If a former administrator ("lengthy inactivity") or bureaucrat ("inactive bureaucrat accounts") has been inactive (defined by zero edits or logged actions) for a period of two years or longer after the removal of permissions (or for two years from the last edit or log action in the case of permissions removed due to inactivity), they must be successful in a new request for adminship or bureaucratship to have the permission(s) restored.
However, the preceding section states:
If an inactive bureaucrat returns to Wikipedia, they may request restoration of the permissions at the bureaucrats' noticeboard provided they have not been inactive from bureaucrat activity for three consecutive years.
Over two years with no edits. If an editor has had at least two years of uninterrupted inactivity (no edits) between the removal of the admin tools and the re-request, regardless of the reason for removal, the editor will need to request reinstatement through the WP:RFA process. In the case of an administrator desysopped due to a year of inactivity, only one year of continued uninterrupted inactivity (no edits) from the removal due to inactivity is required before a new WP:RFA is necessary.
Here are my questions:
Suppose a bureaucrat makes their last edit or logged (administrative) action in February 2024. This edit or logged action counts as bureaucrat activity. The bureaucrat flag is removed after 13 months (March 2025). After another 13 months (April 2026), the former bureaucrat asks only for their bureaucrat flag back—not their administrator flag. May a bureaucrat give it back, given that it has only been 26 months, not three years?
Suppose an administrator falls short of the 60-month editing minimum in February 2024 and stops editing at that point. The administrator flag is removed after three months (May 2024). After 23 more months (April 2026), the former administrator asks for their administrator flag back. May a bureaucrat give it back, given that it has only been 23 months from when the administrator flag was removed, as opposed to 26 months from the last edit or logged action?
It is an inconsistency but not a contradiction; one is not required to be an admin to be a 'crat; your first hypothetical is a "yes", though it will likely be changed now that someone's noticed. Your second is also a yes, because it's a "both" situation. In neither situation, however, would I expect the 'crats to not grant such requests without significant editing on the part of the recently-returning admin (which we have done in the not-distant past). Primefac (talk) 07:08, 27 February 2024 (UTC)[reply]
I think the answer to the first question has to be "no": point 5 of WP:RESYSOP is pretty clear that the two-year rule applies to bureaucrats too, so it makes more sense to treat provided they have not been inactive from bureaucrat activity for three consecutive years as a necessary rather than a sufficient criterion. (This could definitely be made more clear, though.) I agree that the answer to the second question is "yes". Extraordinary Writ (talk) 07:30, 27 February 2024 (UTC)[reply]
Thank you for confirming that "two years from the last edit or log action", as opposed to "one year after the removal of permissions", is not in line with policy, at least for administrators. I was about to go ahead and edit the page, until I realized that there are at least three more questions that should be answered:
In the second question, replace "administrator" with "bureaucrat". Would the answer remain "yes", assuming that the last edit counts as bureaucrat activity?
Suppose that, in February 2024, a bureaucrat closes a successful RfA and uses the Special:UserRights tool to promote the user to admin, then makes 100 edits. They are largely inactive for the next 63 months—they make one edit to an unprotected article every other month or so, close two unsuccessful RfAs in February 2026 and February 2028, but do nothing else, leading to the removal of their bureaucrat flag in May 2029 for failure to meet the 60-month editing minimum. In June 2029, the former bureaucrat increases their editing activity to a significant level, and the next month (July 2029), they ask only for their bureaucrat flag back—not their administrator flag. Should a bureaucrat give it back, given that it has only been 17 months since the most recent bureaucrat activity, rather than 65 months since the last logged administrative action?
If the answer to the above question is "no" (meaning the five-year rule applies to bureaucrats, assuming that closing an unsuccessful RfA doesn't count as a logged administrator action), would an action such as deleting a page count as a logged administrative action for the purpose of the five-year rule as applied to bureaucrats? For example, if in addition to closing an unsuccessful RfA, the bureaucrat (as an administrator) also deleted a page, would the answer then become "yes"?
