Wikipedia talk:Administrative action review

RFC text

From Wikipedia:Requests for adminship/2021 review/Proposals:

Create a new process, Wikipedia:Administrative action review (XRV), that will determine whether an editor's specific use of an advanced permission, including the admin tools, is consistent with policy. XRV will use a structured discussion format, open to all editors and closed by an uninvolved administrator, to reach a consensus on whether an action or set of actions is endorsed or not endorsed. Acting on this consensus, if necessary, is deferred to existing processes.

• The goal of XRV is to provide a focused and constructive venue in which admins and other advanced permissions users can be held accountable to the community.
• Any action, or set or related actions, requiring an advanced permission and not already covered by an existing process (e.g. WP:DRV for deletions), may be referred to XRV.
• A structured discussion format, closed by an uninvolved administrator, will be used to reach a consensus on whether the action should be endorsed or not endorsed.
• Participation in XRV is open to all editors.
• The purpose of XRV is solely to reach a consensus on whether the use of the permission was appropriate, not to remove permissions. Acting on that consensus is deferred to existing processes:
  • Individual actions that are not endorsed can be reversed by any editor or administrator;
  • Permissions granted at WP:PERM may be revoked by an administrator if XRV finds them to be misused;
  • Repeated or egregious misuse of permissions may form the basis of an WP:AN, WP:ANI, or WP:RFARB report, as appropriate.

References

Proposal at village pump

See Wikipedia:Village_pump_(proposals)#RFC:_Shut_down_Wikipedia:Administrative_action_review where it has been proposed that this board be closed. 2409:4071:4E12:1DCA:0:0:4388:2201 (talk) 18:52, 12 June 2022 (UTC)

Lets revive XRV

The village pump proposal mentioned above has been closed per IAR. No problem, I dislike wiki-bureaucracy and think we should be using IAR more often, like how this board is defacto dead and could use with an IAR application of Template:Historical.

Anyway if it is worth keeping this open, lets at least try to get it active again. It was mentioned in the village pump that one of the main reasons why this has become inactive is that this does not have prominent incoming links from other pages. For a brief period XRV was linked from ANI header and it got a flurry of activity. I think we should add the links back to attract more attention here. Thoughts? 2409:4071:E9B:BF48:0:0:43C8:C305 (talk) 16:00, 13 June 2022 (UTC)

I disagree; those links were removed for the reasons outlined at Wikipedia talk:Administrative action review/Archive 2#Two suggestions (and after a consensus to do so was established). I don't think we should re-add the links until this is ready for prime time. And I don't think this will be ready for prime time until my comments at Wikipedia talk:Administrative action review#Forcing speedy archiving of AARV discussions above have been addressed: "We still don't know if this is supposed to be (a) a process somewhere between ANI and ArbCom (to quote the RFA RFC close), or (b) a lightweight process to review one single admin action (to also quote the RFA RFC close). It can't be both. We still don't know if it's going to be modeled on AE or ANI or RFAR or DRV. We still don't know if it's literally going to be open to reviewing every single disputed rollback (which, IMHO, would be crazy, but has been seriously proposed, and is the simplest way to read the RFA RFC close)". Those are fundamental issues; they are not nits that we can pick after AARV is up and running. One of the reasons given in February for not having a series of RFCs about some of these issues was that it would take too long. But it has now been 4 months with no action to try to figure these mission-critical issues out. If you want to try, it seems to me a much more logical starting point would be to hold one or two RFCs to clarify the fundamental issues I raise above. If you go that route, I'd recommend the RFCs be crafted with care, so they aren't derailed right out of the gate. Caveat: others who opposed AARV as it was initially set up may have different priorities than me, so there's no guarantee AARV can be revived; the result of the RFCs might be to continue to have it twist slowly in the wind.--Floquenbeam (talk) 16:50, 13 June 2022 (UTC)

One of the problems is that not everybody agrees that those fundamental issues actually exist (see comments in the village pump discussion). Unless and until there is agreement on what problems exist then I don't think an RFC to fix them has any hope, which in turn means that I don't think XRV has any hope. Thryduulf (talk) 17:16, 13 June 2022 (UTC)

@Floquenbeam @Thryduulf so maybe that's the RfC question. Spitballing here but something along the lines of "Should XRV be developed by BOLD editing, using the RfC as a starting point with further work made through the traditional consensus making process or through discussion of major issues followed by an RfC to ensure there is consensus before the board starts operating". Best, Barkeep49 (talk) 18:19, 13 June 2022 (UTC)

I don't know, Barkeep, my first thought is "it seems pretty clear how that would turn out, but I could be wrong, and in any case, I can't stop anyone from starting an RFC". But then I look below at North8000's accusations of bad faith (which echo Joe Roe's accusations of bad faith at the Village Pump discussion, and several others' accusations above and in the archives), and S. Marshall pretending I didn't already list the issues I'm concerned about, and wonder why I'm supposed to continue to engage in good faith when a majority of those who think it should be live, as is, constantly make bad faith accusations about my motivations and play passive-aggressive games. It's almost a self-fulfilling prophesy; eventually, people who oppose making this page live as is, are going to stop engaging and just oppose everything. That would not be their fault. it will be the fault of the people making the bad faith accusations and playing passive-aggressive games. I don't expect you to fix that problem - you can't control other people - but I want you to understand why I might choose to stop talking about it with you. --Floquenbeam (talk) 18:55, 13 June 2022 (UTC)

I didn't have anybody particular in mind, and my apologies if I was not clear enough on that. Also, if it does crop up somewhere, it's just human nature, not something severe like bad faith. North8000 (talk) 19:44, 13 June 2022 (UTC)

There are two issues here, Floq, that make your position really seem like bad faith even if it isn't. First, when we already have a consensus among many editors to do something, it shouldn't be blocked because one or a small group of editors have objections. Second, the issues you assert we "still don't know" where in fact clearly stated in the original RfC. Levivich has explained this excellently below, but to summarise: XRV is a lightweight process to review one single admin action (or a set of related actions); should be modelled on structured processes like DRV and AE and not from unstructured processes like AN(I); can be used to review any use of any advanced permissions (whether you personally believe that to be "crazy" or not, it's what was agreed). From my perspective, the good faith way to raise concerns like yours is to say, I think these aspects are problems, can we do something about them? Not: if something sounds strange or unclear to me it must be a fundamental flaw, and the entire process must halt until I am satisfied it has been fixed. – Joe (talk) 09:10, 14 June 2022 (UTC)

The relatively complex proposal that this came from is inevitably not going to be perfect. In reality, about 95% of what the literal wording covers doesn't need this venue. We should take the 5% that does need it and develop along those lines. IMO it is admin mishandling of things (via tools or roles) that is not egregious enough to get other admins (who have a natural reluctance to do so) or arbcom to fix the situation. And we need to recognize the gorilla in the living room that there are people who do not want such a review to exist and can use lots of wiki-clever ways to work towards that end. North8000 (talk) 17:54, 13 June 2022 (UTC)

• Well, perhaps we can get a list of these so-called fundamental issues.—S Marshall T/C 18:20, 13 June 2022 (UTC)
• Unless, of course, we're saying that Floquenbeam's list of concerns is comprehensive? I wouldn't want to answer those and then get handed a whole new list of fundamental issues, so let's put all the fundamental issues on the table now.—S Marshall T/C 18:48, 13 June 2022 (UTC)

  That's not a comprehensive list of concerns (see the archives for other lists, posted multiple times, by multiple people.), and nobody can say for certain that the outcome of one RfC will not lead to other questions. Thryduulf (talk) 21:21, 13 June 2022 (UTC)

Well, OK, over my whole wiki-life here are the approx 10 cases I can think of that I probably would have or should have bought here which no other venue would have properly handled:

1. A type A personality involved admin badgering a meeker admin into a wrong decision. Saying multiple times "this is what I'd do if I wasn't involved"
2. A nice-person admin was both incompetent on analysis took a wrong (non-tool) admin action because a wiki-friend of theirs involved in a dispute led them to do it.
3. About 5 cases of an admin using their imprimatur to intimidate the other editor in an editor-to-editor debate/dispute that they were engaged in.
4. A highly respected admin went rogue and reverted a proper admin close on a major RFC saying "it should have been a panel" and started admin saber-rattling against people who reverted their revert. Nobody stopped it until Jimbo did.
5. About 2 times a well-meaning editor who screwed up and got blocked and was long-past ready to come back but gets turned down because reviewing admins a combination didn't learn the situation well enough and were were too cautious.

#1 - #4 didn't involve tools. They just needed a finding to say "don't do that" or a nudge course correction. #5 Involved a tool-involved mistake but not "mis-use". It just needed a closer review of the situation which another admin won't do (unless it's egregious) because it's considered impolite.

Sincerely, North8000 (talk) 20:43, 13 June 2022 (UTC)

• Only #5 is within scope for XRV, because that's what the community decided. The venue is about the use of tools that need advanced permissions (including the decision not to use them). It's not the Court of Admin Admonishment.—S Marshall T/C 20:53, 13 June 2022 (UTC)
• Oy vey. If those 5 bullet points are emblematic of the types of issues that are intended to be brought to this place for discussion, then I think this place is doomed to become a more extreme drama board than ANI. Are we now condemning admins because of their personality type? Differentiating between type A and type B admins, and nice admins vs. meanies? I don't think this is a good start to this discussion. In my opinion, the first step in making AARV a useful venue is to differentiate it from all the other venues that already exist to discuss similar issues, like ANI, AN, Arbcom, or even User Talk / Article Talk pages. The two specific questions to be answered are:

1. What are the shortcomings of venues like ANI and Arbcom, specifically in terms of their ability to host a discussion about whether an admin action was improper and come to a decision about how to best respond to it?
2. How will AARV be different from those venues, such that it will address the shortcomings of the other venues and provide a substantially more effective venue for discussing these incidents and responding to them appropriately?

