Wikipedia talk:Administrator recall/Archive 3

"RfA is not a vote" "ArbCom is not a vote"

ArbCom results are based on the amount of support each principle gets: 8-1, 9-0, etc. That's a vote according to the dictionary. The debate leading up may not be, but the end result... its number for, number against. Where/how is this not a vote? Please illuminate it for me.

RfA results are based on a concensus based on the number of responses for, against, or abstain. Comments and commentary from reading them seem to indicate that the purpose of said commentary is to sway votes. Where/how is this not a vote? Please illuminate it for me. Also, please give me an example of an actual already run RfA that is not a vote, and why it's not a vote.

I am suspecting that the whole "They are not votes" comments here are simply because people don't/won't acknowledge that in ultimate function they are votes, but I will be happy to be proven wrong with examples and a good explanation of my apparent wrong-headedness. rootology (T) 20:38, 15 August 2006 (UTC)


RfA isn't supposed to be a vote, but a discussion with Beauracrats having leeway to evaluate consensus amongst respondants. Like deletion debates, individual "votes" are evaluated on their contribution. Someone with few edits, who merely posts "*Confirm" or "*Oppose" would not carry the weight of a more established editor willing to explain their reasoning. The current reality of RfA may or may not approach this idea.
ArbCom is "not a vote" for similar reasons. The arbcom is presumed to deliberate each point, in private, before casting a ballot. It's sort of like the Supreme Court, I supose. I think the ArbCom currently lives up to their idea more closely than RfA, simply because it's such a small (fixed) group that does have private avenues to discuss cases without getting lost in massive discussion pages on Wiki. Again, it's open to interpretation. --InkSplotch 20:50, 15 August 2006 (UTC)
While there is of necessity a resulting, binding decision that is a vote and is functionally similar, nevertheless discussion precedes that conclusion, and "votes" can be changed and are supported by reasoning. RfA is not simply placing a "Yay" or "Nay" in an unchangeable ballot box. In the case of RfA, a better analogy would be an entire political campaign, including the conversations and considerations that lead up to the election; regardless of the "vote", the non-citizen is not elected and the incompetent is vetted out. In the case of ArbCom, a better analogy would be an entire legal trial, including the courtroom evidence and procedures that lead up to the jury decision; the jury is carefully selected and the result of the decision is based on sufficient evidence and reasoning that proceeds from policy.
Neither of these processes are anything like a CNN online poll or whatever else you may think of as an example of bad voting. These processes admit the specific expression of an opinion, but that opinion is informed by reasonable consideration. It may be erroneous, it may not meet the ideal, but that does not mean labelling it a "vote" is a meaningful explanation of flaws.
Note also that nearly all ArbCom decisions are unanimous: they are applying a clear standard to specific actions in full agreement; in such cases, naming it a "vote" is absolutely meaningless. Similarly, RfAs require a strong supermajority—for every 1 person who opposes, there must be at least 3 who support—and most are nearly unanimous, above 95%. The greater the majority, the less like a typical vote it is; one person has less and less power as a supporting voter whereas an opposing voter has 3 times more power in preventing the conferment. —Centrxtalk • 21:20, 15 August 2006 (UTC)

Thank you. How then is a reverse RfA, in which people can change their points/opinions/"votes", inherently flawed compared to a standard RfA? The only numbers based thing for the Recall RfA itself was that consensus should be at least 80%. Is the only disagreement with that side of it then the "binding" wording? If so, easily amenable to be closed by beurocrat review as a normal RfA is, with the extra emphasis that a desysop should not happen UNLESS the support to do so is at LEAST 80%. rootology (T) 21:28, 15 August 2006 (UTC)