I think I'm going to have to recant on my answer to your original second question: WP:RESYSOP does seem to contradict WP:ADMIN#Restoration of admin tools (or at least is very ambiguous) as to when the two-year clock starts for admins desysopped under the 100/60 rule. The two need to be reconciled one way or another. As for your three new questions: 1) I don't see why bureaucrats would be treated differently than admins on this issue, 2) yes, I don't see any current policy basis for applying the five-years-since-last-admin-action rule to crats (for better or worse) and 3) not applicable given my answer to #2. In terms of fixing the ambiguity I mentioned earlier, would anyone oppose changing In the case of an administrator desysopped due to a year of inactivity, only one year of continued uninterrupted inactivity (no edits) from the removal due to inactivity is required before a new WP:RFA is necessary at WP:ADMIN to "In the case of an administrator desysopped due to inactivity, the two-year clock starts from the last edit or log action prior to the desysop" to align it with WP:RESYSOP and account for the new 100/60 rule? That would resolve the ambiguity, I think (although my eyes are slowly glazing over, so it's possible I missed something). Extraordinary Writ (talk) 00:26, 29 February 2024 (UTC)[reply]
I'm going to be honest, these are insanely specific hypotheticals which I have never seen in my 5+ years as a 'crat. That being said, we (the 'crats) are pretty good at figuring out how the rules apply to a situation, which is one of the reasons why there is a hold on all re-sysop requests. As long as the policies are all in alignment with each other (which, as we have found, might not be the case) we shouldn't be playing "but what if" for every possible scenario given that 90% of them are almost guaranteed not to happen. Primefac (talk) 07:26, 29 February 2024 (UTC)[reply]
I have no doubt that the 'crats will make good decisions, even in unusual cases, along the lines on which the community has agreed. The reason I brought this matter up is that I have been writing a page summarizing the inactivity policy for all local groups that have it (draft now posted at User:PleaseStand/Removal of permissions due to inactivity). The inactivity criteria themselves are clear enough (save for some insignificant points, such as "administrative actions" vs. logged actions). The prior notification requirements perhaps could be a little more specific, though it's not a big deal. Where I ran into this issue was the section in which I try to describe the re-sysop criteria for administrators and bureaucrats together. Unfortunately, given the apparent inconsistencies between the two pages and within this page (which is only a "summary of information", not a policy page), I cannot merely refer to WP:RESYSOP. PleaseStand (talk) 06:16, 1 March 2024 (UTC)[reply]
I think a wider discussion on updating Wikipedia:Bureaucrats#Inactive bureaucrat accounts is in order; some of the things which count as "crat activity" are wildly out of date. SUL finalization is a decade old at this point, so responding to requests in their capacity as a global renamer is hardly indicative of enwiki 'crat activity. Removing signalling that they remain actively engaged and available for bureaucrat tasks as a way of being considered "active" is also worthy of discussion.
While we are at it, I think we should also discuss allowing 'crats to de-crat. I don't see any real discussion of this issue in the archives of this page, but I think it makes sense for the same reasons we gave 'crats the ability to desysop: namely, it would unify the user rights log (so +crat and −crat actions are recorded in the same place). And I do not think this would be a "security risk": I am infinitely more worried above improper granting of the bureaucrat flag than I am improper removal of the bureaucrat flag (to say nothing about other more sensitive abilities 'crats have). To address the elephant in the room, this is listed at m:Limits to configuration changes. However, that page is not a policy; it only describes the practice of how system administrators handle configuration changes. Given that many wikis (such as ptwiki) already give 'crats this power, I think there is a good chance we will be allowed to deploy such a change; the worst case scenario is we get told "no".
HouseBlaster, thank you for starting this discussion; I agree it is overdue. I am 100% in favor of ’crats being allowed to remove the bureaucrat bit when needed, primary because it increases transparency: user right changes made on Wikipedia (as opposed to meta) show up in the local user rights log. Yes, there is a link on that page that takes you to the global logs, but why not just keep everything on-wiki that we can? As you say, allowing ’crats to remove the administrator bit was a positive change that (as far as I know) caused no problems whatsoever. 28bytes (talk) 01:41, 3 May 2024 (UTC)[reply]
I also agree that renames should no longer count as bureaucrat activity, since those things have been uncoupled for many years. 28bytes (talk) 01:43, 3 May 2024 (UTC)[reply]
I agree that allowing a crat to do -crat is not a material security concern, it does mean that a compromised crat account can, in theory, de-crat every other crat and then roam free until someone finds a Steward to de-crat the rogue crat. Not very likely especially if the MFA rules are enforced for crat accounts. MarcGarver (talk) 07:59, 3 May 2024 (UTC)[reply]
A -crat action is relatively rare, don't see a need to request an exception from filing SRM's when the situation arises. — xaosfluxTalk09:31, 3 May 2024 (UTC)[reply]
Allowing crats to de-crat other crats probably wouldn't help security much. Imagine a bad actor getting access to a crat account, then de-cratting all the other crats. –Novem Linguae (talk) 12:49, 3 May 2024 (UTC)[reply]
(Replying to NL) I don't think it would help security much (if at all), but it would hardly make the current situation worse. As I said in the original message, I am infinitely more worried about a rogue 'crat granting people cratship than removing 'cratship.
