→This issue is bigger than expected: don't agree |
|||
Line 131: | Line 131: | ||
::Though I've just noticed that tool isn't correct anymore. It's tracking the 'redirect over move = G6 deletion' thing as an admin action and inflating the number of users listed there... [[User:Samwalton9|'''S'''am '''W'''alton]] ([[User talk:Samwalton9|talk]]) 01:14, 2 January 2017 (UTC) |
::Though I've just noticed that tool isn't correct anymore. It's tracking the 'redirect over move = G6 deletion' thing as an admin action and inflating the number of users listed there... [[User:Samwalton9|'''S'''am '''W'''alton]] ([[User talk:Samwalton9|talk]]) 01:14, 2 January 2017 (UTC) |
||
:::...Sort of ([[Phab:T154408|T154408]]). [[User:Samwalton9|'''S'''am '''W'''alton]] ([[User talk:Samwalton9|talk]]) 01:20, 2 January 2017 (UTC) |
:::...Sort of ([[Phab:T154408|T154408]]). [[User:Samwalton9|'''S'''am '''W'''alton]] ([[User talk:Samwalton9|talk]]) 01:20, 2 January 2017 (UTC) |
||
:Beeblebrox' thoughtful and well-written comment makes me, for what little it's worth, wade in to the fray with a contradictory opinion. I think there's an implicit vision that to "Be an admin" means you're going to spend significant time running around the 'pedia doing adminny type stuff, and that if you're not, you're not really an admin, and it would be better/safer/more accurate/cleaner to take the admin bit away. That model (wielding the mop regularly and with dedication) is one version of what adminship means. But it's not the only one. An admin is someone who has the trust of the community to not screw up with extra tools and to even do some good. If they've done that x times, it's always a shame they haven't done x+1 times, or 10x times. But even if x=1, it's a net positive. And while we might debate whether it was worth the community's time to review someone at RFA if once promoted x is too small, it's even more guaranteed to be a waste of time if we cut them off. I'm happy if we have an admin who once in a blue moon deletes some obvious spam or blocks a disruptive vandal that needs it - that's one less action someone else needs to do. |
|||
:I'm sensitive to security issues and so administrative, easily-reversible bit removal after inactivity is a sensible thing. However, for the above reasons, I'd advocate the community (and therefore bureaucrats) continue to be very lenient in returning the bit for admins/users "in good standing" (however defined) even after lengthy inactivity. However, since the admin bit is an expression of trust, I would be less lenient about returning the bit in the case of "under a cloud", where plausibly that trust has actively been lost. [Noting I have read the Cyp example.] [[User:Martinp|Martinp]] ([[User talk:Martinp|talk]]) 18:38, 3 January 2017 (UTC) |
|||
===Publicising this now the holiday period is over=== |
===Publicising this now the holiday period is over=== |
||
I think more voices would be useful. Any appetite for posting this discussion at BN, maybe CENT... anywhere else? --[[User:Dweller|Dweller]] ([[User talk:Dweller|talk]]) <small>Become [[User:Dweller/Old Fashioned Wikipedian Values|old fashioned!]]</small> 10:53, 3 January 2017 (UTC) |
I think more voices would be useful. Any appetite for posting this discussion at BN, maybe CENT... anywhere else? --[[User:Dweller|Dweller]] ([[User talk:Dweller|talk]]) <small>Become [[User:Dweller/Old Fashioned Wikipedian Values|old fashioned!]]</small> 10:53, 3 January 2017 (UTC) |
Revision as of 18:39, 3 January 2017
Wikipedia Help NA‑class Top‑importance | ||||||||||
|
|
This page has archives. Sections older than 30 days may be automatically archived by Lowercase sigmabot III. |
If an administrator is WP:INVOLVED with an editor in content dispute, then is it okay for them to request another administrator on their talk page to topic ban the editor with whom they have content dispute, instead of visiting WP:ANI and WP:AE? This is like gaming WP:INVOLVED, if they ask their administrator wiki-friend (to topic ban the editor) instead of reporting the user at WP:ANI and WP:AE. --Marvellous Spider-Man 16:25, 23 December 2016 (UTC)
- Generally, no administrator can unilaterally topic-ban an editor. Is this a hypothetical, or is it real?--Bbb23 (talk) 16:40, 23 December 2016 (UTC)
- It's real. An administrator "A" topic-banned editor "B" for three months, after they received a request from administrator "C" on user talk page. Administrator "C" was involved in content dispute with editor "B". --Marvellous Spider-Man 16:51, 23 December 2016 (UTC)
- So who are the administrators?--Bbb23 (talk) 17:56, 23 December 2016 (UTC)
- (MS-M e-mailed me) The topic-banned editor hadn't edited for 3 days before the ban was imposed and hasn't edited since. It's their responsibility to appeal the ban if they wish, not yours. Just so you know, administrators often ask other administrators to look at a situation in which they are involved. They expect the other administrator to exercise their independent judgment, i.e., not just act as a proxy for the first administrator. In this particular instance, the administrator who imposed the ban is absolutely independent in their decisions.--Bbb23 (talk) 14:23, 24 December 2016 (UTC)
- So who are the administrators?--Bbb23 (talk) 17:56, 23 December 2016 (UTC)
- It's real. An administrator "A" topic-banned editor "B" for three months, after they received a request from administrator "C" on user talk page. Administrator "C" was involved in content dispute with editor "B". --Marvellous Spider-Man 16:51, 23 December 2016 (UTC)
Workshopping an RfC on the inactivity policy
The community appears to be unsatisfied with this request for resysop, and it seems like it may be time to strengthen our inactivity policy for administrators. I propose the following changes. I welcome counter-proposals, but let's try not to make things so complicated that we get no consensus for anything. Note that this is a discussion, not a vote, so let's discuss before throwing down votes. Consider this a work-in-progress, and this is not an RfC at the moment.
