information is just as bad as quoting that information on-wiki. I imagine the final product reading something like It is forbidden to post or link to personal information unless the candidate has posted or linked to that information on-site. ("Off-site personal information" is redundant: given the exception provided, applicable information is inherently off-site.) —Compassionate727 (T·C) 16:49, 31 August 2018 (UTC)
 * About how I mentally define "personal information", it is simply what it says on the policy page. And I think the policy page is the official guidance on the matter, so anyone who wants to change what it says there needs to propose it there. About the wording of ALT2, all I did was change the one phrase from ALT1. Frankly, I don't care about the differences, otherwise, between ALT0, ALT1, or your suggestions for ALT2 (although I think a link to something is a posting about it, but whatever) – so long as it says personal information. For all this discussion, that's really all I've been arguing for. --Tryptofish (talk) 20:57, 31 August 2018 (UTC)

Is there a consensus here?
Well, given this just sat abandoned for a week, and not doing anything would a be a tragic waste of a clear consensus to do something, would proposing this require another survey?

I'll argue it's an insubstantial modification of ALT2 (the clear favorite) for precision and tidiness, but someone may disagree. —Compassionate727 (T·C) 19:50, 6 September 2018 (UTC)
 * For me, that sounds just fine. But I'm not seeing the consensus that you see. It kind of looks to me like insufficient consensus to have an edit notice. --Tryptofish (talk) 20:20, 6 September 2018 (UTC)
 * Well, I'm counting 5 supports for every oppose (17 supports to 3 opposes, including nonbolded comments that clearly favor one side or the other), if we consider votes to be relevant. If we're more interested in which arguments are more cogent (as we should be, although I've seen more than a few admins argue that a large majority has as much weight as a policy-based argument), it becomes considerably muddier, because the challenge that this won't accomplish much of anything has not been satisfyingly answered, although you can effectively counter-argue that a single line in an edit notice is hardly what the Avoid instruction creep is writing on. Depends on how you look at it, hence the question. —Compassionate727 (T·C) 21:03, 6 September 2018 (UTC)
 * I think there’s consensus for a line in an edit notice: we aren’t inventing a new policy here. The main opposition was not to create a new rule for RfA, which this wouldn’t do. All it’d do is add a line to the (existing) RfA edit notice about existing policy, which does help when it comes to enforcement (i.e. no one can claim to be surprised if something they post at an RfA is suppressed.) To isaacl’s point below, it might change some behavior, but even if it doesn’t, it would make it easier to respond to when it happens, which is a net positive. TonyBallioni (talk) 14:15, 8 September 2018 (UTC)
 * I think the trickiness in enforcement is due to arguments on what constitutes redactable information, and not on the principle itself on not posting unrevealed personal information. I don't think a revised edit notice will change this. isaacl (talk) 17:00, 8 September 2018 (UTC)
 * So if I went to AN and asked an uninvolved administrator to add the line, I'm not going to be blindsided for a de facto involved closure? —Compassionate727 (T·C) 19:07, 9 September 2018 (UTC)
 * The best approach would be to ask for an uninvolved admin to assess whether there is consensus to add the line, and to add it if there is consensus. --Tryptofish (talk) 19:10, 9 September 2018 (UTC)
 * Consensus seems fairly clear to me, as an involved participant. Not sure why you're disputing the idea of closing with ALT2, Tryptofish, since you said in the discussion that you wouldn't mind supporting it. But anyway, no matter. I've requested a close at Administrators' noticeboard/Requests for closure. Thanks &mdash; Amakuru (talk) 22:09, 9 September 2018 (UTC)
 * What a strange comment to direct at me. I've said that I am OK with ALT2, having in fact proposed it, and all that I said about closing is that the consensus isn't clear to me and there should be an uninvolved closer. --Tryptofish (talk) 22:19, 9 September 2018 (UTC)
 * , per your earlier comment, I'll ping for his thoughts since template requests and closings are kinda his thing. If you have an objection to that, I'm sure he wouldn't mind, but it seems easier than the ANRFC route. TonyBallioni (talk) 16:44, 10 September 2018 (UTC)
 * Thanks, Tony. I think that's a good idea, and I agree that ANRFC is likely to be too slow. --Tryptofish (talk) 20:31, 10 September 2018 (UTC)

I agree with Opabinia regalis that I don't see a particular need to single out this specific page, since personal information not revealed by the editor in question shouldn't be posted on any pages. It probably wouldn't do any harm, either, but does anyone honestly think having an edit notice with this message will actually change behaviour? I think we need to redact this info faster and, as Opabinia regalis alluded to, engrain it into the culture. isaacl (talk) 20:55, 6 September 2018 (UTC)
 * I can't remember when I have last read an edit notice. So the proposed change is probably as harmless as it is useless. —Kusma (t·c) 21:22, 10 September 2018 (UTC)

Regarding no excuse for not following it: the point is for certain basic principles, English Wikipedia culture doesn't consider the absence of an edit notice to be an excuse. Once we start with one basic principle, then an argument can be made to include another, and so forth until the edit notice isn't sufficient focused to be useful. isaacl (talk) 01:34, 17 September 2018 (UTC)

Admin count over time
Is there a page that lists the total number of admins on en.wiki by year, or any other time period? The closest I could find is the graph under the "visualization" section on WP:Desysoppings by month, but that graph doesn't include the actual data, so you'd have to guess it (i.e. the number of admins in a given year) by eyeballing the graph. IntoThinAir (formerly Everymorning) talk  21:38, 5 September 2018 (UTC)
 * You could work it out easily enough by looking at the history of List of administrators, which has the total and active figures updated daily. We used to keep a graph, but it was abandoned circa 2008 as nobody had a practical use for it. &#8209; Iridescent 21:42, 5 September 2018 (UTC)
 * You could work it out easily enough by looking at the history of List of administrators, which has the total and active figures updated daily. We used to keep a graph, but it was abandoned circa 2008 as nobody had a practical use for it. &#8209; Iridescent 21:42, 5 September 2018 (UTC)