The "Purpose" text at the top of AARV attempts to address some aspects of these questions, but not nearly completely enough in my opinion. Someone needs to put together a concise statement for why this place needs to exist alongside venues like ANI and Arbcom. What specific problems need to be solved, and how does AARV propose to solve them? —⁠ScottyWong⁠— 21:27, 13 June 2022 (UTC)

I don't know how you could have misread my post that badly. I described some real problems, mentioned some extra hopefully-useful sidebar info, and you wrote as if I said that the problem was the sidebar items. North8000 (talk) 02:25, 14 June 2022 (UTC)

• That doesn't seem difficult to me. In the discussion that created this venue, the community was thinking about the low numbers of new sysops being promoted. The community felt that one of the obstacles to promoting sysops was a widespread concern that once promoted, sysops are hard to get rid of if they make bad decisions. So we tried to solve that by making a venue for community scrutiny of sysop decisions.

  I can think of nothing more useless than yet another drama board. This is not AN/I or Arbcom. It's for review of individual decisions about the use of the tools, not review of the totality of an editor's behaviour or character. The community has established this board but given it no specific powers, so all it can do is reach consensus to overturn an individual decision.

  We don't really have a problem with evil or corrupt sysops. Arbcom is sufficient guard against that.

  We do have a problem with inconsistent sysop decisions. Often, editors who want a particular outcome will post directly on a sysop's talk page, and if they know how Wikipedia works, they'll select the sysop with some care so as to get the outcome they want. By having a place to review actions or decisions, we can hopefully establish some norms with a view to getting more consistency of outcome.

  The models for this place are DRV and MRV. We particularly want fixed-duration discussions, partly so there isn't the AN/I-style incentive to rush to comment before someone closes, and partly so there's no accusation of someone choosing a tactical moment to close when the numbers have swung one particular way. And as with DRV we particularly want a focus on the decision rather than the person making the decision, with a complete ban on using the venue to cast aspersions, because we want to encourage and enable some reflective practice.—S Marshall T/C 22:00, 13 June 2022 (UTC)

• I'm not really sure more RfCs are the way forward here, but I agree with Floq that the core problem is that this board doesn't really have a definition of what it is aimed at. Ideally what would happen would be the community naturally would decide what type of cases to bring here, what should be closed as out of scope, what was a waste of time, etc. and it would be handled as it arises so an organic consensus on how an additional review board should operate would develop. The problem is that if such a system developed organically, it would probably look remarkably similar to AN, which is in contrast to the wishes of the proponents of creation.
  My reading of why this failed, fwiw, is that the RfC passed, the proponents of creation mistook the consensus for an additional review board to be consensus for a new bureaucratic system more structured than AN and more expansive than the community wanted, those of us who opposed it in the RfC pointed out that the issues we said would happen were happening, and then the people who were mildly supportive of it got turned off because the DRV or MRV for most admin actions that are appealed is overkill, and generally speaking the community is opposed to "rulings" on actions that have been quickly undone, which is something that was promoted here.
  Short of it: if you want something that the community would support here, it is going to look like AN. The idea of a more bureaucratic XRV doesn't have broad community support. Neither does the idea of a duplicative AN. That's why no one uses it. Blame those of us were opposed if you want, but the proponents easily had as much to do with this failing as the opponents. TonyBallioni (talk) 23:27, 13 June 2022 (UTC)
  • the core problem is that this board doesn't really have a definition of what it is aimed at but at the top of this page is the RfC proposal text and it has a very clear statement of purpose. And parameters. It's really very specific. Can we stop treating the proposal as if it were vague? Levivich 00:18, 14 June 2022 (UTC)
    • If you want a more precise problem — there's no definition of how this board is supposed to be different than any other existing process. AN exists with the same purpose stated here, and if someone ran an RfC to remove the purpose at the top of this board from AN it would fail. I've read the responses on this page as to how it is supposed to be different, but those are the vision of the most vocal supporters of this process, and I don't think those have community consensus. If they did, you wouldn't have had people just not use this. Link removal or not, it wasn't exactly a low-key page when it was created. TonyBallioni (talk) 00:41, 14 June 2022 (UTC)

      3 key differences between XRV and AN:

      • AN does not require, or even really permit, a structured format
      • There is no requirement that AN thread is closed let alone closed by an uninvolved administrators
      • AN has virtually unlimited scope. I note this last one because there are large legitimate disagreements about what the scope of this board should be, but no disagreement that it's smaller than AN.

      The real place that there is no difference are the editors most interested in participating. The crowd you get at DRV has overlap with AN, but is materially different and the crowd at MRV has even more differences. But for a variety of reasons that didn't happen here, and I would agree that some of the fault for that lays on the crowd that supported the concept but wasn't interested in participating. Best, Barkeep49 (talk) 01:17, 14 June 2022 (UTC)

      I'm as interested in participating here as I am at DRV, where I am intermittently fairly active. I *hate* AN/I and am rarely interested in participating there; even when admins tell me I should take a matter there, I usually don't. I suspect that the current class "editors most interested in participating in one of XRV and AN/I but not the other" is maybe empty if you set "most interested" high enough, but if XRV succeeds in establishing itself as a forum with an (actual or perceived) substantially higher chance of fair outcomes, that will change. — Charles Stewart (talk) 03:01, 14 June 2022 (UTC)

      @Barkeep49 you said a key difference is that XRV is structured. However looking at the archived threads, none of them have been in structured format. 2409:4071:4D86:10A1:0:0:4348:380B (talk) 08:36, 14 June 2022 (UTC)

  the proponents of creation mistook the consensus for an additional review board to be consensus for a new bureaucratic system more structured than AN: this is where we're really on different pages, Tony. The addition of structured discussion, a limited scope, and a limited range of outcomes (so in other words yeah, "bureaucracy") were the key elements of the proposal as I saw it. If those in favour of the proposal weren't expressing support for that, then... what were they supporting? And how were we supposed to know that what they said they wanted is not what they wanted? – Joe (talk) 09:16, 14 June 2022 (UTC)

  Joe Roe, I actually agree there was consensus for a structured format and that is what I think the community wanted. Sorry for the unclear wording. If you want my honest view on how to help this page succeed, it would be just to change the AE template to be applicable here so its structured, and see how people use it. That'd probably get a working page fastest.

  But my responses was about stuff like this set of principles by S Marshall which per my reply below, I think has some things in that diff are not what was proposed in the RfC and couldn't gain support if they were put to the community as a whole. I don't particularly think this was a good idea; but I'm fine with an 'implement it and see if it works' approach.' My larger point was that things like the diff I just cited are/were proponents overplaying their hand, and that is at least part of why this didn't take off. TonyBallioni (talk) 18:18, 14 June 2022 (UTC)

• I'd like to see any feedback on how this page has done what is was born to do: improve the RFA process, can anyone point to a success measure there? — xaosflux ^Talk 00:23, 14 June 2022 (UTC)

  If what you're really saying is "this shouldn't have been allowed to pass at an RfC about the RfA process" well you're not the first to say that but this is the wrong forum for that discussion. If you're actually saying you want to understand how it's improved the RfA process, I don't think there's anyone who is suggesting this board has actually been in operation. So of course it's not going to have improved the RfA process. Best, Barkeep49 (talk) 01:08, 14 June 2022 (UTC)

  @Barkeep49 This board was supposed to somehow make RFA better, and yes it has been stagnant - but this discussion is about reviving it. What I'm saying is that the revival should include success measures about how it will achieve the initial goal of improving RfA - so before we go back to another shutdown discussion its effectiveness can be judged. — xaosflux ^Talk 09:36, 14 June 2022 (UTC)

  The idea was that XRV would, in the long term, improve the accountability of existing admins and thus reduce the community's reluctance to appoint new admins. It would be great to measure that, though I can't think how and it would be unusual and unreasonable to expect specific metrics before we even try. XRV went stagnant because there was a concerted effort to smother it before it got started that resulted in all incoming links to the page being removed. That's not a measure of its success or failure and, before that happened, we had a surprisingly quick uptake. – Joe (talk) 09:47, 14 June 2022 (UTC)

  It didn't "become stangnant". An example of that happening is Wikipedia:Current events noticeboard. A decision was made that this board was not yet ready to be advertised as more development was necessary. Further development of the board definitely became stagnant after that point but this circles back to my original reply. Best, Barkeep49 (talk) 15:35, 14 June 2022 (UTC)

• I can see two open questions about XRV: when to close (should there be a fixed time and if so how long), and what (if any) format or structure for discussions (the RfC said "structured" but didn't specify a format). Are there any others? Levivich 01:23, 14 June 2022 (UTC)

  I think some people have questions about scope - see Floq's quote of himself above, to which I would add whether an administrator deciding not to do an administrative action can be reviewed (several people opposed Fram taking me here on that ground). I think there is also a question of whether multiple related actions can be reviewed in a single report or only a single action. Also from Floq, some sense of when you go to XRV vs some other forum (i.e. is it before AN/ANI or after AN/ANI but before ArbCom). I think that last one has an obvious answer but enough editors have expressed concerns I think it is worth noting. Finally, for when to close I think the nuance you need to consider is not only should there be a fixed time, and if so how long, but how to handle something that is out of scope. This latter piece is particularly contentious given the disagreements about what is in scope in the first place. Best, Barkeep49 (talk) 01:34, 14 June 2022 (UTC)

  Sorry there's really one more underlying question which I noted above, which is a more fundamental question about how should XRV be developed. Should it be developed while the board is operating or should all of the procedural details be worked out through consensus, with the likelihood of a subsequent RfC to formally determine that consensus, before the board is widely promoted? Best, Barkeep49 (talk) 01:37, 14 June 2022 (UTC)

  My take is that some of these questions are answered already by the #RFC text; some have clear answers that I think everyone would agree with based on existing policies/practices; and some are "open" questions.
  