(Replying to people in general) I realize that we don't have a high volume of 'crat removals, but I think it would be good to have 'crat removals logged locally when they do happen. It is not a "we are bothering the stewards too much" argument, but a "the addition and removal of the 'crat bit should be logged in the same place" argument. HouseBlaster (talk · he/him) 13:05, 3 May 2024 (UTC)[reply]
Yeah. I've always thought it self-evidently poor design that local rights changes of any sort - even if they still have to be made on and logged at meta - don't also create entries in the local user rights log. —Cryptic19:52, 3 May 2024 (UTC)[reply]
I think it would make sense to allow crats to -bureaucrat locally. Not because stewards take forever to action those requests or for security reasons or anything like that, I just don't think that enwiki bureaucrats need steward oversight on that area of permissions management. -- Ajraddatz (talk) 16:24, 3 May 2024 (UTC)[reply]
Support removing global renames as clearly not a 'crat function. Oppose local de-cratting as unnecessary. Oppose removing "signalling that they wil be actively engaged" as the 'crats are a small group with a few very specific tasks, and I can assure you that ArbCom uses this exact reason to exempt members of the functionaries team from their activity requirements pretty regularly. I'm usually a bit of a "hawk" on activity issues but I think we can be a little flexible here. Just Step Sidewaysfrom this world ..... today22:28, 3 May 2024 (UTC)[reply]
Support. If I'm reading it right, the crat activity requirements are only slightly more than the admin activity requirements (the if a bureaucrat does not participate in bureaucrat activity for over three years, their bureaucrat permissions may be removed part of it). I think simplifying things so they are equal to the admin activity requirements is reasonable and makes things less complicated. –Novem Linguae (talk) 12:55, 3 May 2024 (UTC)[reply]
Thinking about it, I would actually support this. I am still of the belief that "if we do something we may as well do it right" – in this case, if we are going to measure 'crat activity, we should not count renaming.HouseBlaster (talk · he/him) 16:30, 3 May 2024 (UTC)[reply]
I'm opposed to anything which links admin and crat. In reality only admin have become crats but we've also had at least one non admin crat (Xeno who could and eventually did ask for sysop back). Best, Barkeep49 (talk) 18:04, 3 May 2024 (UTC)[reply]
I know I'm not Barkeep49, but that does raise an interesting question: if there were theoretically a non-protest non-admin 'crat (e.g. I decided I didn't want to be an admin but I still wanted to do BAG stuff so I keep 'crat), and no 'crat activity requirements existed, would I still be held to "admin" standards? I have no issue with both roles having the same requirements, but in this not-totally-unreasonable hypothetical situation, it seems strange to have a non-admin held to admin standards. Primefac (talk) 12:03, 14 May 2024 (UTC)[reply]
A crat with minimal admin activity would still be an admin so that's not relevant to what's being proposed here or my comment. Best, Barkeep49 (talk) 12:08, 14 May 2024 (UTC)[reply]
I've brought up this idea before at least three times in different contexts, though I can only immediately find the most recent. Each time it's gone down in flames because people think it's somehow a good thing that crats could make a statement by resigning admin in protest (but make it meaningless by keeping crat), or that someone could pass an RfB despite being too timid to try RfA (or actually trying and failing), or - new in that last discussion - that arbcom could create a Super Super Mario effect by decratting while leaving admin in place. —Cryptic19:46, 3 May 2024 (UTC)[reply]
This makes sense to me as well. We did talk about theoretically Arb only removing one bit, due to a reduction in the level of trust that wasn't enough to desysop someone, but still was enough to lose the Crat bit (splitting hairs a bit, but it was all theory). But really, being a Crat is mainly just about being trustworthy, the ability to stick to the letter and spirit of policy and be highly accountable. As long as they are active enough to keep the admin bit, and nothing is demonstrated they have lost the higher level of trust from the community, that should be enough. Yes, this does provide a loose link between admin and crat, and some don't want to do that, but realistically, they ARE linked in practice, even if the letter of policy doesn't explicitly say they are. Xeno may have experimented with being a Crat without being an admin for a bit, but we have never promoted a Crat that wasn't an admin at the time, just as we don't have CUs or OSers that aren't admin. We could, but we don't because the lack of admin tools would hinder their ability to perform their tasks. So the admin bit is somewhat linked to Crat, not as a policy requirement, but as a practical requirement. Dennis Brown - 2¢22:20, 3 May 2024 (UTC)[reply]
SupportIff it is formally tied to the same standards as adminship and one would for sure be removed from both groups when desysopped. We have had at least one non-admin 'crat over the years but that was an exceptional circumstance. Just Step Sidewaysfrom this world ..... today22:32, 3 May 2024 (UTC)[reply]
I think that some want to have the two more isolated, and some want to keep them tied together. Perhaps a fear out of Crat being viewed as a "senior admin" position. As a practical matter, I agree they are very tied and we shouldn't try to artificially decouple them with separate criteria for activity. Dennis Brown - 2¢23:23, 4 May 2024 (UTC)[reply]
I think there are three separate but related questions here that needs answering before we can decide what standards should apply:
Should a crat who is deadminned for inactivity have the crat tools removed at the same time?