Well, right now many feel RfA is flawed because even with the 3-1 supermajority, it can still be flooded with popular votes. If true (and I'd agree that it is), than a reverse RfA would be vulnerable to the same thing. And on that basis, the more wide-reaching an admin's action might be, the more likely an unpopular vote might remove them for it, even if the action was right. The pedophilia crises would be an excellent example of how many admins on both sides of the issue might have been desyssoped in the heat of things. Even Jimbo probably would not have been safe. --InkSplotch 21:41, 15 August 2006 (UTC)
I can absolutely guarantee that ArbCom deliberations are not mere "votes". Certainly, I think I'm in more of a position to know than anyone. :-)
OTOH, yes, RfA is an absolute hell-hole, and has been for a long time. Why do you think that most of us don't vote there anymore?
James F. (talk) 22:22, 15 August 2006 (UTC)
So decisions are not made based on the number of support/agree votes each proposed decision gets? A proposed decision could have less than majority support and still be implemented? Karwynn (talk) 22:28, 15 August 2006 (UTC)
90% of the ArbCom work is done in private where it is only seen by current and former ArbCom members. The "voting" on the on-wiki pages is basically just a formality that informs us of their decision. --Cyde Weys 02:06, 16 August 2006 (UTC)
You can get a feel for the negotiation that goes on if you look at edits to Wikipedia:Requests for arbitration/Irishpunktom/Proposed decision. The case could not be closed for days because some arbitrators dug in their heels about the remedy directed against an administrator. With the help of more evidence, they eventually won a rethink from one of the other arbitrators, who changed his vote on the desysopping proposal. As a result, that administrator will lose his sysop bit if the case is now closed in its current state (which seems likely).
Also worthy of note is the fact that the arbitration case was not brought against that administrator. He was an involved party whose conduct as an administrator was found lacking. --Tony Sidaway 02:22, 16 August 2006 (UTC)
I realize there is discussion on ArbCom cases, just like there is on RfA's and such. BUt in the end it comes down to who supports a decision and who doesn't, right? Is a proposed decision with a 2/1 opose/support vote ratio going to come into effect? Karwynn (talk) 17:06, 16 August 2006 (UTC)

Desysoppng trend?

Hmm. Desysopping is being proposed more and more often. Maybe this is an indication of a greater willingness to do so? This policy may not be needed after all. Karwynn (talk) 22:36, 15 August 2006 (UTC)

Perhaps the solution is to turn Wikipedia:Admin recall into a simply explained layman's guide to filing an arbitration case against an admin which one feels has violated policies and deserves to be desysopped, and allow direct consideration of such cases by the Arbitration Committee. Yes, it is rare that admins lose their bit but it is commensurately rare that admins commit egregious abuses of their buttons. Should there come a case in which the ArbCom refuses to desysop a truly abusive admin, then perhaps community consensus will support some form of more direct oversight. But until that time, I feel that this is a solution looking for a problem. FCYTravis 02:39, 16 August 2006 (UTC)
Not really, it has a problem(s), stated in the "SOme principles" heading. THe problem may not be fact, but there still is a perceived one, and this is not "looking" for one. Karwynn (talk) 16:39, 16 August 2006 (UTC)

Moving forward

There is consensus here that:

  • The proposal as it stands is not acceptable to the current talk-page participants, and
  • That it will fail any wider audience as well.

I see also:

  • There has been some cautious approval of the DocMkII version in the grey box above, and
  • A large amount of this discussion is repetative and/or relates to the "31/10/2000" version of the proposal

I'm going to replace the main page with this version very soon if there is not substantial screaming. Then I'd like to (again) archive this massive talk page, creating a summary/FAQ at the talk so that we can move on.

brenneman {L} 03:57, 16 August 2006 (UTC)