 * Ah, I lied, the graph is still being updated. The dip looks alarming at first glance, but it largely reflects the decline in the editor base; as a proportion of the editorship the percentage of admins is fairly constant. &#8209; Iridescent 21:45, 5 September 2018 (UTC)
 * Does it mean that there are approx. 1,200 total admins, of whom a little less than half are active? That sounds high to me. --Tryptofish (talk) 21:51, 5 September 2018 (UTC)
 * Actually according to WP:Admins, the about 1200 number is right (1202 to be exact). IntoThinAir (formerly Everymorning) talk  22:50, 5 September 2018 (UTC)
 * As of last time the bot did its daily update, there were 1202 admins of whom 524 were active. There were also 469 semi-active admins (less than 30 edits in the past month but still edited recently); the semi-active number always rises at this time of year as people go on vacation. &#8209; Iridescent 23:42, 5 September 2018 (UTC)
 * I didnt look at the graph carefully, and couldnt understand from the look i took at it. But generally speaking, an admin is de-sysoped after an year's activity. Loosely speaking, we have at least 2 de-sysops every month. That is, zero admin activity logged in the period of one year. Based on random observations, I'd say around 300 admins are barely active. And around 300 admins accounts are now totally inactive, just waiting for the 12 month period. That leaves us with ~600 accounts that are "technically" active. But what they are actually doing is a different thing altogether. — usernamekiran (talk)  00:15, 6 September 2018 (UTC)
 * I asked this question because I started wondering: if you extrapolate current trends in the number of en.wiki admins, in what year would you expect this number to reach zero? In other words, I was curious as to how fast the number of admins has been declining. Looking at the graph shown above, let's say it's been about 4.5 years since we had 1400 admins in early 2014, and now we have 1200 admins (rounded down of course), that means we are losing an average of 200 admins per 4.5 years, or about 44.4 admins per year, or about 3.7 per month. IntoThinAir (formerly Everymorning) talk  01:05, 6 September 2018 (UTC)
 * Looking at WP:DBM, we've lost on average just over 40 admins per year for the last 7 years. --Hammersoft (talk) 01:40, 6 September 2018 (UTC)
 * It will never reach zero, it should be an exponential decay, not a linear decay.--Ymblanter (talk) 09:14, 6 September 2018 (UTC)
 * Whenever discussing admin activity it's always interesting to look at the breakdown of admin actions. 40% of non-bot logged admin actions in the past month were done by just 7 administrators, a majority by the top 14, and 77% were done by the top 50. Obviously a simple count of logged actions isn't a perfect representation of admin activity, but it should give a sense of scale. Some could take this as an argument that the project could probably get by with a surprisingly small number of admins (and a few bots) if they were all as active as this cohort. Others might say we're already functionally reliant on a fairly small team as-is before any future decline is projected. But in any case, imagining a team of 500-600 active admins (let alone 1200) is probably even less useful than one would think. MarginalCost (talk) 01:23, 6 September 2018 (UTC)

I was going to say something similar to what MarginalCost said. If one takes a look at the archives of this page, a long time ago a similar issue was discussed. In nutshell, in the early days of enwiki, a lot admins were Jimbo's friends, literally "handed over". A lot of others were never active much. My point is, most of the actual admin activity has always been done by a very small number of accounts. But this discussion is not about activity or workload (the previous discussion was about that). The discussion is about after a certain time period, how many admins would be out there? I think, the users who became admins in last two years, wouldnt get inactive in near future. But thats just a personal opinion (not a hunch), based on observation of their activity. So - statistically speaking, the de-sysoping based on inactivity is expected to decrease, say two or three years from now. I dont know how many successful RfAs we would have in future, but lets just hope they are long term players. If the successful RfAs have current rate, and they are long term players, or at least "longer"; then I think we would never go below 300 admins. Thats roughly the number of "active" numbers anyways. (Iridescent says 524 admins were "active", and I agree with him; but thats a "technical" number. Loosely speaking, at a given moment we have ~300 (active) admins who actually do "adminy" stuff.) — usernamekiran (talk)  01:42, 6 September 2018 (UTC)
 * I'd like to second this. Also, the number of 'active' admins has not appreciably dropped in the last five years. I'd like to see someone crunch the numbers and find out how much the group of active admins changes from year to year. Are admins swapping back and forth from the 'active' to 'semi-active' or 'inactive' groups much? If there is a lot of exchange between these groups, than a lower number of admins overall would be expect to eventually impact the 'active' group, but if the groups don't have much exchange of personnel, the better question would be how many we are losing from the cohort that are generally in the 'active' group. I suspect that the numbers that we are losing from this group are probably much closer to the numbers that we are currently getting through RfA. Given that even this group of 'active' admins has a power law distribution in how much they edit, the majority is done by relatively few of them. In terms of logged actions, how many names change amongst the top 300 admins year-on-year? the top 100? the top 50? the top 20? Answers to these questions would likely reveal how many active admins (of different levels of activity) we need to add each year to keep up. I'm going to seriously consider running some queries for the above data and putting some charts together. —  Insertcleverphrasehere (or here)  02:17, 6 September 2018 (UTC)
 * So, taking just a first stab at it, I did a quick check of logged actions in August 2015, 3 years ago. The top 50 admins account for 73% of logged actions then (slightly, but not dramatically, less than the 77% today). Of that group of 50, 39 have any logged action in the past 30 days, 30 are in the top 100, and 22 are still in the top 50 today. Or, you could say that same cohort of 50 admins still accounts for 38% of our logged actions today. So, to wildly generalize to a first approximation, the half-life of the admin corps is about 3 years. We lost something like half of the most active admins in 3 years, and about half their logged actions, and would need a replacement rate of about 1 new top-tier activity admin about every 1.5 months. A top-50 admin is something like 1/5 of the admin corps that logs at least 10 actions per month, so we'd be looking at something like adding 3 admins a month to keep pace. But this is just one quick and superficial comparison. I'd be interested in any conclusions you come to from a more comprehensive study. MarginalCost (talk) 03:07, 6 September 2018 (UTC)