- In WP:INACTIVITY, change "no edits or administrative actions" to "no administrative actions". This has the effect of considering an administrator inactive even if they edit unless they actually use the tools.
- In WP:INACTIVITY, add a line stating that an administrator will be considered active for an additional year if they respond to the requisite user talk messages with a note that the admin wishes to continue holding the tools. This is intended to benefit administrators who may use their administrator status to close difficult discussions, participate at WP:AE, or other administrator activities that don't show up in the logs as an administrative action.
- In Wikipedia:Administrators#Lengthy_inactivity, remove all references to needing continued inactivity in order to require a new RfA. Essentially, if a former administrator voluntarily gave up the tools, they would need a new RfA if they don't reclaim them in three years. If a former administrator was desysopped for inactivity, they would need a new RfA if they don't reclaim the mop within an additional two years.
Thoughts? ~ Rob13Talk 02:33, 27 December 2016 (UTC)
- Not a terrible idea, but I think we would need to have a clearer idea of what would constitute "administrator actions". Bu Rob13 has mentioned a few. As an oversighter, I could do a lot of actions that don't result in any logged activities, such as turning down requests, or participating in discussions about a specific oversight request which is then acted upon by another oversighter. Reviewing unblock requests that don't result in unblocks (which is the majority of reviewed requests) doesn't require the use of tools, but does require an administrator bit. Reviewing deleted content requires the bit but most times does not lead to undeletion (e.g., when responding to a request to email the content of a deleted article to the originating editor). Closing complicated or controversial XfD's is in the scope of administrators but unless the page is deleted, there is no logged action.
I do, however, share BU Rob13's concern that people who appear to maintain just a little more than the bare minimum activity in order to keep the bit are not what we're really looking for in an administrator. I'd support a minimum 150 edits per year or a minimum 150 combined admin actions/edits per year in order to retain tools, with one opportunity to regain tools after inactivity, following the completion of at least 100 edits prior to the request for resysop. Risker (talk) 04:25, 27 December 2016 (UTC)
- @Risker: I agree with a higher threshold of activity requirements. Not as high as you say, but maybe something like 10 admin actions per six months. That's far more controversial, though, so I think we should be trying for something of an improvement over a massive improvement right now. Administrator actions are defined as showing up on AdminStats on Commons, and I suggest borrowing that definition. Note that my second bullet point above covers those who are making non-logged actions. It's as simple as saying "Yup, still using those" when you get an email about inactivity. That's a lot more simple than requiring that a bureaucrat dig through contributions to determine the last edit made in an administrative role. ~ Rob13Talk 04:37, 27 December 2016 (UTC)
- The AdminStats tool on Commons, last I checked, ignores both edits to protected pages and "no-action" decisions such as closing an AfD as keep. It is utterly unsuitable as a gauge for admin activity. The last RfC we had on the subject failed in part because of this problem. Also, didn't know I was that high on the activity list... Jo-Jo Eumerus (talk, contributions) 09:55, 27 December 2016 (UTC)
- Which is why we have the elegant solution that an administrator can just say "Yo, still using the tools" to retain them, as noted above. It's impractical to ask a bureaucrat to review all edits by an inactive administrator to ensure they're not editing in an administrative area, so this is probably the best we can do. Besides, if an admin is closing as "keep", they are likely also closing as "delete". This is an extreme edge case. ~ Rob13Talk 10:11, 27 December 2016 (UTC)
- The AdminStats tool on Commons, last I checked, ignores both edits to protected pages and "no-action" decisions such as closing an AfD as keep. It is utterly unsuitable as a gauge for admin activity. The last RfC we had on the subject failed in part because of this problem. Also, didn't know I was that high on the activity list... Jo-Jo Eumerus (talk, contributions) 09:55, 27 December 2016 (UTC)
- @Risker: I agree with a higher threshold of activity requirements. Not as high as you say, but maybe something like 10 admin actions per six months. That's far more controversial, though, so I think we should be trying for something of an improvement over a massive improvement right now. Administrator actions are defined as showing up on AdminStats on Commons, and I suggest borrowing that definition. Note that my second bullet point above covers those who are making non-logged actions. It's as simple as saying "Yup, still using those" when you get an email about inactivity. That's a lot more simple than requiring that a bureaucrat dig through contributions to determine the last edit made in an administrative role. ~ Rob13Talk 04:37, 27 December 2016 (UTC)
- I would be open to writing a bot to generate reports tracking administrator edits to fully protected pages. If there's interest, let me know and I'll get started. -FASTILY 00:03, 28 December 2016 (UTC)