  What is in scope? The RFC set out three criteria for what is in scope:
  1. Any action or set of related actions
  2. Requiring an advanced permission
  3. Not already covered by an existing process (e.g. DRV)
  The language in the RFC that sets out this criteria is:

  Any action, or set or related actions, requiring an advanced permission and not already covered by an existing process (e.g. WP:DRV for deletions), may be referred to XRV.

  And it's reinforced in five other places:

  • determine whether an editor's specific use of an advanced permission, including the admin tools, is consistent with policy

  • whether an action or set of actions is endorsed or not endorsed

  • whether the action should be endorsed or not endorsed

  • The purpose of XRV is solely to reach a consensus on whether the use of the permission was appropriate, not to remove permissions.

  Of course this doesn't mean there aren't still questions about what is and what is not in scope, particularly in the form of, "Is [specific example] in scope?" But I think a question like "What is in scope?" is too broad to be helpful, and instead we should look at the specific examples (such as those you've already raised).
  
  Multiple related actions? The RFC text set of related actions and set of actions means that yes, multiple related actions can be addressed in a single thread.
  
  When to use XRV vs. some other forum? I think that's answered by the criteria quoted above: it depends. If it's out of scope, don't go to XRV; e.g., if it's covered by another forum, go to the other forum. Enforcement of XRV consensus is also addressed by the RFC text. Anyone who can do it may do it, and if another forum is needed, use the other forum:

  Acting on that consensus is deferred to existing processes:

  • Individual actions that are not endorsed can be reversed by any editor or administrator;
  • Permissions granted at WP:PERM may be revoked by an administrator if XRV finds them to be misused;
  • Repeated or egregious misuse of permissions may form the basis of an WP:AN, WP:ANI, or WP:RFARB report, as appropriate.

  Basically, there is no overlap between XRV and other forums, per the RFC text. I'm having trouble imagining a scenario where someone would be genuinely confused about whether to go to XRV or another forum; an example would help clue me in.
  
  Is a decision to not do an admin action in scope? I think we all agree on the answer to this: it depends on what we mean by "deciding not to do" an action.
  "Doing nothing" is not an "action", it's the opposite. It's a "decision", but it's not an "action", and the RFC text says "action" not "decision". We couldn't review anyone for doing nothing per WP:VOLUNTEER anyway, and if we did, we'd have too many people to review. We can't be held accountable for the edits we don't make. (It's kind of nice, right?)
  However, declining a request that requires advanced perms to decline (e.g., an admin declining a PERM, RFPP, or unblock request, or a template editor declining a protected template edit request) isn't "doing nothing", that's taking an action, and would be reviewable (if it wasn't already covered by an existing process).
  Removing a CSD tag is not an action that requires an advanced permission, and thus not in scope, and it doesn't matter that some people argued about it in the past. I do not think this is a live controversy or open question.
  
  How should out-of-scope threads be handled? This is another one that I think we all agree on: threads that are unambiguously out-of-scope should be promptly closed with a brief explanation as to why they're out of scope (which of the criteria they fail, e.g. which other venue covers the reported action or set of actions).
  If a thread is arguably out of scope but not unambiguously, it should be discussed. This can be done in the thread itself, or on the talk page (I think better on the talk page, to keep the thread "structured" per the RFC text, but structure is an open question).
  If an admin "quick-closes" a thread as out-of-scope and others disagree with that, and the admin doesn't self-revert and isn't reverted by another admin, that quick-close itself would be in scope for review at XRV.
  All roads end at the same place: an admin can close a thread as out of scope but ultimately whether a specific report is in scope at XRV will be decided by consensus at XRV.
  
  Because I'm of the opinion that the above questions are already answered, I think that leaves three open questions:
  1. Should threads be opened for a fixed time and if so how long?
  2. How should threads be structured or formatted (inc. where to discuss whether a thread is in scope)?
  3. Do we need to answer some/all of these questions before the board begins operating (before the board is widely promoted, e.g. with inbound links)?
  Agree/disagree that these are the "open" questions and the others are already answered? Levivich 04:03, 14 June 2022 (UTC)

  • This summary is useful. My own opinion on those questions is: (i) I lean to a fixed minimum of 7 days, given how frequently closes before that time passed were complained about, but that's a question we are quite likely to want to revisit, (ii) We should have a preferred template, either the expediency-friendly DRV-like one page per day, or the precedent-friendly AfD-like one page per case, but we shouldn't refuse ill-formatted requests, rather encourage more experienced editors to fix submissions that don't fit, and (iii) No, I regard the page as active already, and both DRV and AfD evolved over time; there seems to be the unreasonable suggestion that if XRV is not born fully formed, like Athena, then it is a failure, but I find the idea silly. — Charles Stewart (talk) 04:44, 14 June 2022 (UTC)

• I regard this page as operational. The lack of regular cases can be taken as evidence of a lack of issues with administrative actions. This is a good thing, and not a reason to shut it down. Does someone think there is a problem with barriers to bring bona fide complaints about administrative actions?

The history of cases is messy. In my opinion, by far the biggest problem has been speedy closes. Closers have appeared to consider this page to be an offshoot of ANI, and suitable for speedy closing. Reviews should be ponderous. Reviews should not be dominated by big personalities making quick decisions. —SmokeyJoe (talk) 02:45, 14 June 2022 (UTC)

It was a mult-faceted proposal which isn't going to be perfect at the detailed level. This came out of an RFA discussion and is titled "Administrative action review" It's almost unthinkable that the intent of the respondents was to exclude review of admin actions as an admin because they didn't involve use of a tool. Clearly all administrative actions should included. North8000 (talk) 11:40, 14 June 2022 (UTC)

• XRV was an unmitigated disaster, and entirely redundant to existing processes. I knew when it was proposed that it wouldn't last, its purpose, scope, and authority were unclear at best. We don't need it, it certainly did not "fix" RFA, at all, and never would have. Let it go. Beeblebrox (talk) 20:50, 14 June 2022 (UTC)

Crystal clear scope

So to be clear, if I am reading the section above correctly, nobody still thinks this is "a process somewhere between ANI and ArbCom" (to quote the RFA RFC close), right? That part of the RFA RFC close was either incorrect, or has been superseded? It's actually a less severe option than AN/ANI? There will be no sanctions or removal of permissions issued here, only endorsing or overturning admin actions? If sanctions or removal of permissions are desired, some other process must be used? If someone tries to get someone else sanctioned or their permissions removed, either the thread will be closed, or that part of the thread will be closed (details TBD)? It is a review of actions that only the holder of the permission could make (so, for example, we're not reviewing actions taken by an admin that they could have taken as just an editor)?

If it really is true that this is just DRV/MRV for other admin actions, it removes about 1/2 of my concerns. But I don't think this is a universal opinion; I think there are people who believe (either instead of, or in addition to, the first paragraph) that this should be a place to deal with "problem" admins (or, apparently, "problem" rollbackers ). If nobody wants it to do that anymore, and everyone agrees on this lighter scope, that's significant progress towards addressing half of my concerns. But I'd want to see it confirmed below, rather than just assume.