I like DocMkII, but it *might* be worth switching to Wikipedia:Administrator recall or something in order to preserve the history of this page as a rejected format (going by comments and the straw poll). There's no harm in noting that it grew directly out of this proposal, though, as a show of good faith in the original proposal. On the other hand, I could be looking too hard at this. I just don't want people to ignore it out of a misunderstanding for what drastic changes have/will have occurred when said replacement takes place. -- nae'blis 05:12, 16 August 2006 (UTC)
*sound of crickets* Hmm... it's quiet. *wind blows, tin can rolls down empty street* Too quiet. Why do I feel sometimes like people prefer to argue, and when a proposal is semi-sound they clam up? ^_^ One reason for not creating a new proposal is that "proposal fatigue" sets in. People say things like "Well, the last one failed after two days, so junk this one as well." Instead if we stick in one spot we've got some continuity to show that this was a thoughtful process. - Aaron Brenneman 06:58, 16 August 2006 (UTC)
The gray box? It's basically arbitration with an intermediate step. Why bother with the intermediate step? The Arbitration Committee is perfectly capable of identifying and desysopping poor administrators (it's apparently about to desysop someone right now). --Tony Sidaway 10:14, 16 August 2006 (UTC)
Tony, I think the problem most objectors have with ArbCom being the primary de-admin process is that it indirectly encourages cabalism and a disregard for the community. Admins well-liked by the ArbCom stand very little chance of ever being removed - even if they have lost the trust of nearly the entire community. This is not meant to suggest 'corruption', but merely human nature - complaints against our friends always seem less valid than those agaisnt people we do not know or dislike. That creates an imbalance which leads some admins to be more hostile to the community than they would be if they were still answerable to/part of it. Clearly it is inherently damaging to Wikipedia to have the community at war with administrators, and thus I believe we should work to restore a closer connection between the two. See my suggestion below, which (in parallel to the above) you might term 'basically request for adminship with an intermediate step' (intermediate step to weed out the frivolous). --CBD 11:07, 16 August 2006 (UTC)
I think 'DocMkII' is better, but still too 'processy' for my taste. I'd prefer something like;
  1. One or more users submit a 'request for admin review' ('RfR') with diffs explaining why. Basically similar to the initial presentation of an RfC or RfAr. No rebuttals, statements of support/opposition, et cetera. Just the reasons for requesting review stated as briefly as possible and primarily with diffs.
  2. A Bureaucrat reviews this request one week after it was first posted and judges whether the grounds for reviewing the admin seem sufficient that there is some reasonable chance of the admin being removed. There is no numeric basis for this... one diff from Danny or Jimbo where the admin does something possibly very damaging to the encyclopedia might be found sufficient while a hundred diffs of content disagreement reverts mis-labelled as 'vandalism' by a dozen recently created accounts might be dismissed. Determinant is whether the bureaucrat thinks the listed issues constitute something the community might de-admin over. This step exists solely to weed out/discourage frivolous requests. Obviously bureaucrats connected with incidents in the request and/or the admin should recuse themselves, but we choose bureaucrats precisely to make these kinds of decisions impartially.
  3. If review is called for then the admin has a month to prepare a normal 'request for adminship' form with nominators, acceptance, and answers to a standard set of questions which would differ slightly from the current ones ('What admin tasks do you perform' instead of 'would', 'Please explain your views of the conflict(s) leading to this review', et cetera). If no RfA is prepared after a month it is the equivalent of the candidate not accepting the nomination and the user is de-adminned.
  4. The RfA runs the normal course in the normal way and is then evaluated by a bureaucrat (preferably a different one than approved the review). I'd prefer to see the same standard of approval required, but given views about 'adminship inherently creating enemies' I suppose it might go as low as 'roughly 50% approval' required to remain an admin. If it did start out at a lower percentage I'd like to see it slowly raised over time so that admins are people who continue to enjoy high community support rather than just those who had it at one point.
  5. If the reviewing bureaucrat does not find consensus support for the admin they remove the person's admin status (which became theoretically possible for bureaucrats a few days ago). The person is then free to re-apply at any time - though like other repeat RfA's a waiting period would be recommended.
Throw in some structure where no reviews can be filed until six months after the policy is adopted, that admins can't be reviewed until at least six months after they were promoted/last reviewed (not just review requested, but actually reviewed), and all current admins are notified on their talk page when the policy is adopted and I think that's it. Somewhat similar to DocMkII, but doesn't involve a full RfC, a 'vote' to initiate the review, or the ArbCom. Obviously, in most cases an RfC should be attempted before a 'RfR' is filed and Bureaucrats might take the lack of any effort to resolve the issue into account in evaluating whether to approve the review, but I don't think RfC should be directly tied in. The above is very close to the existing RfA process with the same decision makers (bureaucrats) and a minimum of extra procedure around it due to the special circumstances. --CBD 10:45, 16 August 2006 (UTC)
Looking over this latest suggestion actually helped point out the same problem I see in Aaron's Grey Box proposal that had been niggling at my brain. Your suggestion hinges on a Bureaucrat's approval to move forward. The Grey Box proposal rests on Arbcom's shoulders. I think in either case it's going to make these parties unpopular. What becomes of the Bureaucrat who decides not to forward on an unpopular admin? What happens when Arbcom declines to hear a removal case just based on community vote?
I believe what this page really wants is community oversight over the admin removal process, and while both these proposals seem to offer it, it isn't all-the-way. And while it may go over well at first, I think it will just cause more grief down the road. --InkSplotch 13:17, 16 August 2006 (UTC)