 * Just pointing out that "logged actions" is an entirely useless metric. Interesting, don't get me wrong, but still utterly useless. One admin may work for an hour and log dozens of actions while another may work for eight hours and log only 2 or 3. If one admin spends two hours investigating and sorting out a single AN/I complaint that ends in no logged actions, that does not make them less active than an admin who spent two hours deleting pages and blocking vandals. For the most part, it only requires a basic level of competence and intelligence to work AIV, RfPP, CSD, etc., where you sort of just grind away at the backlog, spamming logged actions the whole time (no disrespect to the admins who do so; I am one of them), whereas other tasks such as investigating reports at ANI, AN3 or AE, handling CCIs or SPIs, answering Adminhelp requests and protected edit requests are equally unending parts of the backlog that eat up hours of Admin time yet don't necessarily correlate with logged actions. We can even treat AfD as an example: one admin may easily spam 50 uncontroversial closes in the time it takes another admin to review and assess a single contentious discussion. Can we really say the admin who closed the 50 uncontroversial ones is more active than the admin who closed the 1? In reality, those latter admins form backbone of the project just as much, if not more, than admins who rack up huge log stats. S warm   ♠  08:52, 6 September 2018 (UTC)
 * And couple that with the fact that the majority of admin actions aren't logged at all, since the commonest admin action—probably by a full order of magnitude—is "closed as no admin action required". &#8209; Iridescent 09:36, 6 September 2018 (UTC)
 * Edit Count is a similarly useless metric, yet we use it to estimate how active people are all the time. I get your point though. If I put together some statistics, I'll also look at edit counts among admins, perhaps restricting to Wikipedia and Wikipedia Talk spaces. I'm a bit busy at the moment, so I don't expect to get around to it any time soon. —  Insertcleverphrasehere (or here)  10:56, 6 September 2018 (UTC)
 * The above points are acknowledged, but I should like to note that because we are looking at admin activity over time, this metric is still useful. Most admins find a particular area or group of areas that they like and settle down, and the number of logged admin actions that are needed on the project has probably remained fairly consistent for the past five or so years (that's probably a function of editor activity and readership, but I don't know where to find data on any of this). Thus, for the purposes of determining who is active then and now, the list is useful. And to say we are interested in who does the tedious, repetitive stuff, how much of it and for how long is not to undervalue the work of those who do other things: both are absolutely vital. The fact that the data isn't a comprehensive understanding of all scenarios does not mean we shouldn't use it for those things about which it is insightful. —Compassionate727 (T·C) 16:42, 7 September 2018 (UTC)


 * Actually, it would be great if we could maintain smth similar to Commons:Commons:List of administrators by recent activity, but this obviously needs a different formatting and a dedicated bot.--Ymblanter (talk) 09:11, 6 September 2018 (UTC)
 * Does this introvert navel gazing have a point? It looks like dubious data for the sake of dubious data. Yet another pointless tangent, IMO. Leaky  Caldron  11:25, 6 September 2018 (UTC)
 * I do not see how admin activity is dubious data, and, if you do not like the metric, you are of course welcome to propose a different one.--Ymblanter (talk) 11:29, 6 September 2018 (UTC)
 * 3/5 of this chart is based on guesswork. There is no robust, accurate way of measuring Admin. outputs / productivity. It is the very definition of dubious. Just like the nonsense spouted every time a few inactive Admins. drop off the list. They have been INACTIVE - we have therefore lost NOTHING when they are finally deactivated. As for the daft idea that I should propose a different one - you have missed my point - it is unnecessary to constantly pontificate about this. Leaky  Caldron  11:49, 6 September 2018 (UTC)
 * I think Leaky does have a point - the number of total admins has dropped off, and there are certain areas where admin activity is quiet, but I think it's generally always been that way, and that I think it would take years and years for the levels of admins to drop off to a critical level. <b style="color:#7F007F">Ritchie333</b> <sup style="color:#7F007F">(talk) <sup style="color:#7F007F">(cont)  14:34, 6 September 2018 (UTC)
 * is right: given the current rate of net admin loss of 44.4 per year (based on the 4.5 year period I mentioned above), it would take 1202/44.4 = about 27 years for en.wiki to totally run out of admins. Assuming the current rate remains totally unchanged, of course. IntoThinAir (formerly Everymorning) talk  16:41, 6 September 2018 (UTC)
 * I wish to note that because the project is trying to preserve the sum of human knowledge for the entire conceivable future, if we accept that we'll run out of admins in 27 years, that would be a monumental problem. Obviously, nobody accepts that, but "saying it will be a while before it's a critical problem" is not really an acceptable response if we can determine now that it will be a problem before our readership dissipates. —Compassionate727 (T·C) 19:44, 6 September 2018 (UTC)
 * Well then, there's an easy solution: make sure that our readership dissipates. { --Tryptofish (talk) 20:22, 6 September 2018 (UTC)