• Subject to the other fundamental issues I still have concerns about being addressed, I don't object in principle to a noticeboard with this lighter scope. If people want to expand it to sanction or remove permissions, then I still oppose it. --Floquenbeam (talk) 12:45, 14 June 2022 (UTC)

  I am fairly sure that I've seen people commenting in the talk page archives about the purpose being a kind of pre-RFAR too, but (a) I suppose I could be wrong, or have misunderstood, and (b) (more importantly) I'm not willing to dig thru the archives to find them. If that is very clearly not the goal and not going to be the goal, then yay. --Floquenbeam (talk)

• Agree, but IMO it's main use is even lighter than that, let's call it doubly lighter. (except that "sanctions" is too broad of a word) Where a common finding is the community simply telling the admin to "don't so that" or "stop doing that". North8000 (talk) 12:54, 14 June 2022 (UTC)

  Indeed. I think Ritchie333 set the tone for this really well early on, by bringing one of his own actions here for review. It would be a huge achievement for XRV if users felt comfortable enough to come here to say, "hey, I'm not sure I did the right thing here, can I get a second opinion?" without being scared of repercussions. Because I don't think anybody in their right mind would do that at ANI right now. – Joe (talk) 14:04, 14 June 2022 (UTC)

  It doesn't happen at ANI, but people do ask for second opinions at AN and in fact there is one such thread there now. This isn't to dismiss the idea of XRV as a place where that could also happen but to be accurate about what happens now. Best, Barkeep49 (talk) 15:39, 14 June 2022 (UTC)

• For me the answer to all the questions in your first paragraph is "yes". I hadn't realised that people were interpreting that part of the RfC close as meaning that XRV was to be some sort of higher court than ANI and in that sense yeah, I don't think it's reflective of the consensus at all. I tried to be very clear in the proposal that the only outcome of an XRV is that an action is either endorsed or not endorsed (or no consensus) and that any resulting sanctions would be deferred to existing processes (with the caveat that sometimes that 'process' is just an individual admin doing something like revoking a PERM). XRV is only a venue for dealing with 'problem' admins/rollbackers/etc. in the sense that the hope is that if we address problem actions early, it will help stop such unconstructive labels emerging. I really don't think anyone thinks otherwise—if they do they're badly misreading the original proposal—and this goes a long way to explaining why we've been talking past each other, so really thanks for this Floq. – Joe (talk) 14:00, 14 June 2022 (UTC)
• My understanding has always been what you have just summarised, Floq; this is a board to basically decide on a particular action with respect to improving the encyclopaedia, avoiding anything to do with the person who made it. The problem I've seen with XRV is that it requires both a) a strong agreement and consensus that is the case and b) a general willingness to clamp down on anyone doing otherwise in the same manner as starting an AfD with "This article was created by [person I don't like] who is such a ignorant douchebag!" would. Ritchie333 ^{(talk) (cont)} 14:18, 14 June 2022 (UTC)
• The middle ground between the incidents noticeboard and arbitration (as mentioned by Joe in his support statement in the RfC) was referring to a process that isn't completely ad hoc as the incidents noticeboard but also not as formal as arbitration. As far as I can tell, all the support statements understood that the proposal wasn't intended to create a process that was a middle ground in terms of severity of sanctions that could be imposed. isaacl (talk) 15:34, 14 June 2022 (UTC)

  Oh crap, I did say that didn't I. Yes, that is what I meant: between ANI and ArbCom in their formality; not their severity. Man, this really is a lesson that if you're proposing something new you should choose your words very carefully. – Joe (talk) 08:14, 15 June 2022 (UTC)

• Maybe the "higher than ANI but below below Arbcom" concept came only from an ethereal / collective sense and not in the sense of remedies. With ANI being self review, and self-review avoids handling most of these, and Arbcom being overkill, this could be just one step above the self-review of ANI. In any event, I would not only assent to ruling out any heavy duty sanctions from here, I would strongly advocate ruling those out. The strongest thing coming out of here would be reversal of an action or a "you were wrong, stop doing that" statement. North8000 (talk) 15:58, 14 June 2022 (UTC)
• Meatballwiki has a useful page at FairProcess which is helpful to bear in mind as we develop this. My understanding of the principles is:-

  We expect most decisions to be endorsed.

  It's important that a nominator who's unfamiliar with Wikipedia should be able to use the process.

  It's important that the nominator (or person complaining) can see that we've reviewed the decision properly and fairly. They should receive a clear explanation of the decision during the discussion.

  The duration of the discussion should be seven days, in the same way that AfD is a seven-day discussion. This is specifically meant to allow the occasional snowball close and possibly, where necessary, relist, in the same way as AfD.

  Discussions should be archived in a way that's indexed and searchable. We should plan for a volume of discussions comparable to DRV or MRV, which I would see as a likely maximum.

  We should expect it to be rare that an XRV thread would lead to sanctions on the sysop or other advanced rights holder. But that could happen. I can envisage an XRV that leads to the closer revoking someone's NPP rights, for example.

  It's possible that XRV might identify a pattern of problems and refer that elsewhere. A good parallel is RHaworth's arbcom case, which was heavily attended by DRV regulars because DRV was fairly concerned about RHaworth.

  We've agreed that a decision not to use the tools is reviewable. It has to be an active decision. So if I post on %sysop's talk page to say: "Please could you semi-protect my user talk page", but %sysop doesn't respond and the thread gets archived, then there hasn't been a decision and I shouldn't open a thread here. If he does respond but his response is, "Why do you need this?" or "I'm not able to consider what you're asking, so please post that request on AN", then there hasn't been a decision so I shouldn't open a thread here. But if he says "Yes, sure," or "No, I won't", then those are reviewable decisions.

  The drama level needs to be kept as low as possible, but as this board would review blocks and other high-stress outcomes, it's likely to be a higher-drama board than DRV or MRV are.

  There's a basic expectation of decent behaviour, respect and restraint, and mocking complainants with a hilarious image caption or piece of doggerel in burma-shave format should be unacceptable here.

• I believed that we'd already decided and agreed on most or all of these things.—S Marshall T/C 17:42, 14 June 2022 (UTC)
  • I don't have a particularly strong opinion on what this should look like, but there are two points above that I think if put to an RfC would have a majority oppose or at the very least wouldn't gain consensus:
    • The duration of the discussion should be seven days, in the same way that AfD is a seven-day discussion. This is specifically meant to allow the occasional snowball close and possibly, where necessary, relist, in the same way as AfD. A 7 day discussion about an individual action, especially if the action has expired or been reversed, is outside the mainstream of community opinion. We specifically don't do that except in cases where it is about an admin reviewing consensus (i.e. DRV and MRV), because those are generally discussions about people closing discussions (DRV gets some speedies, but the norm is AfDs.)
    • We've agreed that a decision not to use the tools is reviewable. I suspect this would have a supermajority of the community oppose if put to an RfC. It would fundamentally change the relationships administrators or any user of advanced permission has with the tools, because an axiom that has gone back to basically the start of the project is that admins can never be forced to use the tools. That can't be changed by a talk page discussion.

This is what I was talking about when I said the proponents took the consensus for something it was not. TonyBallioni (talk) 18:05, 14 June 2022 (UTC)

There are two kinds of decisions not to use the tools. There are cases where I choose, for whatever reason, to do nothing and there are cases where I choose to decline a request for me to use my tools. Someone used PERM as an example above: I can choose whether or not to reply to any given PERM request, but if I decline someone's request for a PERM is that decline a use of my tools or is the only use of tools if I accept it? I agree with you that the community would not support review of a "do nothing" decision. Are you suggesting the community would also oppose this second kind of decision not to use the tools? Best, Barkeep49 (talk) 18:23, 14 June 2022 (UTC)

Actually anticipated someone would ask this after I typed it up :-)

I think formally declining something (i.e. an unblock request, RFPP request, or a perm request) is arguably an action, though historically we've just told people to re-apply/re-appeal rather than appealing the decline since its faster, and much more light weight but yeah, its something that taking someone to AN for if it was particularly abusive or involved would be reasonable.

What I don't think the community would support is someone coming to my talk page, me saying "No, I will not protect that page" or "No, I will not make you a pending changes reviewer" and taking it here rather than just going to RfPP, PERM, AIV, etc. TonyBallioni (talk) 18:30, 14 June 2022 (UTC)

Per Tony, I would absolutely oppose being able to bring someone here for not doing something, except I would be opposed to, but could grudgingly live with, "official" decisions at PERM/RFPP/etc where the admin made a decision and closed the discussion. --Floquenbeam (talk) 19:11, 14 June 2022 (UTC)

That would indeed by very silly, but I don't think we need to pre-emptively tell people not to do silly things. Looking again to DRV, technically you can bring any disputed reading of consensus there you like, but when someone complains about e.g. a no consensus close they'd prefer to be keep, the community is pretty good at telling them to go away and focus on things that matter. – Joe (talk) 08:18, 15 June 2022 (UTC)

Declining a request (e.g. an unblock request) often IS doing something if it terminates the request. Very different than "doing nothing". Sincerely, North8000 (talk) 19:32, 14 June 2022 (UTC)

There are several sorts of declines:

1. No response. These definitely should not be in scope here. If there is a pattern of evading responsibility then I think everyone agrees that should go to AN or arbcom.
2. Ask elsewhere. Cases where the person asked cannot do the required action for some reason (e.g. they lack the ability or time, they are WP:INVOLVED, etc.) or simply do not want to get involved. I don't think these should be in scope here, as the best course of action for everyone is for the question to be asked elsewhere. Very occasional exceptions may apply, but I can't think off the top of my head that would be appropriate for this venue (something like RHowarth's "I don't talk to IPs" is something that I think is better suited to AN or Arbcom).
3. I want more opinions. (see e.g. [1]) This is not so much a "no" as a "maybe" and reasonable time should be allowed for those other opinions to be given before thinking about a review. The final decision, once made, may of course be in scope.
4. Nobody can do it. This is an active decline, but as the only way that it could be overturned is if policy and/or the software is changed there is nothing anybody here could do about it (I hope we can all agree that this is not a venue to try and change policy). If the admin is wrong and it can be done then the solution is to educate them about their mistake and ask them again or ask someone else.
5. No, decline and that's final. This is an active decline and should be in scope. Thryduulf (talk) 20:11, 14 June 2022 (UTC)

In your list a decline at PERM (again just to keep using a consistent example) a 5? Best, Barkeep49 (talk) 20:17, 14 June 2022 (UTC)

@Barkeep49 Yes, unless the request is explicitly left open for others to comment or make the final decision (in which case it would be a 3). Thryduulf (talk) 20:39, 14 June 2022 (UTC)