Aaron/all, I made an archive at Wikipedia:Administrator_recall/Tenure_Versions for whenever/if we make a major change to work on developing the Grey Box solution. We can just drop a straight copy the "final" version of Tenure if we go away from that to this archival page with the diff structure to that subpage and then link it off the main one. For historical/ongoing comparison purposes etc. Just need to put in the URL of the "final" version afterwards to match the one for the first. rootology (T) 15:07, 16 August 2006 (UTC)

Aaron, why would you wipe out this policy to introduce another one? WHy not just start a new policy proposal? And if you're going to archive, please don't move everything and please don't archive active discussions this time. Karwynn (talk) 16:52, 16 August 2006 (UTC)
Struck out already answered question above. Here's the deal: it may not be the case that everyone wants to abandon this "version", so a new proposal would be better. Besides, why have two choices and not just one? There can easily be cross-links to both proposals and their respective talk pages. Unless there is consensus to completely abandon this, and there's not, it makes no sense to shove one proposal out of the way when you can just as easily make a new page. Besides, rejected proposals are supposed to be kept, not steamrolled. Karwynn (talk) 16:58, 16 August 2006 (UTC)
I don't think Aaron is attempting to steamroll anyone, I think he's just trying to focus everyone's efforts on the most popular proposal (a judgement call, to be sure). The risk in splitting out competing proposals is that apart, each one would take longer to develop, longer to find consensus, and eventually risk some sort of competition for acceptance. If everyone stays on one proposal, no matter how much it evolves, it will evolve more quickly with more momentum. --InkSplotch 23:34, 16 August 2006 (UTC)

Just curious what most admin complaints on this policy would be about?

Would they be about assuming bad faith, personal attacks, deletion abuse, banning too much, or something else? I think that most would be about assuming bad faith. I was reminded of something I asked in ebay's forums about buyers. I was told that if you approach things that a problem buyer (e.g. one that causes lots of trouble, never pays, negs you, that type. I almost refered to them as trolls.) is a crazy person then you will never reach a resolution or understanding. Anomo 04:49, 16 August 2006 (UTC)

Probably all of the above. THey're all applicable policies. It depends on the filer. Karwynn (talk) 16:49, 16 August 2006 (UTC)
You got it. I have too much faith in the hard core trollers, Wikipediareviewers who claim thousands of edits, sockpuppets, et cetera. My suggestion would be to require a RFCU for every person who votes for the recall to check for sockpuppets, but that would never fly. User:Zoe|(talk) 01:24, 17 August 2006 (UTC)
Why not just apply the same standards they do for RfAs? Are there any effective methods at troll-countering there? Karwynn (talk) 14:15, 17 August 2006 (UTC)
No. User:Zoe|(talk) 21:20, 17 August 2006 (UTC)

Temporary desysopping

The German Wikipedia has a recall procedure where an admin's sysop flag is first removed temporarily (it is removed by a stweard and automatically restored by a local bureaucrat) after a recall vote with 2/3 consensus (AFAIK) of editors with more than 200 mainspace edits. This acts as a warning (analog of a short block), and only after this has been done and problems persist, a new RfA can be initiated. Maybe we should also have a method for temporary desysopping. That both sysopping and desysopping are currently quasi-permanent is perhaps not optimal. Kusma (討論) 14:12, 16 August 2006 (UTC)

"Grey Box" vs. "Tenure" versions - pro/con?