- To be honest, I do not think this is a real problem. Indeed, there are admin actions which are not logged such as editing protected pages. However I can hardly imagine an admin who would perform these actions but no logged actions during a year. Moreover, if such an admin exists and would be about to be desysopped, we can add a clausure that they may present their non-logged actions to be counted against desysop. Usually we deal with people who do not log in the project for years, not with someone who is active on a daily basis and just for whatever reason does not perform logged admin actions.--Ymblanter (talk) 22:55, 27 December 2016 (UTC)
- This was the most recent RFC on the matter - Dec. 2012 - Feb. 2013 RFC on Resysopping practices. Leaky Caldron 10:48, 27 December 2016 (UTC)
- There were two more recently than that. See here and here. Any new proposal will need to address those concerns. Sam Walton (talk) 11:40, 27 December 2016 (UTC)
- I don't think we should get too hung up on the fact that certain types of admin actions generate fewer logged actions than others. Admin actions are like edits: while number of edits is not an accurate indication of level of contribution to Wikipedia and certain areas require people to make far more edits than others, it's rather hard to argue that you contribute to Wikipedia if you don't edit it at all. Similarly I don't think someone with zero logged actions can claim to be a vaguely active administrator, whatever areas they work in. Sure, closing an AfD as Keep doesn't count as a logged action, but anyone who spends any time at all closing AfDs will close some as Delete and get at least some logged actions. Other types of admin action will either generate some level of admin actions or won't be the only thing an administrator does. I find it rather hard to believe that the large number of admins who perform few or no admin actions are actually all sitting around reviewing unblock requests or editing protected pages. Hut 8.5 11:52, 27 December 2016 (UTC)
- Commons has had disputes on the inactivity deadminning because it arbitrarily counted some actions and not others. And "no actions" are rather common. It's not a merely philosophical issue. Jo-Jo Eumerus (talk, contributions) 11:54, 27 December 2016 (UTC)
We should adopt the maxim of qui tacit consentire and dispense with the 1 month advance warning. As our own article puts it, "silence gives consent. ......when he ought to have spoken and was able to". No reminders are needed and would prevent the current overly generous and gameable situation. Leaky Caldron 14:32, 27 December 2016 (UTC)
- The more I think about it the more this seems like a good idea. Sending an editor a message informing them they're going to be de-sysopped for being inactive just encourages gaming the system. Sam Walton (talk) 15:26, 27 December 2016 (UTC)
- I have closely monitored the inactivity policy since it was initially created four or so years ago. My concerns then were about administrators gaming the system with 1 edit equaling 1 resysop whenever they please. And just like I suspected, we right now have 1,274 administrator accounts, and 738 of them don't make 30 or more edits every two months. Making it one administrative action, while better than 1 single edit, is not exactly going to prevent any situations that have similarly come up. What is to prevent an administrator logging in to protect their userspace, unprotecting it and logging back out to keep it? And what about performing one trivial CSD and logging out for a year? If we are truly dissatisfied with the current inactivity policy, one administrative action or one anything won't be satisfactory. Being an active member of the community doesn't mean logging in once a year to reclaim your bit, no matter how good that one edit is. If an administrator makes ten edits or deletes ten things for ten years, they are not demonstrating a need for the tools. They also don't have a body of work to base whether they are knowledgeable of policy anymore. There is really only one option here. First, there has to be a set number of logged actions counted every month/6 months/year which count towards activity (preferably 10-12 actions/edits or more). This would give a very small body of work for the community to look at to judge competency when using the tools. The other change that needs to be made is the length of time. It needs to be changed to something such as two years, and after which they can't automatically regain adminship because they have been inactive for too long. One inactivity desysop and then they would need a new RFA. No regaining adminiship, disappearing and reappearing ad nauseam. Also, warning of a pending desysop unless they meet X criteria is counterproductive and encourages the current gaming. Like I said when this discussion happened the first time: years from now, we're going to have a bigger problem than dormant accounts if we don't make it more strict. Right now, we're seeing the consequences of it because it's going to get harder to confirm identities of older administrator accounts when there is no policy to restrict them from getting the tools back. Regards, — Moe Epsilon 14:37, 27 December 2016 (UTC)
- I'm just wondering why inactivity needs tinkering with particularly? Is there any particular reason why the current standards are inadequate? All it seems to do in my mind is to add to red tape. RfAs are at an all time low. An every shrinking pool of active admins. And yet all the recent RfCs and so forth seem to focus on ever decreasing the pool of admins. Until someone can come up with a convincing argument otherwise I shall oppose such measures. --Jules (Mrjulesd) 17:14, 27 December 2016 (UTC)
- Because we're about to give administrator rights to a user who was given the rights to see what they looked like and has hardly edited in 10 years let alone used the tools and that makes some of us uncomfortable. Sam Walton (talk) 17:16, 27 December 2016 (UTC)