I'm assuming you mean on the actual PERM page vs. on a user talk, etc. I'd agree with this on PERM, just not if someone is bugging me on my user talk and I say "I don't think you're qualified, so I'm not granting it" but don't do anything formal at PERM or the like. TonyBallioni (talk) 20:47, 14 June 2022 (UTC)

If someone is asking you outside of a formal process and you are the first person they've asked (as far as you are aware), then your "I don't think you're qualified, so I'm not granting it" is basically #2 on my list (they can ask somewhere else if they want). If you are the second person who they've asked, then it's strong advice to leave it there (especially if you explicitly agree with the first person) but not a prohibition on doing so. I'm not opposed to reviewing those, but I don't think they should be encouraged. If you are the third person, then they are forum shopping and your decline would be a #5. That would definitely be reviewable, but in almost all circumstances your decision is going to be resoundingly endorsed. Thryduulf (talk) 21:07, 14 June 2022 (UTC)

Key words are and that's final. In my view, the sine qua non of a "decline action" (as differentiated from a "nonaction" or "decline to act") is whether the requestor can ask someone else without running afoul of WP:FORUMSHOP. If the decline is the kind where asking someone/somewhere else would be WP:FORUMSHOPing, then it's a "decline action" and is reviewable (as in #5 on the list above). If the decline is the kind where the person asking is free to ask elsewhere (which is #1-#4 on the list above, as I read it), then it's a "decline to act" and is not reviewable. The reason for this being where to draw the line is that in a "decline action", the editor with the advanced permission is using the advanced permission to limit the actions of another editor, and thereby is taking an action that requires an advanced permission. Whereas, in a "decline to act", the editor with the advanced permission is not limiting the actions of anyone else. Levivich 20:32, 14 June 2022 (UTC)

I generally agree with this; the one thing I would like to point out though like I hinted at in response to Barkeep's initial query, the other side of the question is "If it is in scope, is it a good idea to bring it here?" For many 'decline actions' the answer will be "no." For RfPP you just submit another request at RfPP a few hours later when there's been more disruption. AIV, the same (or take it to ANI if its urgent. PERM, our example here, is a bit of an outlier, but for something like rollback it is probably a better choice for the person being declined just to wait a week an re-apply (i.e. creates less bad feelings and people are less likely to be suspicious of a 2nd request if the appeal here is declined.) For unblocks, it is quicker to file an additional unblock rather than appeal the decline of the appeal.

That's the part of the discussion that I think is missing - even if something is theoretically in scope, is it a good idea to bring it here? The answer for a lot of decline actions is going to be "no". That's a pragmatic judgement, and the answer might be "yes" sometimes, but its something we also need to keep in mind when talking about scope. TonyBallioni (talk) 20:47, 14 June 2022 (UTC)

I agree if the purpose is to get the action reversed, there are better ways than to go to XRV, e.g., as you suggest, make another request later. However, if the purpose is to determine if there was community consensus for the action in the first place -- not to get it reversed, but to find out if it was proper, if it is endorsed -- then that's what XRV is for.
I feel you're coming at this from a perspective that the only reason anyone would take anything to XRV is because they want an action to be reversed. This is belied by the very first XRV thread by Ritchie, which was a self-review. There is, indeed, a value in having a place where we can just ask, "Was this action proper? Was this action a good idea?" without asking "Should the action be reversed?" That's what XRV is for, and that's apparent in the #RFC text (which, in multiple places, notes that enforcement is deferred to existing processes), the discussion, and the close.
So I would argue that even if the action was already reversed, there may be value in discussing it at XRV anyway. Because the point is to find out what the community thinks of certain actions; it's not to reverse actions. It may be that a reversed action is nevertheless endorsed; knowing this is valuable to other advanced permission holders, who may be faced with the same dilemma in the future with regards to whether to take a specific action. The purpose of XRV isn't to directly fix mistakes, it's to determine what is and isn't a mistake in the first place, for the sake of everybody's education, so mistakes aren't repeated in the future. Levivich 21:27, 14 June 2022 (UTC)

The purpose of XRV isn't to directly fix mistakes, it's to determine what is and isn't a mistake in the first place, for the sake of everybody's education, so mistakes aren't repeated in the future.

I agree that's what you and other more vocal proponents want who are active on this page want. I don't agree that view has community support or that the RfC demonstrated a consensus for that. We don't have review forums that review moot issues. We don't have cases with binding precedents, and the desire to use this board to become that is a big part of the reason it failed so quickly. TonyBallioni (talk) 21:40, 14 June 2022 (UTC)

I've re-read the support statements once again, and it still seems clear to me that they supported the creation of a place to review administrative actions in a place other than the administrators' noticeboard or the incidents' noticeboard with the hope that it would be more effective, and multiple commenters underlined their support for reviewing the action and not the administrator. There were some comments on it being a way to provide feedback. While there was no explicit statement of preventing future problems, I think this is a natural consequence of reviewing actions. I don't feel one-off reviews will serve as binding precedents. As per English Wikipedia tradition, consensus agreements is required to establish guidance. However, just as reviews taking place today in the various venues are points of data that help illuminate consensus views, reviews on this page would contribute towards figuring out community opinion. isaacl (talk) 23:10, 14 June 2022 (UTC)

I have mixed feelings about reviewing settled issues. There's a saying that you can get a lot more done if you don't worry about who gets credit. On Wikipedia, you can get more done if you don't worry about being proven right. Reviewing an issue that has been resolved in practice can create acrimony. I also wouldn't want review on this page to be a reflexive default action. Take, as an example, a case where a mass message was requested and performed, and then someone feels that the message wasn't of sufficient interest to the targeted audience. Technically, this could be reviewed on this page, but a discussion on the mass messaging talk page would probably be more effective, both from a perspective of reaching editors with experience in the area, and in generating future guidance. An individual discussion might not in itself lead to any definitive guidance; multiple discussions though would help provide more examples to work through and gradually coalesce to a consensus. I think this would be easier to do when the conversations are held on the specific talk page for the relevant area.

There have been occasions, though, where editors who repeatedly make decisions counter to community consensus have defended their actions by narrowing focusing on whether or not their actions were specifically overturned. In this situation, it would be helpful to review the issue in order to provide timely feedback. Under the assumption, though, that this is a minority of editors, I'm not sure how to strike the right balance in figuring out what settled issues warrant a review anyway. isaacl (talk) 23:27, 14 June 2022 (UTC)

Minimum time

I'm curious how much agreement/disagreement we have amongst page watchers about this. Should threads be open for a minimum length of time, and if so, for how long, and are SNOW closes still acceptable? Levivich 18:40, 14 June 2022 (UTC)

• Well, a SNOW close is a decision to ignore the rules, so we don't need to legislate for them. IAR is policy...—S Marshall T/C 18:47, 14 June 2022 (UTC)
• It doesn't really make sense, especially for something that's already been reversed. We don't have courts and precedents here, and the natural reaction to something that is moot, even at DRV and MRV, is to close it without reaching a conclusion. Even real life courts don't deal with moot questions. If this is truly an "action review" board rather than a behavioural board, then there's no reason to continue. If there's a clear consensus that something was correct/incorrect or at the very least that a consensus isn't going to develop to overturn an action, there's not really a need for 7 days.
  The reason DRV and MRV last 7 days is because they are reviewing closes of discussions that lasted 7 days. You need to give all participants a time to comment. You don't really need 7 days to decide "oh, this removal of rollback wasn't justified" or on the flip-side "yep, he was abusing rollback, justified." It just becomes a pillory of the losing side once you keep it open too long. TonyBallioni (talk) 18:51, 14 June 2022 (UTC)

  The natural reaction to something that is moot, even at DRV and MRV, is to close it without reaching a conclusion. But not a speedy close. Calling something “moot” and boldly closing can be controversial. No harm comes from a moot case being left open a week or more. Speedy closes allow biased closes on narrow perspectives that disenfranchise editors who prefer to take their time. SmokeyJoe (talk) 22:45, 14 June 2022 (UTC)

  Little plea from the stalls here: calling something "moot" can also be controversial if it's not clear if someone's speaking AmEng or BrEng, or unaware of the difference. (Strangely similar to tabling a discussion; that one can really sour a meeting.) NebY (talk) 23:00, 14 June 2022 (UTC)

• For my part I don't really have an opinion, other than that things should be open a minimum of 24 hours (I'd like to see 48 hours) to allow global participation before anything is SNOW closed (this is a general principle applicable to all discussions IMO). I can see the benefits of keeping things open longer, just as I can see the downside of doing it. I wouldn't oppose a 7-day minimum, but I'd be fine with a 24-hour minimum, too. Levivich 18:56, 14 June 2022 (UTC)
  • Only "problem" (and I use that in quotes because its a minor one) that I have with 24 hours is that you'd have people still wanting to review moot actions or minor stuff like removing rollback that's quickly solved. If you frame it as a suggested minimum, and then give examples of situations when the minimum might not happen, I'd be good with it.​ TonyBallioni (talk) 19:13, 14 June 2022 (UTC)

    Suppose I want to contribute a statement to the review of a moot action. Why do you want to use bureaucracy to lock me out? Why do you push notions like “quickly solved”? “Quickly” is not an important feature of a review process. “Timely”, yes, “quickly”, no. SmokeyJoe (talk) 22:49, 14 June 2022 (UTC)