Just to make a standalone section; what criticism can be offered of Doc's Grey Box version (good name for it, by the way; my collective incrementing versions that spawned this can be called the Tenure Versions just for clarification) as compared to the Tenure version? What is better about the Grey Box version than the Tenure Version? If possible I'd be curious for specifics. rootology (T) 14:57, 16 August 2006 (UTC)

The problems with the "Tenure" version are all over this page. No provision for dismissing frivolous complaints, insufficient protection of admins from users who hold a grudge against them, and the chill this puts on admins who perform controversial tasks like AfD closings, speedy deletions, bans and page protections.
The "Grey Box" version is, as Tony says, our standard dispute resolution system with a pointless extra step thrown in which will only increase tensions between users and administrators. Fagstein 16:16, 16 August 2006 (UTC)
Oy, I'm getting a headache. The dismissal of frivilous claims is for the admins to not certify/not get 31 total... rootology (T) 16:18, 16 August 2006 (UTC)
I'm unconvinced that 30 people agreeing to something makes it non-frivolous or automatically gives it merit. Fagstein 21:53, 16 August 2006 (UTC)
In your opinion, what would be a fair, bullet proof, community-level (editors and admins) to make this determination? rootology (T) 21:55, 16 August 2006 (UTC)
People keep bringing up your same concerns about the "tenure" version, Fagstein, and then they ignore replies. Read some of the talk and you'll see that there are some solutions. Karwynn (talk) 16:48, 16 August 2006 (UTC)
There are wildly contradicting proposed solutions which haven't made it into the "tenure" version. And none of them tackle the core problem that the purpose of this is to de-admin people that ArbCom isn't willing to de-admin. Fagstein 21:53, 16 August 2006 (UTC)

Several points

This proposal seems to treat checkuser and oversight rights as being "admin plus"; they're not. Checkuser and oversight are entirely independent of being an admin (one can actually have checkuser without being an admin, although the odds of that are rather remote). Secondly, checkuser authority is granted by the Foundation, not by the community. The Foundation policy on granting and rescinding checkuser rights does not give the community the power to grant or rescind them in a wiki which has an Arbitration Committee. Therefore, to the extent that this proposal purports to be able to revoke checkuser rights, it conflicts with Foundation policy. A similar argument applies to oversight.

In general, the entire "exceptions" section is grossly misfounded and should be scrapped. Kelly Martin (talk) 15:08, 16 August 2006 (UTC)

Based on your knowledge of policy, what should an exclusions section look like? I'd certainly think that some people (i.e. Jimbo) should be except from Recall. But thats just me. rootology (T) 15:11, 16 August 2006 (UTC)
I view this entire proposal as misfounded, and as such am not inclined to participate much in its refinement. In my opinion, the current policy (let the ArbCom decide when it is appropriate to revoke adminship) is sufficient; the only alternative I would contemplate is an administrator's council which both grants and revokes adminship based on a deliberative process, taking input from the community but not obliged to even entertain, let alone count, votes. Kelly Martin (talk) 16:51, 16 August 2006 (UTC)

Ok, from someone interested in bettering this policy, I'd say exclude stewards, bureaucrats, Jimbo, arbitrators - basically everyone who is now excluded except Checkuser people, per Kelly Martin's comments. Karwynn (talk) 16:53, 16 August 2006 (UTC)