- Any particular reason for that? By the way I'm not overall opposed to WP:INACTIVITY clauses per se, but I can't see why it needs to be tinkered with. --Jules (Mrjulesd) 17:23, 27 December 2016 (UTC)
- Because I think it's fair to say that the state of the encyclopedia when this user was active is very different to how it is now, and a 10 year editing gap leaves them almost as clueless as a brand new editor. We're simply trusting that they will read through the rules and understand them well enough to use the administrator toolset properly, which is completely at odds with the high bar set by the community at RfA. There are hundreds of editors who wouldn't pass an RfA due to lack of tenure/experience that I would trust more than someone returning from a ~10 year break. Sam Walton (talk) 17:29, 27 December 2016 (UTC)
- OK well let's say we have this hypothetical admin "x", who hasn't been particularly active for the last 10 years, but has done enough to keep the bit even so. But they want to return to, let's say, WP:AIV or some other area that requires the admin toolset. Now I'm willing to bet that they would reacquaint themselves with the relevant policy and procedures before doing so. Why? Generally people don't change enormously. If they were trustworthy enough 10 years ago it is very likely they are trustworthy now. And even if they made mistakes they would likely learn quickly, or get dragged before a relevant noticeboard to dealt with. The thing is Wikipedia has mechanisms in place for errant admins, whether or not they have been active recently or just returned after a hiatus. For some reason returning admins seem to arouse suspicion, but in my mind they should be encouraged to contribute. --Jules (Mrjulesd) 17:44, 27 December 2016 (UTC)
- This is not a hypothetical situation. People may not change much, but the standards for evaluating trustworthiness have changed drastically since the user in question was promoted via this RfA. At BN, someone stated this user has only made three admin actions ever and none in over a decade. As Sam Walton observed, this user is effectively a brand new editor. No way should we be giving them the mop. Lepricavark (talk) 18:10, 27 December 2016 (UTC)
- I'm not saying hypothetical examples don't occur. I'm just using one to clarify discussion. --Jules (Mrjulesd) 18:34, 27 December 2016 (UTC)
- Fair enough. But I am highly uncomfortable with the notion that we should still trust this user because he was deemed trustworthy at a three-participant RfA from 13 years ago. Standards from 13 years ago are far too light to be a useful reflection on whether Cyp is qualified to be an admin today. Lepricavark (talk) 18:48, 27 December 2016 (UTC)
- This is not about "decreasing the pool of admins." Inactive admins are not doing anything anyway, so desysopping them ultimately has no bearing on actual admin activity. This should not impact our ability to promote new admins. We are not trying to add red tape, but instead we want to refine the present measures relating to how ex-admins are handled. Lepricavark (talk) 18:10, 27 December 2016 (UTC)
- This is about the decreasing the pool of available admins, as I was discussing returning admins. And I see further tinkering as red tape, as I see it as solutions in search of problems.--Jules (Mrjulesd) 18:34, 27 December 2016 (UTC)
- Back in July 2015 I looked at the numbers and 99.9% of administrator actions were carried out by the 45% of administrators who made more than 10 administrator actions in the past year, so tightening the requirements and desysopping a few hundred is unlikely to have a significant immediate impact on admin load. Sam Walton (talk) 18:39, 27 December 2016 (UTC)
- But you'd also be shutting out returning admins. That is my point, it seems to me editors are constantly changing their editing commitments. --Jules (Mrjulesd) 18:45, 27 December 2016 (UTC)
- That is a fair concern, but something must be done to prevent the gaming of the system that is currently possible. Lepricavark (talk) 18:51, 27 December 2016 (UTC)
- This is not a solution in search of a problem. It has been clearly demonstrated above and at BN that the problem is very real. I seriously question whether we want someone who has barely edited at all in the past decade to be in the pool of available admins. The solution to the lack of available admins is to nominate and promote currently active editors. The solution is not to hang on to former admins who longer actively participate on this project. Also, Sam is quite right that desysopping inactive admins will not have an immediate impact. Indeed, I doubt if it would have a future impact either. Inactive admins, by definition, are not actively doing admin tasks. Lepricavark (talk) 18:49, 27 December 2016 (UTC)
- I think we should move away from arbitrary statistical requirements like x edits in y days. We should expect administrators to be active members of the community. That may be too difficult to define or too easy to game to use for desysopping, but could be used for resysopping. If a former admin returns after a long absence, it's not unreasonable to ask that they spend a little bit of time reacquainting themselves with the place before they're restored to a position of trust. Having said that, though, we perhaps ought to be stricter still with admins who got the bit before there was a formal voting process. HJ Mitchell | Penny for your thoughts? 22:33, 27 December 2016 (UTC)
- (edit conflict) Two observations. Not sure these translate easily into formulations for an RfC, and even less sure any is politically feasible, but still.