 * Everybody keeps saying that "RfA is harsh process", and/or somwhere along the lines of "it is getting harsher with time". I agree admins should be good, but nonetheless the current situation of RfAs is very weird. The expectations are too high, the scrutinisation too in-depth, and seldom there is drama too. For whatever reasons, the number of RfAs has gone way down. The good candidates arent coming forward. They just keep editing, but dont go for RfA. Not even if they are offered a nomination. According to Requests for adminship by year 594 RfAs in 2008, 355 in 2007, and 231 in 2006. And then there were 36 in 2016, 40 in 2017, and 14 in 2018. With the current rate, I think in 2018 we would have 30 RfAs at most; possibly even 25.The problem is not about "having admins", most - almost all the problems are caused by the harshness of RfA. But this problem cant be solved by changes in policies, or changes in view of individuals; it can be addressed only if the "community as whole" changes its perspective/view. A lot of RfAs were failed based on "what candidate hasn't done" or "isnt willing to do". Instead of that community should concentrate on what the candidate can do with tools, and if there is a chance of misuse/abuse. Even though the process is lengthy/cumbersome/dramatic, there is a process to revoke admin privileges. I meam, I dont mind if pbsouthwood keeps deleting pages in projectspace and then moves on to CSD, and stuff that he his knowledgeable of. Likewise, I dont mind if Bbb23 does nothing but handling SPIs, Anthony doing hist-merges, and so on. I think I babbled a lot lol. My point is: we as community, should change our approach towards RfA. We shouldnt expect the candidate to be an "all-rounder", if they want to work in certain fields; we should give then the tools. If something goes wrong, admonishing, and in further cases revoking the privileges is always an option. This a lot better than the perennial suggestion of splitting the admin tools (eg rollback, pagemover). — <span class="monospaced" style="font-family: monospace, monospace;">usernamekiran (talk)  00:57, 7 September 2018 (UTC)
 * I maintain that the issue of low RfA numbers could be partly solved by looking at the problem less as "good candidates arent coming forward" and more as "good nominators arent looking for candidates". I nominated 7 editors in 3 months and built up a long list of 'in 6-12 months' candidates because I sat down and spent some time looking for editors who would make good candidates, discussing the process and their concerns, and encouraging them to run. Browse AfD, AIV, RFPP, and you'll quickly build up a long list of editors who might make good candidates. Most RfA noms come from the same few people, and I'd encourage everyone else who browses and discusses on this talk page to spend some time looking for candidates. A few of us actively looking for candidates led to 9 new admins in a single month in 2017 - if we kept that up for the rest of the year it would have been quite something. Sam Walton (talk) 09:17, 7 September 2018 (UTC)
 * I can't speak for how stressful RfA is and whether people are scared to do it; I have no personal experience with that and am dispositionally tolerant of stress. I'm also relatively new to RfA as a whole, but I've voted in every in the past several months, and from what I see, people pass whenever they are capable of doing good work and there aren't serious quality or temperament concerns. Yes, there's always a couple of fools who stubbornly oppose because of content creation and "deletionist tendencies." If 1-2% of voters behaving foolishly genuinely stresses candidates, the candidates are the ones with the problems. Maybe it used to be a problem? seemed to receive a peculiarly hard time of it? Is that what RfA once looked like? I have no idea. But since him, three people of high qualifications have passed easily (with many of the flimsy objections raised in Pbsouthwood's RfA never returning), and  will pass easily if he runs again in nine months. We're lamenting that something needs to change in the community, but perhaps it has. —Compassionate727 (T·C) 16:42, 7 September 2018 (UTC)
 * Not so stressful for me. Interesting study in human behaviour, though. I think people who opposed on what were seen as petty or irrelvant points had a far harder time of it. Maybe there is a tendency for the less easily stressed to run for admin.
 * I quite agree with the opinion of that we should not be expecting all admins to be all-rounders. If I find the candidate to be trustworthy, civil, and responsive to reason, I will probably support them. If lacking in one of these, I will not. What they do with the bit is of secondary importance as long as it does no harm. &middot; &middot; &middot; Peter (Southwood) (talk): 18:20, 7 September 2018 (UTC)
 * With the current rate, I think in 2018 we would have 30 RfAs at most; possibly even 25. The end of the year is closer than you think. We have only 3.5 months of 2018 left and we've only had 14 RfAs. I think around 17 or 18 total RfAs is a more realistic estimate, assuming we continue the trend of ~1 per month. Unless we add at least 5 administrators RfAs per month over the next 3 months, we will not reach 30 RfAs. For 25, we'd have to add at least 3 admins RfAs per month. Mz7 (talk) 07:59, 15 September 2018 (UTC)
 * lol, yes you are right. When I did the math, I stupidly thought it was August month going on murmurs in mysterious voice: "dang it, i must get that machine fixed!" — <span class="monospaced" style="font-family: monospace, monospace;">usernamekiran (talk)   18:11, 19 September 2018 (UTC)


 * I endorse what said. RfAs of suitable candidates usually pass (I can only think of one RfA in the last 6 months that failed where I disagreed, and that went to crat chat.) The issue is that we need more current admins reaching out to suitable candidates to run. This is in my opinion the single most important factor in RfA in practice. TonyBallioni (talk) 14:19, 8 September 2018 (UTC)
 * Actually, any experienced user should be encouraged to nominate, not just current admins. Aiken D 15:52, 8 September 2018 (UTC)
 * Has anyone come up with a list of the non-admins with the highest admin score? Might be a good place to start.  Λυδ α  cιτγ  06:59, 11 September 2018 (UTC)
 * I think gets top marks for everything. <b style="color:#7F007F">Ritchie333</b> <sup style="color:#7F007F">(talk)  <sup style="color:#7F007F">(cont)  15:52, 11 September 2018 (UTC)
 * Heh :)  "Observe my works and tremble, ye Wiki", etc. That is really bizarre. To be honest, I imagine that (something like) anyone with over 10K edits and 5 years' tenure probably scores the same, considering everything is capped at 100...also, it doesn't seem to include block logs? Which I find particularly insulting! :p  —SerialNumber54129  paranoia / cheap sh*t room 16:09, 11 September 2018 (UTC)
 * No, I checked myself, and I do not score all top grades--Ymblanter (talk) 16:24, 11 September 2018 (UTC)
 * Same, although I'm close (though not as close as ). Not that the metrics are useful (size of userpage, anyone?). We've had this conversation before, although I admit that as a panning tool it might have a little bit of utility. Unless, of course, nobody's willing to make the nominations, which is probably the real problem. —Compassionate727 (T·C) 16:30, 11 September 2018 (UTC)
 * very surprisingly I scored higher than SerialNumber54129; but just because of the stupid userpage. Also, it didnt consider my recent-ish block. But rather than these stats; i think understanding of policies, and behaviour are the qualities that we should look for. — <span class="monospaced" style="font-family: monospace, monospace;">usernamekiran (talk)  17:44, 11 September 2018 (UTC)
 * I am too afraid that we might lose a valuable contributor to a stupid RfA. Probably that fear is misplaced (the only RfA nominee of mine that failed successfully self-nominated a while later) but I have become overly cautious. Am I the only one? —Kusma (t·c) 08:05, 11 September 2018 (UTC)
 * Probably? I doubt anyone who would be so traumatized by an abortive RfA would run in the first place. —Compassionate727 (T·C) 16:25, 11 September 2018 (UTC)
 * And this is why so few people run in the first place, and why we are excluding all but the thickest-skinned editors from crewing on the the good ship Admin; by the nature of RfA toxicity. The few admins I have approached for nomination are so afraid that I'll run away after a potential failure that they aren't keen to accept the smallest bit of risk. I've got thick skin though, I'll probably self-nom sometime soon. —  Insertcleverphrasehere (or here)  21:32, 11 September 2018 (UTC)
 * Funnily, I also score higher than Serial Number 54129, though only by 33 points. Apparently this is because of my 29 edits to AIV-related pages. IntoThinAir (formerly Everymorning)  talk  18:56, 11 September 2018 (UTC)
 * [ec] I have very little confidence in the scores generated by that tool. It gives no indication whattsoever of whether the person is able to deal with conflict, remain civil, neutral, polite and rational in discussions, express themself clearly, accept consensus they don't agree with, work out a compromise that both sides can accept, gain and retain the trust of fellow wikipedians, have a need or desire for any of the tools, and work collaboratively. These are traits that I would hope to see in any admin. The specialist skills come on top of that. Others may see it differently,and it would be interesting to see how different they might be. I would not be surprised if the focus varies according to what work one does the most. &middot; &middot; &middot; Peter (Southwood) (talk): 19:38, 11 September 2018 (UTC)