Why would bureaucrats and arbitrators be excluded? I don't like the idea of "normal" admins versus "higher than admin" (the proposal also mentions "basic" admins). For example, arbitrators are not technically required to be admins. Desysoping an arbitrator would not remove them from the ArbCom. I mention this because one of the major problems I have with this proposal is that it's been built in a Frankenstein manner, grafting on arbitrary rules, exclusions and limitations whenever there's complaints. Aren't I Obscure? 17:28, 16 August 2006 (UTC)
I'm just aiming to exclude anyone who has a say in office decisions/bans. If B'cats and arbitrators don't apply to that, forget them. THe point here is not to present a challenge to WP:OFFICE. Karwynn (talk) 17:41, 16 August 2006 (UTC)
There is no logical reason why you should exclude bureaucrats and arbitrators, but not checkusers. As I stated above, the community has no authority to revoke checkuser authority anyway, so all you can do is strip the admin flag without stripping the checkuser flag. Arbitrators can be recalled as admins without affecting their authority as Arbitrators. There is no process to remove an Arbitrator, other than the Will of Jimbo (all Arbitrators serve at Jimbo's pleasure). I see no logic to exempting bureaucrats, either, since a b'crat is just someone who has been trusted to make judgment calls on RfA and to handle a couple of relatively uninteresting ministerial functions related to usernames and bot flags. The checkuser right embodies the highest trust level available on any single project (since the disclosure of information discovered by checkuser is just about the only truly irrevocable act on Wikipedia); for you to suggest that it should be nonexempt when these others are exempt merely indicates that you are really quite clueless as to the manner of governance of Wikipedia and the other projects within the Wikimedia Foundation umbrella. As to WP:OFFICE, formally that process does not involve anyone other than Foundation staff (which you have failed to exclude); Arbitrators, bureaucrats, checkusers, or have no direct input, and if your goal is to "avoid interference with WP:OFFICE" then you have no reason to exclude anyone other than Foundation Board or staff. (I truly doubt that you understand WP:OFFICE at all, however.) Kelly Martin (talk) 17:48, 16 August 2006 (UTC)
I'm sure rootology will take that into consideration. You're knowledge of policy would be helpful, though I have to say I don't see what you're getting so belligerent about. Karwynn (talk) 17:55, 16 August 2006 (UTC)
(edit conflict) Shouldn't removing sysop access and removing checkuser access be separate things? There's a difference between if someone does a bad job at one task than the other. Sysop is about handling power and Checkuser is about trust. I read comments on a forum that one person without checkuser was getting private checkuser data from their friend who had checkuser, not using the requests for checkuser just getting IPs themselves. It was none of the people who have posted on this talk page, though, just FYI. Anomo 18:03, 16 August 2006 (UTC)
They already are, read Kelly's comment. Under the current systems, the two are different processes. Karwynn (talk) 18:05, 16 August 2006 (UTC)

A historical perspective

I note that most of the participants in this discussion weren't here when the Wikipedia:Requests for de-adminship page was open for business. I was, and I was here for the "quickpolls" experiment that followed. In general, these processes were a haven for anyone who had a grudge against an administrator, regardless of its merits. Most of the listings were retaliatory. Since there have always been people involved in Wikipedia who are critical of substantially all admins as a group, a certain baseline level of support was present for any request for de-adminship.

Much of the reason the arbitration committee was formed was to deal with user conduct issues in a way that would not be prone to retaliatory listings and other abuse. I believe that it is effective.

The Uninvited Co., Inc. 18:47, 16 August 2006 (UTC)

As do I, which is my problem with the current proposal. I do think there is a role for some sort of community reaffirmation of existing administrators, to ensure that they still carry the trust of the project-at-large, but structuring any proposal in such a way that a) it's going to deal exclusively with issues not really appropriate for the arbitration committee, and b) it's not going to run rampant with abuse, is no small order. I'm always receptive to proposals to address these problems, but while I do think a problem exists, on some level, I think most of the solutions I've seen proposed so far are as bad (or worse) in various ways. – Sean Daugherty (talk) 19:24, 16 August 2006 (UTC)
One of the interesting things about the perennial proposal for a de-adminship/adminship renewal/adminship recall process is that the handful of bad admins in the back of everyone's minds during these discussions invariably end up leaving the project after a period of some months anyway. They usually flame out when they realize they don't have much support in the community for what they're doing, and then they leave. In a couple of cases such individuals have been stripped of their adminship through existing means, e.g. arbcom, or have voluntarily given up their adminship. The existing processes are more effective than one might realize, it's just that they're slow. The Uninvited Co., Inc. 21:40, 16 August 2006 (UTC)
The problem, though, is that there are a few who don't leave, and even in the case of those who do, the damage done before they leave is often irreversible. But you do have a valid point, and while I'd like to see a more efficient process for dealing with problems, speed shouldn't come at the expense of accuracy or fairness. – Sean Daugherty (talk) 22:36, 16 August 2006 (UTC)