I think nearly everybody would answer "yes" to this.
Should there be specific (additional) activity requirements that need to be met to retain the 'crat tools?
This discussion suggests to me that "yes" would probably have a slight majority in a straight vote but I'm not seeing a clear consensus.
If the answers to the above are "yes" and "no" respectively, what activity requirements should apply to a non-admin crat?
There hasn't really been discussion of this (as we have no non-admin crats this is currently entirely theoretical anyway).
That sums it up nicely. 3 is the tricky one, but the least likely to be an issue as a practical matter. Not convinced we need a rule for that. Dennis Brown - 2¢23:53, 4 May 2024 (UTC)[reply]
The simplest way to handle 3 would be to apply the standards as if they were an admin, i.e. if they make no edits or logged actions for one year or fewer than 100 edits/5 years then remove the crat tools for inactivity. The tools/bots that track activity would require a bit of tweaking but I'm guessing that wouldn't be particularly difficult. In addition to simple it also seems reasonable to me, but of course others may think differently. Thryduulf (talk) 01:24, 8 May 2024 (UTC)[reply]
Getting back to basics, the underlying idea is for a user to have some level of involvement with the community in order to maintain advanced privileges. Editing activity level is being used as one measure. If the degree of involvement expected from an admin or a bureaucrat is the same, then the same measures can be used for either, whether or not a given user holds one or both sets of privileges. If more involvement is expected from one role versus the other, then two separate measures can be used. isaacl (talk) 15:11, 13 May 2024 (UTC)[reply]
For admins the desire is actually two-fold - admins should have some continuing level of involvement with the community (for which the easy-to-measure editing activity has been deemed an appropriate proxy) but also continued familiarity with the policies and expectations applicable to admins. The adopted measure for the latter is a combination of editing activity and logged admin actions. Thryduulf (talk) 16:41, 13 May 2024 (UTC)[reply]
Yes, for simplicity I only listed one of the criteria. The point is the measures are being used to establish if certain characteristics are met that are considered necessary for retaining privileges X and Y. If the required characteristics are different for X vs Y, then different measures may be needed. Otherwise, the same measures can be used. This avoids any assumption that holders of Y also hold X. isaacl (talk) 18:12, 13 May 2024 (UTC)[reply]
@Dweller I really don't understand why this is an issue to obsess about. We don't have many 'crats and when have we had rogue 'crats? The only time I can remember an actual rogue 'crat was in the very early days of 'cratship, when there was an argument on whether a certain persn should be an admin, more or less because he was "famous", a discussion I participated in. One of the 'crats decided to simply promote him because he. personally, wanted him to be an admin. That admin and crat were both deleted. End of story.
I think Dweller's proposal to remove activity requirements to retain bureaucrat privileges aligns with your viewpoint. isaacl (talk) 15:02, 13 May 2024 (UTC)[reply]
Comment - Should this be advertised more, or converted into a full RFC? I think we have a nice start to forming consensus, but I think more participation would be desirable. I would hate to lose the momentum and have it peter out without a consensus one way or another. Dennis Brown - 2¢10:08, 14 May 2024 (UTC)[reply]
Support removing these redundant rights. It makes it easier for users to understand what the job of a 'crat is (especially when compared to regular admins) when their permissions are kept to what they do in practice. I also support removing (override-antispoof), (move-subpages), (noratelimit), (suppressredirect), and (tboverride), which are remnants from the days when they were responsible for renaming users, and have nothing to do with the modern duties of a bureaucrat. All 'crats are admins who already possess these rights. If a 'crat wants to turn in the mop temporarily (while keeping 'crat), they can be given WP:PMV which covers most of the rights. HouseBlaster (talk • he/they) 06:07, 19 December 2024 (UTC)[reply]
I think it should probably be changed to granting/revoking administrator, interface administrator, bot/copyvio bot, and extended confirmed (EC because of adding and removing EC upon adding/removing administrator rights), and of course granting bureaucrat. The rest are seemingly holdovers from the old days. The permissions changes seem fine, though perhaps somewhat redundant given they're a mix of page mover and account creator permissions anyways. EggRoll97(talk) 17:59, 20 December 2024 (UTC)[reply]
Pretty sure these are inherited from mediawiki default - changing them means maintaining a special config to have them removed on just this project. This seems to be a make-work request with no practical benefit. — xaosfluxTalk21:48, 20 December 2024 (UTC)[reply]