 * Can I just point out to the room (or anyone else that considers it remotely interesting) that was clearly ribbing me about the so-called highest-score, and no-one should take it seriously whatsoever. As I pointed out, the algorithm used makes it virtually useless as a comparator. So please...after all, biggest isn't always best  :p   —SerialNumber54129  paranoia / cheap sh*t room 19:15, 11 September 2018 (UTC)
 * Oh? Is this a number comparison?! Why didn't anyone tell me? Not cool guys! Let me jump in! *Looks at numbers* "two plus........two......is........five...hundred million! There! I win! "
 * To add to the "ribbing", if I managed to be an admin and get a higher score with less edits, so can you! -- The SandDoctor  Talk 20:16, 11 September 2018 (UTC)

This isn't about RfA, but...
...this is a page that is widely viewed by administrators so I thought I would post it here. I just happened to stumble across an obscure page, Community health initiative on English Wikipedia/Per-user page, namespace, and upload blocking, where I learned that the WMF is planning a major change in blocking. They are planning to introduce something they call "granular blocking" or "partial blocking," whereby we can choose to block someone from a specific page, group of pages, or category of pages, rather than from Wikipedia as a whole. (Site-wide blocking will remain as the default - I think.) Am I the only one who didn't know this was going on? They've been discussing it at the talk page since May, and are now into the phase of designing what the new system will look like. It sounds enormously complicated. I do hope that at least they are planning to provide some training to admins before they launch it. Maybe even some kind of trial or beta test? I know that isn't WMF's style. Comments? --MelanieN alt (talk) 16:33, 19 September 2018 (UTC) (yes, it's really MelanieN, just away from my regular computer)
 * Interesting information. Though, I think perhaps WP:AN would be a better location for this? That page has more watchers (4525) than this page (3390) and I would think is watched by more administrators. This was mentioned in the June Administrator's Newsletter. --Hammersoft (talk) 16:42, 19 September 2018 (UTC)
 * This has been posted lots of pages. One of the noticeboards asked for input on what it would look like a few days ago. Don't remember which one. Natureium (talk) 16:43, 19 September 2018 (UTC)
 * Found it. Really not the most widely viewed place. Village_pump_(miscellaneous). Natureium (talk) 16:48, 19 September 2018 (UTC)
 * Obscure indeed - the actual page has had 113 lifetime views, though the talk page has had some 300. Johnbod (talk) 16:44, 19 September 2018 (UTC)
 * I knew this was happening and I've campaigned for this for ages. It means that if two users are edit-warring on an article where other editors may want to look at it, we can just block them from that article for a short time, and they are free to edit away from that, instead of getting a standard edit-warring block. In turn, it means a site-wide block would no longer be appropriate for those sorts of editors, and many reports at WP:AN3 would only warrant partial blocks. Also, the site-wide block is a much more major admin action, and consequently ought to be easily overturned to a partial block instead. Hopefully, you'll never again get massive ANI threads complaining about a block, because the action that probably led to the block would be resolved via a partial block instead. I'm not really bothered how it's implemented, I can just learn stuff by rote like the other admin tools. <b style="color:#7F007F">Ritchie333</b> <sup style="color:#7F007F">(talk) <sup style="color:#7F007F">(cont)  16:45, 19 September 2018 (UTC)