- "Edit" vs. "administrative action" requirement: if the spirit is that the candidate must show admin activity, it does not really matter whether it is logged or not. Allow any admin action to qualify, and let candidates for a resysop point to the diffs where they closed a keep AfD, denied an unblock request, or whatever, if the logs fall short.
- Edit conflict: basically a reformulation of HJ Mitchell's point above. More importantly, the problem right now is 10-year-old returning admins that may or may not be out of touch with recent changes. I think it could be left to bureaucrat discretion to evaluate ; something along the lines of
provided the bureaucrat is satisfied that the returning administrator has enough familiarity with significant changes in guidelines that occured since the procedural desysopping
(clarify to mean this is a lower threshold than RfA if needed). If a number is absolutely needed to constraint bureaucrats, addas a general rule, a candidate to resysopping with more than x (edits/admin actions) (per year/since the desysop) will usually have kept up with the changes, while a candidate well below this threshold might not have
- what is important is that when the candidate has x+1 edits bureaucrats may still pocket veto the resysop without having to resort to WP:IAR. In the case that made this discussion happen, I feel the bureaucrats most leaning to restore sysop rights do so binded by the guidelines rather than by their judgement.
TigraanClick here to contact me 22:56, 27 December 2016 (UTC)
- I'm not sure the problem is really 10-year-old returning admins; I'd say it's admins who are making one or two token edits a year just to keep the bits. If we manage those accounts more closely - and at this point I think there's a reasonable chance that we could put some additional activity levels on the account, even if it's just 50 edits a year - then we won't have the issue of 10-year-old accounts asking for the admin bit back. I'd also shorten the length of time between removal and reinstatement to a maximum of one year if there are fewer than 50 edits in the intervening year, with only those having 50+ edits/year being eligible to reclaim the bit for up to 3 years. We're one of the few larger projects that doesn't have some sort of genuine activity requirement; one edit a year really isn't evidence of activity. Risker (talk) 04:14, 28 December 2016 (UTC)
- IMO the recurring discussions about how Something Must Be Done about the horrible scourge of people not volunteering enough of their time are far more of a nuisance than anything ever done by the admins in question. I'm still waiting for examples of actual problems caused by this pattern. "It looks like we have more admins than we really do" is nonsense; it's easy to measure who's actually active. "They might screw up" is just handwaving; you need to point to examples of screwups that are at least plausibly related to long inactivity, and demonstrate that these problems are widespread or serious, and demonstrate that your proposed solution is not worse. As someone who did return after long inactivity, I might well have done the "token edit every year" thing if I'd noticed that I should - I'd see the message and, probably, think something like "Oh yeah, I used to do that stuff a lot; maybe I'll get back to it when I have some more time". I always think it's very strange that people skip over that kind of motivation and assume that inactive admins are "gaming the system" or "hanging on to power" - trust me, if you're not really engaged with Wikipedia, then the politics of adminship are Not Interesting. Maybe before we try more invasive solutions to this non-problem, we could invest some of that time in doing something useful like developing an admin-specific (and sanely formatted) version of WP:UPDATE to point returning admins to. Opabinia regalis (talk) 05:54, 28 December 2016 (UTC)
- There is a possibility that we will end up giving the bits to a user who has done almost nothing in over a decade. Do you see why some of us are concerned? I ask because you didn't really address the specific situation that prompted this discussion. Lepricavark (talk) 06:15, 28 December 2016 (UTC)
- This thread was seemingly created specifically to separate discussion of the general issue from discussion of a particular case. But no, actually, neither thread makes a good case for what specifically is a concern other than account compromise, which was handled. Conversations about inactive returning admins usually seem mostly like opportunities for people to advocate their views on the politics of adminship. Opabinia regalis (talk) 19:49, 28 December 2016 (UTC)
- Huh?! Multiple users are concerned because an editor who has not been especially active for over a decade has been granted administrator privileges. This isn't just about account compromise. This is about the blatant lack of evidence that the user is in any way qualified to be an admin. This has nothing to do with politics. Lepricavark (talk) 20:07, 28 December 2016 (UTC)
- This thread was seemingly created specifically to separate discussion of the general issue from discussion of a particular case. But no, actually, neither thread makes a good case for what specifically is a concern other than account compromise, which was handled. Conversations about inactive returning admins usually seem mostly like opportunities for people to advocate their views on the politics of adminship. Opabinia regalis (talk) 19:49, 28 December 2016 (UTC)
- (edit conflict) @Opabinia regalis: Compromised accounts are the biggest problem I have with having many inactive administrator accounts floating around, but I also abide by the philosophy that an administrator serves at the will of the community and should step aside if the community no longer views them as qualified for the role. ~ Rob13Talk 06:17, 28 December 2016 (UTC)