• AE, AN/ANI, and DRV/MRV are the most frequently mentioned semi-equivalents of AARV, and only DRV/MRV have a minimum length of time. I believe (but am not positive) that DRV will often close earlier than 7 days once a decision seems clear? Hell, for permabans of people - a much more serious discussion - the discussion "must be kept open for 72 hours except in cases where there is limited opposition and the outcome is obvious after 24 hours" (from WP:BAN). I don't object to a minimum length if it is reasonable, but 1 week is just giving any dispute unnecessary time to fester. We would be trying to give the dispute a chance to be reviewed by multiple people in various locations; we should be trying to solve problems, not trying to give every editor a chance to comment on every dispute. Personally, I think no minimum time is OK; I don't mind 24 hours; I could accept 48 hours if people honestly felt it better than 24; I would grudgingly accept 72 hours only if doing so breaks a logjam of disagreement here, and only if there were a SNOW exception similar to the WP:BAN quote above; and I would oppose any minimum time longer than 72 hours. If people are really planning to review rollbacks here () then I super-duper deal-breakingly strongly oppose any minimum time at all for those. That would be extreme overkill (it's rollback, for God's sake). I'm curious if there should be a maximum time, but (unlike fundamental issues) that is the kind of thing that could be hashed out later. --Floquenbeam (talk) 19:02, 14 June 2022 (UTC)

  DRV will often close earlier than 7 days once a decision seems clear? No. “Seems”? “Seems” is a very weak word. Try “is”. I am not sure that “often” is correct either. SmokeyJoe (talk) 22:52, 14 June 2022 (UTC)

  “1 week is just giving any dispute unnecessary time to fester”. Disagree with this. A review process centred on a complaint serves to illuminate and air an issue. This is the opposite of festering. More damage comes from requiring participants to jump in fast and hard or risk being excluded from being allowed to contribute. SmokeyJoe (talk) 22:56, 14 June 2022 (UTC)

• One of the big problems with AN/I is the way AN/Is get closed before you get time to have your say. At AN/I decisions are made by the people who're on Wikipedia the most: it works to the benefit of the prolific Wikipedians who're here for hours a day and for whom a three-day discussion seems like a needlessly long one. Seven-day discussions would be an opportunity to fix that problem and allow the weekend-only people a chance to get heard. Honestly, who's meant to be harmed by letting a review of their choices run for seven days?—S Marshall T/C 19:16, 14 June 2022 (UTC)
  • I disagree with the focus on giving particular people a chance to get heard. The focus should be on giving the issue a chance to be reviewed by a cross-section of people. Once that has happened, any further delay doesn't contribute to actually solving the problem. And the harm in a 7 day review is that, for admin actions like blocking, once a decision is clear, keeping it open only allows bad feelings to fester. And for "admin" actions like rollback, it about an order of magnitude more time than the problem deserves. Even though participation is optional, there is still a cost, in time, work, and happiness, in dragging out a dispute. --Floquenbeam (talk) 19:22, 14 June 2022 (UTC)

    I agree minimum time limits can cause problems. Especially if there is no requirement to informally discuss things before bringing it here you will have situations where the admin concerned reads the report, says "ah yes, I made a mistake" and immediately reverses their decision - there is no benefit to anybody in keeping these reports open any longer. So I cannot support any minimum period that doesn't have allowances for exceptions. Thryduulf (talk) 19:36, 14 June 2022 (UTC)

    “requirement to informally discuss things before bringing it here”. I strongly support this as one of very few conditions for opening an XRV case. There must be at least a thread addressing the issue and evidence of at least two editors in considered disagreement. This is strongly encouraged at DRV and is almost mandatory at MRV. XRV should not be a place to first mention a disagreement, or it will be filled with trivially resolved issues. SmokeyJoe (talk) 23:02, 14 June 2022 (UTC)

    Would we want to keep it open long enough to give the administrator concerned time to respond? If so, what would be reasonable without stringing things out? Maybe likewise for the editor affected, if someone else starts the thread? NebY (talk) 20:14, 14 June 2022 (UTC)

    Boy, I love this idea. As a general rule, I've always felt that whenever someone's edit is reported to any noticeboard, the person who made the edit should have a chance to respond to the filer's statement before everyone else jumps in with an opinion. How could any uninvolved editor evaluate whether an action is proper if they don't even hear the actor's justification for the action? I'd support making that part of the structure (see next thread below): no uninvolved commentary until the person who made the action has a chance to give their explanation. I'm not sure anyone else agrees with that, though :-) One obvious concern is that a "bad actor" could stall review by refusing to comment. This could, in turn, be addressed by saying something like... no comments from uninvolved editors in the first 24 hrs until after the editor whose action is under review has a chance to comment; after 24 hrs, if the editor hasn't commented, we allow uninvolved commentary. That, however, seems like cumbersome rule creep. I'd love to see it as an unwritten cultural practice, though... everywhere, at XRV, ANI, DRV (the closer should have a chance to explain the close before others review it), MR, etc. Levivich 20:26, 14 June 2022 (UTC)

  Wow. So many ways to stop people discussing bad calls.  :(—S Marshall T/C 20:52, 14 June 2022 (UTC)

  The purpose of XRV, of course, isn't to discuss bad calls, which would be a pointless endeavor. The purpose is to determine whether a call was a good call or a bad call (endorsed or not endorsed are the two options, per the RFC text). It's difficult to imagine how editors could intelligently discuss whether a call was good or bad if they haven't yet heard from the person who made the call. In fact, there exists, to my knowledge, no investigation or dispute resolution procedure in the entire world that doesn't involve an opportunity for all parties to state their views before the decision-makers make a decision. Investigators don't conclude their investigation without first attempting to talk to the person they're investigating. Judges and arbitrators don't make decisions -- or even offer opinions on the merits -- without hearing from both sides first. There's a reason the whole world operates this way, and it's because it's a universal principle of fairness. In fact, it's one of the principles in the FairProcess meatballwiki article you linked to ("Involve individuals in the decisions that involve them. Get their input..."). In my view, anyone !voting "endorse" or "not endorse" without first reading the explanation for why the action was taken would be seriously jumping the gun. And, again, this is applicable not just to XRV but to any investigative or dispute resolution procedure. And of course a delay in allowing uninvolved editors to respond is not the same thing as stopping anyone from discussing anything. Levivich 21:34, 14 June 2022 (UTC)

  If it's at XRV then at least one person thinks it could be a bad call. I don't want people getting their balls busted for talking in the wrong order, particularly if they've got something to say and the discussion is going to get closed after an unknown interval.—S Marshall T/C 21:46, 14 June 2022 (UTC)

  @Levivich Something like the person who made the action should generally have the opportunity to respond before others comment. If, after they are made formally aware, they choose not to respond at all others can leave comments after 24 hours. If they make edits elsewhere on the project then others can comment 1 hour after those edits. However, that should not apply to reports that are clearly made in bad faith, trolling, where the reporter has clearly misunderstood something, or where the reportee has unquestionably made a mistake. Putting the two together I think the best thing would be for comments by others to be generally discouraged unless the situation is obvious and/or egregious. Thryduulf (talk) 20:58, 14 June 2022 (UTC)

  Agree, that makes sense to me. Levivich 21:36, 14 June 2022 (UTC)

  Well! I was thinking more that a thread shouldn't be closed for some minimum time that would give the admin a chance to respond, but this is all much more interesting and could lead to calmer more measured discussions. Yes, waiting would preclude rapid corrective action, but making it clear that it's the wrong venue for that could help calm the board too.. NebY (talk) 21:45, 14 June 2022 (UTC)

It's only been a day, but so far I see:

• S Marshall and SmokeyJoe in favor of a 7-day minimum
• Tony, Floq, and Thryd are in favor of no minimum or a short minimum like 24hrs
• NebY in favor of at least long enough for an admin to respond
• I'm kind of ambivalent; I think there ought to be a minimum, maybe 24-48 hrs, but would support trying any minimum
• Everybody (I think) agrees no matter what the minimum, there should still be IAR/SNOW exceptions for obviously bad-faith, out-of-scope, etc., threads

Is this a fair summary? If not, please set me straight. If so, this is a bit of a split, at least numerically. How do we want to resolve this? Do we want to just try one of these options for now and revisit it later (my preference)? Or put it to a talk page !vote? Include it in some future RFC? Levivich 20:15, 15 June 2022 (UTC)

Discussion format

I'm curious how much agreement/disagreement we have amongst page watchers about this. Should one or more specific discussion formats be required, or recommended, and if so, what should the format be? Levivich 18:46, 14 June 2022 (UTC)

• I think there ought to be one hard requirement, which is that whomever posts a new thread seeking review of an action needs to include a link to a diff or log entry of the action under review and identify which advanced permission that action requires (or set of related actions). This will allow us to easily determine if a review is in scope, and also to help remind participants of the purpose of the review (to review a specific action or set of related actions). Beyond that, I would let the person seeking the review set the format of the review, and then we can try out some different formats and see which work best. My guess is that different formats will be better for different kinds of reviews--there are a lot of very different advanced permissions, after all. Examples of possible formats include separate survey/discussion sections (like RFCs), one-section-per-editor (like arbcom), or no format (like ANI). I can see an argument for not allowing the last one per the RFC, which would be fine with me. We could create two or three format templates and allow filers to choose from among them. Levivich 18:46, 14 June 2022 (UTC)