 * I saw the notice, and I thought it was oddly-placed as well. I'm kind of with Ritchie, with tools the WMF has deployed since I've had the bit I've taken a "see what they come up with and use it if it's useful" approach. But I guess with changes to blocks there should probably be a companion on-wiki discussion about actually using them. Remember PC2? Ivanvector (Talk/Edits) 16:51, 19 September 2018 (UTC)
 * We definitely need to think about policy changes at the Village pump before this thing goes live. I've just effectively described my view of best practice at WP:AN3 - does everyone agree? (Probably not). <b style="color:#7F007F">Ritchie333</b> <sup style="color:#7F007F">(talk) <sup style="color:#7F007F">(cont)  16:53, 19 September 2018 (UTC)
 * I don't see why people that are being blatantly disruptive shouldn't be blocked outright. Natureium (talk) 16:56, 19 September 2018 (UTC)
 * My understanding is that blocking outright will remain the dafault. But what do I know? I only found out about this half an hour ago. --MelanieN alt (talk) 17:00, 19 September 2018 (UTC)
 * What I mean is, if we now can block a user just from a particular page, is there community consensus for that action rather than a standard block? And at the other end, are we going to be expected to use these limited blocks instead of the site blocks we're all used to? Is Twinkle going to be recoded to support partial blocks? There should really be more on-wiki discussion about this. Ivanvector (Talk/Edits) 16:54, 19 September 2018 (UTC)
 * (ec)Good point about AN. Although AN doesn't seem like a place for sober discussion about meta stuff; it has become more of a place to deal with user issues. Actually AN is where I saw a notice about it - which didn't really make clear what was happening, just asking for input on something, I think Special:Block log. And the notice got no replies and apparently didn't entice anyone but me to wonder what it was about. I do think there needs to be wider awareness. I'm not saying it's a bad idea - it may be a big improvement - just that we need some time to digest it and learn how to use it before it suddenly gets introduced. A lot of the reason for some of the bad blood between WMF and users is that sometimes WMF makes a change that users didn't know was coming. --MelanieN alt (talk) 16:59, 19 September 2018 (UTC)
 * Posted at AN, where I got my head handed to me in a bag for not knowing about it already. Well, at least maybe a few more people know about it now. --MelanieN alt (talk) 18:13, 19 September 2018 (UTC)
 * For what it's worth, this is my first time hearing about it, so I thank you for posting. Useight (talk) 15:08, 20 September 2018 (UTC)
 * No good deed goes unpunished, right? :) --Hammersoft (talk) 22:24, 19 September 2018 (UTC)
 * The Twinkle thing is a good point (who even uses the normal block interface anymore), but that's more of an issue for the developers of that tool than for the community at large. ansh 666 03:47, 20 September 2018 (UTC)
 * I use the normal block interface.~  ONUnicorn (Talk&#124;Contribs) problem solving 15:48, 20 September 2018 (UTC)

RfA page structure
Currently, the structure is in following order:
 * ===candidate===
 * ====nomination====
 * ====questions====
 * ====discussion====
 * <here rests the user links>
 * =====support=====
 * =====oppose=====
 * =====neutral=====
 * =====general comments=====

I have a few doubts/questions about that structure. Even though RfA is a discussion venue, a new editor (and a lot of not-new editors) might think that the "discussion" section for general comments, as in gossip about the candidate. In the old days, the section "general comments" used to between "questions", and "discussion"; and they had ==== level header.
 * 1) Why does it start with a subsection (===) instead of a primary section (==)?
 * 2) Why does title links back to the same page?  Answered: transclusion. That was a stupid question anyway. — <span class="monospaced" style="font-family: monospace, monospace;">usernamekiran (talk)   03:22, 27 September 2018 (UTC)
 * 3) Why the userlinks provided here are the most inconsistent with links used elsewhere? Is there a reason that rfplinks or usercheck full is not used?
 * Here are some changes that can be done:

— <span class="monospaced" style="font-family: monospace, monospace;">usernamekiran (talk)  02:51, 27 September 2018 (UTC)
 * 1) Reduce the header level by one level. (I dont know why it starts with === though.)
 * 2) Rename "discussion" section to a synonym, or maybe to "survey".
 * 3) Change the current userlinks to rfplinks.


 * The reason for the link to the RfA is likely so that it is easily accessible when looking at the transclusion on the Requests for adminship. We probably could fix this by putting the brackets within tags, though.-- SkyGazer 512 <span style="background: linear-gradient(aqua, #d580ff);">Oh no, what did I do this time? 02:56, 27 September 2018 (UTC)
 * yes. Idiot me. Individual RfAs are transcluded here. — <span class="monospaced" style="font-family: monospace, monospace;">usernamekiran (talk)  03:22, 27 September 2018 (UTC)

The only thing remaining now is "discussion" vs "survey". — <span class="monospaced" style="font-family: monospace, monospace;">usernamekiran (talk)  19:19, 29 September 2018 (UTC)
 * I think === headers are due to transclusion as well. — <span class="monospaced" style="font-family: monospace, monospace;">usernamekiran (talk)  03:28, 27 September 2018 (UTC)
 * Yeah, you're probably right; I didn't even think about that.-- SkyGazer 512 <span style="background: linear-gradient(aqua, #d580ff);">Oh no, what did I do this time? 03:46, 27 September 2018 (UTC)
 * Nay, they're because Template:RfA which is substituted to start every RfA nomination page uses level three headers.. Galobtter (pingó mió) 06:30, 27 September 2018 (UTC)
 * The intent being normal display on WP:RFA in the Current nominations for adminship. ~ Amory <small style="color:#555"> (u • t • c) 10:39, 27 September 2018 (UTC)
 * And which links are you taking issue with? usercheck-short or RfA toolbox ~  Amory <small style="color:#555"> (u • t • c) 10:47, 27 September 2018 (UTC)
 * I did a huge mistake. I never even hovered my pointer on "count" in usercheck-short. Apparently, it links to edit counter of xtools. Till now, in all the RfA's that I participated (except my first one [golden ring], where i casted a pile-on vote), I used to edit a random page to see preview of usercheck-full with candidate's name. The RfA toolbox isnt visible in RfAs in mobile view though. But still, that "'links' point" is now moot too.

Introducing the s-index


Are any of you familiar with the h-index? It's a method for measuring contributions in academia, where h is the number of papers someone has published that each have at least h citations. It has its flaws, but I've had it in the back of my mind for a while as something that might be interesting if applied here. I've finally gotten around to writing up what I'm calling the s-index, measuring sysop activity. In this case, s is the number of sysops with at least s logged actions in a given period. I've embedded one image here that calculates it on a monthly basis, and includes the total number of actions, both with and without bots; there are a few other graphs at User:Amorymeltzer/s-index. I'm not sure that it tells us a whole lot new or interesting, or even much beyond the total actions, but perhaps it is useful regardless.