- Every known case of admin account compromise in the last two years was of a currently-active account. There's no evidence that old accounts are more likely to be compromised. More to the point, is being an admin a "role"? I keep trying to push back on this idea that adminship is or should be a social class rather than a form of technical access. All of this stuff about returning admins always seems to be an argument about "this person isn't of the right social class", not "this person would use their tools incorrectly". Opabinia regalis (talk) 19:49, 28 December 2016 (UTC)
- Talk about "promotion" needs to be ended then. Not many more socially divisive term than promoting someone. Leaky Caldron 19:55, 28 December 2016 (UTC)
- I'm starting to think that you've done absolutely no review of the discussion at BN that prompted this discussion, which has nothing to do with social classes and everything to do with concerns that an editor may misuse the tools. The editor in question has not been an active editor in over a decade. There is no evidence that he is qualified to use the tools. I don't know where you are getting this "social class" stuff from, but it is completely irrelevant. Lepricavark (talk) 20:07, 28 December 2016 (UTC)
- Every known case of admin account compromise in the last two years was of a currently-active account. There's no evidence that old accounts are more likely to be compromised. More to the point, is being an admin a "role"? I keep trying to push back on this idea that adminship is or should be a social class rather than a form of technical access. All of this stuff about returning admins always seems to be an argument about "this person isn't of the right social class", not "this person would use their tools incorrectly". Opabinia regalis (talk) 19:49, 28 December 2016 (UTC)
- Take the case that spurned this (where a 'crat restored the bit for what I think are valid reasons). The returning admin will probably not go on a vandalizing rampage, that is for sure, but an RfA candidacy with similar stats would have been a snow fail in the 2010-2016 period (at least). I am all for weakening RfA standards, and
I'm still waiting for examples of actual problems caused by this pattern
is a powerful argument for that. But I do not agree with doing so by the backdoor, clinging to a procedural policy that lets older accounts get away with not less, but much much less qualifications that the recent promotions. If anything, that is not fair. - My $0.02: this is only one head of the hydra born from the hysteresis in current adminship requirements, where you need outstanding contributions to pass but outstanding misconduct to be desysopped, so that adminship is effectively a big deal and for life. The only solution remotely feasible I see out of this is to make administrator stand for reconfirmation at regular intervals - with a much simpler process and still a higher threshold to remove an incumbent than to block a candidate, but something that forces them to defend their actions if they want another mandate, and an opportunity to mount a desysop challenge with a lower "personal animosity" malus. TigraanClick here to contact me 17:16, 28 December 2016 (UTC)
- "Wouldn't pass today's RfA" is true of lots of currently active admins. The huge majority became admins at a time when RfA standards were much more reasonable. The conclusion to be drawn from that is not that we should start retroactively applying modern standards to old admins, but that modern RfA standards are not evidence-based. (And by that I mean "a steaming pile of useless bullshit" ;) Opabinia regalis (talk) 19:49, 28 December 2016 (UTC)
- There is a possibility that we will end up giving the bits to a user who has done almost nothing in over a decade. Do you see why some of us are concerned? I ask because you didn't really address the specific situation that prompted this discussion. Lepricavark (talk) 06:15, 28 December 2016 (UTC)
While rooting around, I found an interesting statistical anomaly that I thought I would share. I was going to compare the (in)activity of administrator accounts shortly after creating the inactivity policy to the present day. On July 31, 2011, we had 1,541 administrator accounts broken down as such (according to WP:LA standards): 760 active administrators, 553 semi-active, 228 inactive. During the most recent round, we had 1,274 administrator accounts: 537 active, 551 semi-active and 225 inactive. It's interesting to say the least. Regards, — Moe Epsilon 13:08, 28 December 2016 (UTC)
- The current thread at WP:BN is the perfect example of the weakness of our surrent policy on inactivity. That user was never an active admin by any reasonable standard. They used their tools to delete three pages in 2005. That's pretty much it. No blocks, no protections, no changing of user rights, etc, one single time. No edits anywhere in project space for a period of over ten years. Didn't know until yesterday that rollback is no longer only for admins. (that happened in 2008) And yet, because they made this edit in 2014 they still qualified for an immediate resysop for another three years. That's just ridiculous. This situation is limited toa small number of admins, but it is a loophole that needs to be closed. We should not have admins who never really were admins coming back after ten years of almost no activity at all and all they have to do get all the tools is ask and then wait 24 hours. Beeblebrox (talk) 21:44, 28 December 2016 (UTC)
- Let’s take stock of what’s been assumed or said up to now and dispel some conjecture:
- This is not a hypothetical issue. It really happened, but some commenters have not read this request for resysop, the discussion that led to this. And that’s why it needs to be brought into ‘legislation’.