• Serious suggestion: if you want this to succeed, adapt the AE template, get rid of the extraneous rules that we're debating here on the talk page outside of a larger community discussion (such as time and scope), and tell people to use it to bring cases here. The community will naturally decide scope and time limits on their own by the closes. If you have a relatively small group of people dictating the specifics in advance, the result is what happened when this launched. There was consensus for a structured discussion format. Create a template for that, use it, and see if it works. Most likely way to get this off the ground, imo (and I have no strong ties to the AE format, just think it'd be easiest to try/adapt since its pre-built for you.) TonyBallioni (talk) 18:55, 14 June 2022 (UTC)

  I dislike, strongly, the idea of using a behavioral template as the basis for this page. This is not a behavioral forum. We don't have to use the current template but would suggest something closer to the MRV or DRV template be used. Best, Barkeep49 (talk) 18:59, 14 June 2022 (UTC)

  I don't have a particularly strong attachment to the AE template; but I don't really see the difference between DRV/MRV/XRV templates and AN. If people want a structured format, the only discussion venues I'm aware of with structured formats are SPI, AE, ARCA, and ARC if people want to use those as models. Everything else is freeflowing and the current template is no different than AN from what it looks like in the archives. I guess I'm saying the one thing I think everyone agrees there was consensus for (a structured format) was the one thing that wasn't implemented. TonyBallioni (talk) 19:10, 14 June 2022 (UTC)

  I see DRV etc. as 'structured' because they use dated discussion subpages, expect participants to offer a bolded !vote, and are always closed on a fixed schedule. And when I wrote the proposal, I specifically had something like DRV in mind. (I realise that doesn't mean that others who supported the idea thought the same way). Personally I find AE and arb formats in general to be a bit too structured and bloated, but I'm not opposed to it in principle. Maybe we should do a quick poll to see which is most preferred? – Joe (talk) 08:02, 15 June 2022 (UTC)

  I'll note that people do AE/Arb formats wrong and it requires active clerking, where as I see people get the DRV format "wrong" far less often. Best, Barkeep49 (talk) 13:22, 15 June 2022 (UTC)

  One page per day/week/month doesn't make a venue structured, it just makes it harder to watchlist. For venues with a lot of traffic like DRV it's a necessary evil, for something with the volume we anticipate here it's not necessary. Thryduulf (talk) 14:44, 15 June 2022 (UTC)

  You could say the lack of watchlist-ability is a good thing. Part of the problem with ANI is surely that as soon as something is posted there, every drama hound on the project sees it. I like that AfD and DRV are a bit harder to track, which encourages a slower and less noisy pace. But probably a lot of that is personal preference. And yeah you're right it doesn't make much sense to have dated subpages unless there's high volume. The kind of structure you and Levivich hashed out below works for me. – Joe (talk) 18:10, 15 June 2022 (UTC)

  I mean... that's what we did originally? With DRV templates instead AE, but still. Then we were criticised for not working out all the specifics in advance. – Joe (talk) 07:57, 15 June 2022 (UTC)

• Each type of structured format has advantages and disadvantages. I'm not sure I have a strong preference, but I disagree people should be able to use any format they want. A format should be chosen before this starts, and then if it isn't working you can use trial and error/discussion/whatever to change it. I think it would be a big mistake to mix and match threaded discussion on some issues with sectioned discussion on others. --Floquenbeam (talk) 19:16, 14 June 2022 (UTC)

  I strongly agree that there should be a single structure, possibly with some minor alterations depending on what is being reviewed. Anything else is just going to lead to confusion and disorder. I don't have a strong preference about what that structure should be, but I do think there should be some hard requirements - the request should include all of the following:

  • A clear statement of what action is being appealed, ideally including a diff and/or log action.
    • If there are multiple aspects, which one(s) are relevant
  • Who performed the action (in most cases this will be obvious from the diff/log entry but there might be exceptions, and it also serves as a check to make sure wrong trees are not being barked up)
  • A clear statement of why it is being appealed - why do you think the decision was wrong? What do you think the correct decision should have been and why?
  • Either a link to prior discussion with the person who performed the action, or an explanation of why you haven't had a prior discussion
  • Any background necessary to understand the situation.

  This should, I hope, provide the right balance of not to onerous or bureaucratic but providing structure and making sure that everybody is on the same page. One of the issues with AN(I) is that misunderstandings can quickly snowball. Thryduulf (talk) 20:36, 14 June 2022 (UTC)

  I would support this kind of format with a few minor tweaks. I'd suggest four required sections for a new filing:
  1. A clear statement of what action is being appealed, whenever possible including a diff and/or log action (and if there are multiple aspects, which one(s) are relevant)
  2. Who performed the action, and confirmation that the person has been notified of the report (like we do at AE)
  3. Either a link to prior discussion with the person who performed the action, or an explanation of why you haven't had a prior discussion
  4. A clear statement of why it is being appealed - why do you think the decision was wrong? What do you think the correct decision should have been and why? - including any background necessary to understand the situation Levivich 21:39, 14 June 2022 (UTC)

  Yes, that's good phrasing. Thryduulf (talk) 00:30, 15 June 2022 (UTC)

  Good phrasing, but I think it would be improved at #4 with a requirement for the initiator/complainer to state their “desired outcome”. This must not be read as limiting the outcome to yes/no, as the discussion may find consensus for another outcome.

  There may be a difference between “what the decision should have been” and “what should be done now”. SmokeyJoe (talk) 00:45, 15 June 2022 (UTC)

  Much of the time there can't be an outcome because we'll be reviewing a block that's expired before the XFD closes, or a time-limited page protection decision that's no longer in force by the time the paperwork's been done. In such cases the point of XRV is to endorse or to say what we think should have happened instead -- it becomes a place for learning lessons and reflecting on how we could do things better next time.—S Marshall T/C 12:16, 15 June 2022 (UTC)

  Your comment made me think of Morbidity and mortality conferences, which are about improving future performance without assigning blame. - Donald Albury 14:42, 15 June 2022 (UTC)

  I think i"ve linked to that at some point as well Donald and agree that would be the best case version of this board. Best, Barkeep49 (talk) 15:49, 15 June 2022 (UTC)

  There's an unresolved quibble about the language we use in that case -- "Endorse" or, err, "Not endorse"? Technically I think the antonym of endorse might be "deplore", but I'm hoping to be able to improve on that word.—S Marshall T/C 16:43, 15 June 2022 (UTC)

  I disagree about stating the desired outcome. #RFC text was clear about endorsed or not endorsed, and that's it. No other outcomes. Nothing else to be discussed except whether a specific action is or is not endorsed. Levivich 19:07, 15 June 2022 (UTC)

Formal RFC

Should Wikipedia:Administrative action review be revived, or should it be closed and marked historical? 21:40, 14 June 2022 (UTC)

Option 2: Closed and marked historical

• Support I did not support the creation of this process, which I consider entirely redundant, and how it was supposed to "fix RFA" was magical thinking. However, it was approved at the prior RFC, so I figured we were stuck with it and participated in discussions aimed at clarifying how it would work. While those discussions were underway, the board was in active use, without any clear rules, scope, or authority. I walked away as it seemed to be becoming a total train wreck, and when I checked back on it later I found it was basically closed as a failure. At that point I think it is fair to say that consensus changed and there was no longer broad-based support for having such a process, yet we have conversations ongoing on this page that seem to assume that all that is needed is a few tweaks for clarity of purpose. I do not think that is sufficient, we need a clear consensus that the community still wants this process if it is to go live again. I still have the same objections I did previously, now backed up by the fact that it did indeed fail badly when attempted before. We simply do not need another drama board, this process can't do anything beyond what ANI already does on a daily basis. Beeblebrox (talk) 18:50, 15 June 2022 (UTC)

Option 3: Procedural close

• Procedural close - we literally just had this at Wikipedia:Village pump (proposals)#RFC: Shut down Wikipedia:Administrative action review, which I closed because it was an improper RFC, having a non-neutral RFC statement, and having no WP:RFCBEFORE, was reopened, and Tony re-closed it. This RFC also is not neutral (not even factually accurate in its retelling of history), and also has no RFCBEFORE. In fact, on this page, right now, are editors who are discussing what may become an RFCBEFORE, as a result of the last RFC being closed. Launching this RFC is WP:FORUMSHOPPING, WP:DISRUPTIVE and WP:POINTY. As Barkeep says below, posting this now is extremely poor form. Frankly, if an editor who wasn't on arbcom had done this, I'd be thinking about starting an ANI thread about it. Levivich 21:54, 14 June 2022 (UTC)

  I think ANI would be a silly choice no matter who opened it. That said, being an arb conveys no formal special privilege in this kind of discussion. Obviously arbs have social connections that could make an ANI tricky but that's part and parcel of why I don't think it would be a good choice no matter who opened it (including some other IP). Best, Barkeep49 (talk) 22:00, 14 June 2022 (UTC)

  When it comes to disruptive behavior, I have higher standards for functionaries than regular editors. Levivich 22:02, 14 June 2022 (UTC)