All data comes from XTools' AdminStats, so I am hugely grateful to User:MusikAnimal and the rest of the XTools team. There's a fair bit more explanation and prose at User:Amorymeltzer/s-index, and I'd welcome everyone's input and criticism of it. I have some additional features or options I'd like to add to it, but for now your feedback on usefulness, errors/bugs, or any suggestions (expanding takeaways and insights, design changes, etc.) would be most welcome. ~ Amory <small style="color:#555"> (u • t • c) 16:35, 1 October 2018 (UTC)
 * Because h-indexes aren't controversial enough. Kidding. This is interesting. Natureium (talk) 16:59, 1 October 2018 (UTC)
 * This is interesting. But it looks like the s-index, unlike the h-index, isn't a measure of individual-level productivity, but of the productivity of all admins collectively. In this case does "logged actions" only include things that only admins can do (block, delete, etc.) or does it include things that non-admins can also do (e.g. moving pages)? IntoThinAir (formerly Everymorning) talk  17:23, 1 October 2018 (UTC)
 * I think it'll be the types of actions listed here, which are admin-only actions. This doesn't capture the full scope of admin activity as some types of admin actions aren't logged (e.g. editing protected pages, closing XfDs as anything other than Delete, applying discretionary sanctions, etc) but it's a close enough proxy to be useful.  Hut 8.5  17:55, 1 October 2018 (UTC)
 * Basically. I originally planned to use JamesR's data, but it was more of a pain to parse/process, was only reliable back to 2008, and had a major change in 2012 that overly complicated things.  I touch on this there, so if I may be allowed to quote myself: AdminStats includes (re)(un)blocks, (un)deletions (including revdels), (re)(un)protections, user rights changes, and imports. Anything else (edit filter changes, renames, merges, massmessages) isn't counted as a sysop action... other typical sysop activities, such as closing deletion discussions, are not counted. This is a common issue with sysop-related metrics. ~  Amory <small style="color:#555"> (u • t • c) 18:34, 1 October 2018 (UTC)
 * [T]he s-index treats enWiki as the scholar, with each sysop as a paper and logged actions as each sysop's "citation" count. Sysops can do things like delete pages, but we can't do things that have their own measurements, so we're one level less deep than a researcher. ~  Amory <small style="color:#555"> (u • t • c) 18:35, 1 October 2018 (UTC)
 * Thanks for this. It would also be interesting to create an "e-index" where e is the number of editors with at least e edits in a given period (both with and without bots) and see how it compares. Gizza  <sup style="color: teal;">(t)(c) 01:02, 2 October 2018 (UTC)

Is RFA becoming stricter?
It seems obvious that RFAs are now a lot harder to pass than they were in the very early days of Wikipedia--as put it if you "weren't a bozo", you'd basically have a 100% chance of passing your RFA. But what if you want to know long-term trends in the % of RFAs that are successful? I have compiled data on this topic from several WP project pages and analyzed it to see what the success rate for RFAs was in each year from 2004 to the present. It is clear that RFAs were much more likely to succeed in 2004 than they are now: last year only 53% of RFAs were successful compared to 79% in 2004. Nevertheless, last year's success rate was actually the highest since 2005's (64%). 2017's success rate is also much higher than the average rate (45%) and especially higher than the lowest year, 2012 (only 29%!). IntoThinAir (formerly Everymorning) talk  00:50, 17 September 2018 (UTC)
 * There is a huge complication here, not all unsuccessful RFA candidates are equal. For the first decade or so of RFA there was a steady trickle of complete newbies filing RFAs and failing. But now you have to have extended confirmed status before you can edit the page and start an RFA. That pretty much rules out the candidates with fewer than 500 edits, and the people who have stuck around till 500 edits usually have the nouse to not run until they have a pretty good chance. To truly compare candidates today with candidates of the past is difficult, and I wouldn't start with the success rate. There are a number of different criteria that matter to groups of RFA participants, the popularity of those criteria and the thresholds have varied over time. I'm one of those who thinks that tagging accuracy for deletions, judgment about blocking and an ability to add inline citations to reliable sources are three important criteria. Those three have been well established for well over a decade, It has been a while since we have seen a candidate whose editing shows they haven't yet fallen in with consensus as to what is vandalism and when we should use the block button, candidates who fail because multiple recent deletion tag mistakes indicate they aren't ready for the deletion button are more common, until Rollback was unbundled in early 2008 it was possible to pass RFA as "a good vandalfighter" since then the de facto criteria have included making contributions to the pedia (despite multiple attempts to raise that bar from adding reliably sourced content to has an FA or GA, the de facto test is more at the DYK level). Others, for reasons I don't always understand or respect focus on length of tenure and editcount, both of those criteria were subject to standards inflation through the first few years of RFA, and now if you have a good candidate with either 3,000 edits or only a years tenure they can expect some opposes from those with outlier views a candidate at the minima for both edits and tenure might well fail even if otherwise well qualified. With the greying of the pedia we now rarely see obvious adolescents or teenagers at RFA, but here the threshold has clearly increased in the last fifteen years, and I would be surprised if someone could now pass while they were known to be under 18.  Ϣere Spiel  Chequers  09:43, 17 September 2018 (UTC)
 * The extended confirmed requirement was only added in Nov 2017. It was been semi-protected since May 2008 and for short periods before that (log). you may be interested in RfA inflation. Adrian J. Hunter(talk•contribs) 10:36, 17 September 2018 (UTC)
 * This point is a valid one that I was actually already aware of. But I was trying to compare consistent statistics  across  time,  so  I just  divided  the number of successful  RFAs by  (successful+unsuccessful) for a given year. But the fact that it became impossible  for many users  to even edit  WP:RFA  last November is definitely  an important factor  in  the  success rate  of  RFAs  since then (many new  users  whose RFAs would have been doomed  were certainly prevented from editing the main page by that protection). So I've been  trying  to  also keep track of deleted RFAs but  I  don't have data  for all years yet. I discussed the protection more here: RFA study IntoThinAir (formerly Everymorning)  talk  12:13, 17 September 2018 (UTC)
 * There's also a major factor – how many folks are scared away from the current culture and process and don't even put their name up for consideration? This chart started by speaks volumes about the pool of folks who are even putting their hat in the ring – RFA_by_month. So we have important variation on the pool of candidates over time, but also the (de)motivation of active users who even want to be run through the gauntlet. -- Fuzheado &#124; Talk 16:58, 19 September 2018 (UTC)
 * I actually thought about this earlier, and you're probably exactly right: the number of potential admins who are willing to even volunteer for RFA, as well as the % who do and actually succeed, have both probably declined in recent years. It seems that the increased stringency of RFA standards and perhaps a more hostile environment at RFA has deterred a sizable number of would-be candidates from running. So this can also affect the success rate by reducing the denominator (total RFAs). Consider that of the >4,000 total RFAs from 2004 to the present, 62% of them occurred before 2008. So there is clearly a change in the # of applicants and in their chance of succeeding. IntoThinAir (formerly Everymorning) talk  18:33, 19 September 2018 (UTC)