- It’s not about decimating the pool of active admins. A non active admin is a non available admin.
- Let’s be clear about what constitutes an admin action that counts, and make a responsible decision about it. Risker’s suggestions appear to be realistic. Admin stats per X-Tools or {{Adminstats}}, while tasks that can be done as NAC should not count.
- Sending reminders to admins about to be desysoped encourages gaining the system: probably - run a check against those who have done so and see how many admin actions they have done since.
- There should be more than 1 logged admin action to prevent desysoping.
- All cases of compromised accounts in the last 2 years were to active admin accounts.
- My opinions (FWIW):
- I don’t see any evidence at all supprting this comment ’’All of this stuff about returning admins always seems to be an argument about "this person isn't of the right social class", not "this person would use their tools incorrectly” by Opabinia regalis.
- Term duration would require term limits which again would discourage people from running or re-running - something we cannot afford right now.
- There is no evidence that the huge majority became admins at a time when RfA standards were much more reasonable - have they ever been reasonable since 2007? Are we not confusing standards with the perennial unpalatable environment at RfA that no one wishes to address when mentioned in the appropriate venues?
- There is no evidence that ‘modern’ standards are "a steaming pile of useless bullshit”. There are indeed occasions when a voter’s oppose is based on some ludicrous criterion such as not having been registered for at least 3 years, or not having created 25 articles, or not having made 10,000 edits, or having no GA or FA, but these are relatively rare, come mainly from children, raw newbies, or trolls, and such anomalies have been on RfA since its dawning.
- There is no evidence whatsoever that it is generally harder to pass RfA today, based on user criteria, than it was in the threshold year of 2007. A problem arises however, when a disingenuous vote gives rise to a lot of pile ons.
- Every Wikimedia project has something different in the way it manages its administrator systems; in this instance, making comparisons/drawing parallels doesn’t help
- Back on topic, Beeblebrox sums up perfectly, among all the of-topic stuff, what this discussion is all about, and that's what an RfC should hinge on.
- Kudpung กุดผึ้ง (talk) 03:57, 29 December 2016 (UTC)
Another alternative
Just throwing this out there without comment on how I feel about it. On Commons, they have requests for de-adminship, which is wide open for use for any purpose. That will never fly on enwiki, but how about requests for de-adminship which can only be opened on an administrator with no logged administrative actions in the past year or fewer than 10 logged administrative actions in the past two years? Basically, in addition to the "automatic" inactivity desysops, allow the community to examine cases of administrators with extremely low levels of activities if a community member decides to initiate that. This would allow a more tailored look at the editor's contributions, whether they're out-of-touch due to inactivity, and whether they're acting in an administrative capacity. ~ Rob13Talk 06:15, 28 December 2016 (UTC)
- I don't think this is the right approach. It would be be better to tighten the inactivity desysop settings so that everyone is playing by the same rules. We probably don't want to create a system where some inactive admins lose the mop and others don't. While I strongly believe we should have requests for de–adminship on this project, using such a system for inactive admins would needlessly personalize the process. Personalities and friendships should not come into play in the case of an individual who is no longer active, but I can envision a scenario in which an inactive admin account retains the bit at such a request for de–adminship because the admin in question had made many friends. In other words, I am concerned that requests for de–adminship would turn something that should be routine into a popularity contest. It would probably be best to begin drafting an RfC for the purpose of tightening inactivity desysop settings. Lepricavark (talk) 16:44, 28 December 2016 (UTC)
This issue is bigger than expected
I have just become aware of User:JamesR/AdminStats. It is a series fo sortable tables chronicling all admin actions by current admins. It has really opened my eyes to the scope of admins who are not, and never were actually active in administrative functions. The last table, listing all actions by all current admins is the most informative. There are over 100 current admins who have only used their bits one time, and hundreds more who have used them ten times or less over their entire tenure.