• The one strong consensus in the Village Pump proposal was that an RFC about shutting down this review was not the right thing at this time. Given that there are active, potentially productive, discussions ongoing at the moment means this consensus is likely even stronger than it was. Thryduulf (talk) 00:23, 15 June 2022 (UTC)
• This RFC is unfortunately timed. People clearly don't want endless RFC's, and I still believe that when/if this is all set up, there should be an RFC confirming that what we have is what the community wanted/wants before it goes live, before links to the process are added back to all the places they were removed from. I suspect people who think such an approval RFC is not necessary might have grudgingly gone along with that, just to get broader buy-in, but probably not if we've just gone thru another RFC to close it. Since there is active discussion on several aspects, and since the previous RFC - malformed as it was - seemed to demonstrate a lack of community desire to shut this down, I'd prefer this be procedurally closed, and an "approval" RFC be held instead when a critical mass of people think it is ready for prime time. I guess that's option 3. --Floquenbeam (talk) 00:45, 15 June 2022 (UTC)
• Close RFC Long story short, the RFC should not exist at this time.North8000 (talk) 16:50, 15 June 2022 (UTC)
• Obviously I don't agree. We have discussion here of how to restart this thing, without having a recent consensus that it should exist at all. The previous RFC cannot be leaned on as proof that the community actually wants this after what happened last time it was active. If the community does not want this, defining how it works is a waste of time. That a prior malformed RFC was closed early doesn't change any of that. RFCBEFORE is not a policy to be used to stifle discussion, it is just advice. Beeblebrox (talk) 18:42, 15 June 2022 (UTC)
  • Hmm let's see, the previous RfC was... last October. A little under eight months ago. How long ago was your RfA? Will you be handing in your bits for lack of "recent consensus"? – Joe (talk) 18:49, 15 June 2022 (UTC)

    Last October was before the whole thing was tried and failed. I think your arguments are the ones getting increasingly absurd. You seem kinda desperate to not have this discussion. Beeblebrox (talk) 18:54, 15 June 2022 (UTC)

• I stopped engaging actively in the process, because what was deliberately intended to be a lightweight process suddenly got bogged down in people wanting a vast amount of rules entirely laid out in advance rather than allowed to evolve from the original basics like most of your pages. Multiple individuals also carried out a degree of goalpost moving which discouraged me from staying engaged. The verbose disputes on the talk page likely also dissuaded people from bringing cases to XRV, which is not especially surprising. That all said, that wouldn't make this a procedural close !vote. That's because having this RfC while a discussion is ongoing seems odd, and while it's obviously nowhere as problematic as the recently closed RfC, it doesn't avoid all the issues raised therein. Nosebagbear (talk) 19:07, 15 June 2022 (UTC)

General discussion

• This process was approved as part of a multi-part RFC on various ideas to reform WP:RFA. Rules, scope, and authority of the board were never clearly established, and it quickly fell into disuse. An attempt to mark it historical was reverted, and led to discussion to revive the board instead. While there was a consensus to start this board last year, the rapid failure and abandonment of it suggest consensus may have changed on that point, the purpose of this RFC is to establish if there is still a consensus to have this process at all. Beeblebrox (talk) 21:40, 14 June 2022 (UTC)
• Beeblebrox edited your headers for clarity. Feel free to revert. TonyBallioni (talk) 21:43, 14 June 2022 (UTC)

  I'm not sure what they were but I'm having the wording match the rfc question. Best, Barkeep49 (talk) 21:45, 14 June 2022 (UTC)

• Beeblebrox I think swooping in out of nowhere to start this is poor form. There is, above, useful discussion which you could have joined. Alternatively you could have reverted Tony's admittedly self-involved close if you wanted to establish consensus or not. If this RfC passes it actually doesn't change the need for the discussion above but it seems likely to be derailed by the presence of this RfC. Best, Barkeep49 (talk) 21:49, 14 June 2022 (UTC)
• I thought arbs were elected for their judgment and discretion?—S Marshall T/C 22:10, 14 June 2022 (UTC)
• The process was put on pause by request of various editors who disagreed with it. In the collaborative spirit, links pointing to the page were removed in accordance with the request, to enable more refinement and discussion to take place. This does not indicate that the consensus generated by the original RfC has changed. isaacl (talk) 23:34, 14 June 2022 (UTC)
• ...seriously Beeblebrox, why now? We just had an attempt to start an identical RfC that went down like a lead balloon and the discussions above are by far the most constructive we've had since the MfD nomination. – Joe (talk) 07:54, 15 June 2022 (UTC)

  Why now is pretty simple, because it seems like some folks are acting like there is consensus to re-start this thing,based on the prior RFC. I don't think we can count on that still being the case given what happened when this was last active. Beeblebrox (talk) 18:35, 15 June 2022 (UTC)

  This is getting absurd. Do you organise a household RfC before you take a shit in the morning? We don't need a second (third? fourth?) consensus to try and restart XRV any more than we needed a second consensus to start it in the first place. There was never any consensus to stop it. – Joe (talk) 18:45, 15 June 2022 (UTC)

  I obviously don't think they violate any rules but I will ask Joe if you think the first two sentences add more heat or more light? Barkeep49 (talk) 18:49, 15 June 2022 (UTC)

  Barkeep49, I dunno man, maybe we should have an RfC about it. – Joe (talk) 18:51, 15 June 2022 (UTC)

  Hard to get a consensus to stop it when RFCs on the topic are shouted down. ScottishFinnishRadish (talk) 18:50, 15 June 2022 (UTC)

  @Beeblebrox you might want to at least support option 2 then because right now there are 4 editors formally opposing the existence of this RfC (plus another 3 who have made comments here in GD along those lines) and none who have formally weighed in on the merits. Best, Barkeep49 (talk) 18:47, 15 June 2022 (UTC)

  Yeah, I had to rush off to an appointment yesterday before I had time to do that, I've done so now. Beeblebrox (talk) 18:52, 15 June 2022 (UTC)

  This RFC was started because we're having the most productive conversations here in months. Beebs is trying to stop that. Thankfully, this RFC can be safely ignored until it's snow closed, and we don't need to wait for that to happen in order to proceed with the productive conversations. Levivich 19:03, 15 June 2022 (UTC)

There was a consensus (in a widely advertised RFC) to start it. Viewing that we need a consensus to continue on that is simply not correct. North8000 (talk) 19:02, 15 June 2022 (UTC)

Was being the operative word. There was a consensus to start it, nobody is debating that point. It was started, it was a hot mess, and it was stopped. It is reasonable to ask if at that point, consensus had changed. Beeblebrox (talk) 19:39, 15 June 2022 (UTC)

I see the current discussions as a way to provide a concrete thing that people can make an informed choice to support or oppose, rather than a nebulous cloud of multiple competing ideas of what AARV sort of is or might be or should be. I say that this is the wrong time for an RfC because we haven't currently got a process to stop, we are having discussions about what form a potential process might take. You don't have an RFC to determine whether people should be allowed to continue having discussions about developing ideas that might or might not lead to something. The time for an RfC was either when the old process was active or when there is a concrete proposal for a new process. Thryduulf (talk) 19:56, 15 June 2022 (UTC)

There was a discussion amongst a much smaller number of people than at the original RfC to pause the process. There hasn't been any consensus to permanently stop working on the process. isaacl (talk) 20:03, 15 June 2022 (UTC)

Yeah. I think later I'm gonna write a little history with diffs so we can all remind ourselves about what actually happened, which is quite different than what Beebs is recalling. Levivich 20:06, 15 June 2022 (UTC)

I'm going to use a slightly wiki-incorrect word to make a point. That's like saying that a supermajority decided that it should exist and then any person can come along and say that it needs a second supermajority to continue exist or else it disappears. North8000 (talk) 20:20, 15 June 2022 (UTC)

I'm not trying to put a number on it, and I certainly won't be the one closing the discussion, I just think the discussion should actually be had before trying to reboot this thing. We tried this, it didn't work out, and we've been fine without it for several months, admin actions are being reviewed at ANI and appropriate actions taken, so I don't see why we need a redundant process to that when it already failed once. I could easily be wrong about that, it sure wouldn't be the first time, it may end up that a majority of users fully support restarting this, I just don't think it should be a small group of supporters that makes that decision. And, Joe, I don't know why we need to discuss this by email, so I'll answer your question here: yes, I do think this is harmful, for the simple reason that we do not need another drama board, an we certainly don't need a board with no teeth that can't even do what ANI already does. I'm sure the supporters of this are all here in good faith and we simply don't agree on this point, but I think you may be wasting your time trying to restart this. Beeblebrox (talk) 20:38, 15 June 2022 (UTC)

I think there's no outcome of this RfC that improves on what was happening before it and it is this RfC that is wasting time. Let's say that Option 1 were to gain consensus. The board would immediately restart. Is that a better outcome than people coming to a consensus that addresses some of the concerns that were raised before? (I would suggest) Let's say that Option 2 were to gain consensus, would that consensus evaporate if the board were put forward in a slightly more refined format? (I would suggest not, though it would save the time of other volunteers that seems to be a concern) Let's say that neither option gets consensus right now, what is the status quo to default to? (I would suggest it would turn into a clusterfuck and probably the largest waste of editor time of the three potential outcomes) You had a superior option to ensure that there was consensus before the board was, to use your word, rebooted. That was for you to find a place to say above that you agree with Floq and Thryduulf that an RfC was necessary before the board is re-linked and then not spending much, or any, time doing the work of addressing concerns. You could have had your chance to say that this board would go away without disrupting the consensus building that was (and fortunately still is) occurring. Best, Barkeep49 (talk) 21:07, 15 June 2022 (UTC)