 * I've made this comment before, but I find the wailing and gnashing of teeth about how RfA is broken to be misplaced: most qualified candidates pass RfA with over 90% these days. Of the four serious failed RfAs this year, the reasons for failing were roughly: 1) not enough community engagement/too little experience, 2) concerns with views on how we deal with Nazis and temperament issues as seen at ANI, 3) temperament concerns based on a fight with an arbitrator combined with mess from Wikipediocracy 4) copyright concerns and concerns over understanding of sourcing. You'll note that virtually none of these were because of some numeric standard.Of those four, the only one where I think the process was less than entirely fair was Jbh's, and while most of them likely would have passed in 2003, and I was in support for 2/4 of them, I don't think its a broken process.Compare this to the successful RfAs this year, which are 1) more numerous than the serious failed candidacies and 2) with the exception of one were highly successful and had greater than 90% support.The community wants more admins and is generally willing to promote competent people. That doesn't mean that the community is willing to promote anyone, and when valid concerns are brought up, RfAs can fail. I will also maintain my point of view that the reason we don't have that many RfA candidates is because we aren't asking enough people and that if we suggested that more people run, we would see a lot more successful RfAs. TonyBallioni (talk) 19:23, 26 September 2018 (UTC)
 * The main problem isn't the success rate of the trickle of candidates that we get, though we have had a couple of good candidates fail recently. The problem is that very few of the many qualified candidates are actually running and therefore we aren't maintaining our numbers of admins or bringing newer editors into adminship. Ignoring bots and an admin who has changed an account, we have just 12 admins who created their account since the start of 2012.  Ϣere Spiel  Chequers  20:35, 1 October 2018 (UTC)
 * I only disagree with the outcome of one of the RfAs in the last 6 months, and that went to crat chat, but to your larger point, I agree. We need more people to run, and I wish that they did because qualified candidates tend to pass. On the creation point, I do want to push back a little. Technically, I’m a 2007 account, but I only really became active in 2016. I basically became an admin after 14 months. Same with There’sNoTime. A lot of editors of my real life generation created accounts in secondary school, messed around for 3 months, and then came back after graduating university in part because entry-level jobs are boring and have a fair amount of down time or some similar reason. I wouldn’t classify myself as a “10 year editor” by any means, even though I do appreciate my brief stints on Wikipedia a decade ago. I’m basically a 2016 editor who passed RfA in 2017, and I think there are a few other people who have similar stories. TonyBallioni (talk) 04:06, 3 October 2018 (UTC)
 * Thanks, my unsubtle method of looking by creation date does indeed miss editors such as yourself. I remember an ancient discussion either here or at the London meetup where we concluded that such long dormant accounts were very very rare at RFA, but even one that had been completely dormant from inception till activation met the real reason for the 12 month tenure test; That certain long term problematic people were capable of behaving themselves for many months before adminship, but they weren't capable of planning enough years ahead to create accounts that looked like yours. Of course one key metric that has changed since that discussion is that as the project ages from 11 years to nearly 18 years so the window of opportunity for people to have created an account and left it for about a decade skyrockets. Sorry about the job BTW, there is a very reasonable argument that my generation has somewhat mucked up the world for yours.  Ϣere Spiel  Chequers  06:01, 3 October 2018 (UTC)
 * Heh, just means we eventually end up going back to school to get a better job TonyBallioni (talk) 01:00, 4 October 2018 (UTC)

Where's the second Rfa?
My Watchlist is currently telling me there are two RFAs open for discussion, but I can only see Sir Sputnik's (and the last reported closure is for October 1st, so it seemingly isn't referring to some recent closure). So where's the 2nd Rfa? Tlhslobus (talk) 03:29, 18 October 2018 (UTC)
 * See User talk:L293D Abelmoschus Esculentus (talk to me • my contributions) 03:31, 18 October 2018 (UTC)
 * Thanks. That user (L293D) seemingly no longer exists (if they ever did). At any rate my Watchlist has now gone back to just one Rfa.Tlhslobus (talk) 03:37, 18 October 2018 (UTC)
 * (Hallelujah, it's a miracle - L293D has seemingly been resurrected ).Tlhslobus (talk) 03:42, 18 October 2018 (UTC)
 * Too confusing :) Abelmoschus Esculentus (talk to me • my contributions) 08:59, 18 October 2018 (UTC)

Where's the RfA?
My watchlist is indicating there's an ongoing RfA, but I can't see any indication of it Nosebagbear (talk) 19:53, 2 November 2018 (UTC)
 * It's at Requests for adminship/Flooded with them hundreds. I don't think the RfA table has updated yet, but it's transcluded on the main WP:RFA page. Mz7 (talk) 19:55, 2 November 2018 (UTC)


 * For future reference, you can easily check this for yourself by viewing the page history at WP:RFA. Swarm  talk  21:59, 2 November 2018 (UTC)