Adminship is not a trophy to be put on a shelf for bragging rights, any holder of advanced permissions should be expected to actually use them once in a while. It seems clear that over the years we have selected a large number of admins who it turned out didn't really want to be admins, didn't have the stomach for it, would rather do content work, whatever. If they had just left they would have been removed for inactivity, but apparently large numbers have remained at least marginally active but are admins in name only and have not made siginificant contributions in administrative functions.
This is a probelm in that it creates a security risk while providing no benefit, and it artificially inflates the number of admins, creating the appearence that we have several hundred more admins than we actually have. There is no reason for anyone who has been an admin for a year or more and only has one admin action in their logs to retain their advanced permisssions. Beeblebrox (talk) 00:56, 2 January 2017 (UTC)
- @Beeblebrox: You might find the adminstats tool useful - it's a bit more customisable than those tables. I quite like the statistic that 45% of administrators carry out 99.9% of all administrator actions, with the other 55% making less than 10 admin actions per year. Sam Walton (talk) 01:08, 2 January 2017 (UTC)
- Though I've just noticed that tool isn't correct anymore. It's tracking the 'redirect over move = G6 deletion' thing as an admin action and inflating the number of users listed there... Sam Walton (talk) 01:14, 2 January 2017 (UTC)
- ...Sort of (T154408). Sam Walton (talk) 01:20, 2 January 2017 (UTC)
- Though I've just noticed that tool isn't correct anymore. It's tracking the 'redirect over move = G6 deletion' thing as an admin action and inflating the number of users listed there... Sam Walton (talk) 01:14, 2 January 2017 (UTC)
- Beeblebrox' thoughtful and well-written comment makes me, for what little it's worth, wade in to the fray with a contradictory opinion. I think there's an implicit vision that to "Be an admin" means you're going to spend significant time running around the 'pedia doing adminny type stuff, and that if you're not, you're not really an admin, and it would be better/safer/more accurate/cleaner to take the admin bit away. That model (wielding the mop regularly and with dedication) is one version of what adminship means. But it's not the only one. An admin is someone who has the trust of the community to not screw up with extra tools and to even do some good. If they've done that x times, it's always a shame they haven't done x+1 times, or 10x times. But even if x=1, it's a net positive. And while we might debate whether it was worth the community's time to review someone at RFA if once promoted x is too small, it's even more guaranteed to be a waste of time if we cut them off. I'm happy if we have an admin who once in a blue moon deletes some obvious spam or blocks a disruptive vandal that needs it - that's one less action someone else needs to do.
- I'm sensitive to security issues and so administrative, easily-reversible bit removal after inactivity is a sensible thing. However, for the above reasons, I'd advocate the community (and therefore bureaucrats) continue to be very lenient in returning the bit for admins/users "in good standing" (however defined) even after lengthy inactivity. However, since the admin bit is an expression of trust, I would be less lenient about returning the bit in the case of "under a cloud", where plausibly that trust has actively been lost. [Noting I have read the Cyp example.] Martinp (talk) 18:38, 3 January 2017 (UTC)
Publicising this now the holiday period is over
I think more voices would be useful. Any appetite for posting this discussion at BN, maybe CENT... anywhere else? --Dweller (talk) Become old fashioned! 10:53, 3 January 2017 (UTC)
Bold edit
I added "When an administrator has had an emotional reaction to something they've dealt with in a purely administrative role then this caveat does not apply". Maybe someone who is a native speaker can phrase this more eloquently. (((The Quixotic Potato))) (talk) 06:18, 29 December 2016 (UTC)
- WP:BOLD is a great policy for article editing. As is reflected in the bold editing guideline, it's not such a great idea in an important policy like this. Your addition seemed badly worded and not necessary, so I have reverted it. It ddin't seem to add anything crucial to the policy and was not discussed. Beeblebrox (talk) 13:14, 29 December 2016 (UTC)
- Agreed with reverting this. I can see the ANI posts now conjecturing about how an administrator was definitely involved in a situation because they probably maybe might have had some feelings about it. Sam Walton (talk) 13:29, 29 December 2016 (UTC)
- I hadn't seen that thread, but now that I'v had a look at it I would add that it is always a bad idea to edit policy due to one specific incident that you were involved in and can be considered WP:POINT editing. Beeblebrox (talk) 02:59, 30 December 2016 (UTC)
- Agreed with reverting this. I can see the ANI posts now conjecturing about how an administrator was definitely involved in a situation because they probably maybe might have had some feelings about it. Sam Walton (talk) 13:29, 29 December 2016 (UTC)