Tony's Hybrid: peer/arbcom oversight

This is the most progress yet on this, and a good idea as it's a combination of community endorsement (the RFA you mention), and extra, simple oversight. Do you think adding a new section to the bottom of WP:RFAR with a format like the proposed Filing format on this policy would be a good way to do this? It would be basically a reformatted ArbCom submission, where you simply list who is in question, the basis of dispute, and examples of policy violations (for starters)? rootology (T) 22:24, 16 August 2006 (UTC)

  1. User submits request at ArbCom page, with below filing format (feel free to edit the filing format).
  2. ArbCom in up to 7 days (give or take) reviews this. If 4+ ArbCom agree it's with merit...
  3. ...it moves to a some form of Recall RfA or ArbCom decides to "escalate" it to a full ArbCom review.
  4. The Recall RfA structurally is the same as a normal RfA, except a support/Recall vote and commentary is to desysop.
  5. As a normal RfA, a closing b'cat will make a decision to desysop based on everything presented. A steward will remove access if the b'cat decides on this.

We on the same page so far...? rootology (T) 22:32, 16 August 2006 (UTC)

I think this could be good. --Tony Sidaway 22:38, 16 August 2006 (UTC)
Let's edit off my new path/idea here, to see what can be hammered out. How about now? rootology (T) 22:44, 16 August 2006 (UTC)
Needs a bit more than you and me enthused about it. I've given my input and I'd like to see what others say. --Tony Sidaway 22:52, 16 August 2006 (UTC)
I like it! I think we (me included) were a bit too hasty back in the day to throw the Stevertigo case back to the ArbCom. Haukur 13:14, 17 August 2006 (UTC)
Could work. As long we insist that the filer must show that other admins have attempted discussion/correction of the 'offender' first. Discussion, not discipline, should always be the starting point. Further, arbcom should have the option of repremanding the admin, if there have been breaches, but a full RfA would be excessive. --Doc 22:58, 16 August 2006 (UTC)
That's sensible. Individual arbitrators are habitually unstinting in their expression of opinion about cases. They can reject a case while suggesting that there's a problem. They can also politely hint that the complainant should try to discuss things with the administrator first. From my experience as an arbitration committee clerk, I think this latter is likely to happen quite often. --Tony Sidaway 23:34, 16 August 2006 (UTC)
So, Arbcom approves the reverse RfA instead of a single Beauracrat? Well, a load off the Beauracrat's minds, I'll bet. Well, I hate to sound negative yet again (because I feel I have been this whole page), so I'll say: it's the best proposal yet. It won't stop cries of favoratism or cabalism, but I expect most spurious requests could be quashed by arbcom fairly easily and quickly. I just wish I could be convinced it's necessary. I mean, I understand what the folks who've worked so hard on this want, but I just don't know if this will really fulfill their wishes. I fear many will embrace it now, to feel disappointed later when they realize it really doesn't make removing an admin that much easier. I'll stop the doom, gloom, and italics now. --InkSplotch 23:28, 16 August 2006 (UTC)
Basically I think the arbitrators are trustworthy and I know they are quite happy to have community input on administrators (See for instance the case Wikipedia:Requests for arbitration/Stevertigo). If they're amenable to this, I think it would be better than foisting the responsibility onto a single bureaucrat (who really isn't selected for his willingness to review evidence and express his personal opinion). --Tony Sidaway 23:34, 16 August 2006 (UTC)
I agree with every you've just said. My only fear is that, as noisy as some admin RFCs have gotten, when Arbcom turns down a spurious request with large, vocal support, that could damage their trustworthy reputation needlessly. I certainly think they're more suited to this risk than Bureaucrats, but even so, I feel like it's asking them to take more mud slung their way. --InkSplotch 23:38, 16 August 2006 (UTC)
That's their problem. They shouldn't reject reasonable applications. They're used to rejecting unreasonable ones up to and including demands for desysopping. Such practice helps to keep them connected to the concerns of the community. This would simply provide them with a way of redirecting more obvious problems of administrator trustworthiness to the community. --Tony Sidaway 23:46, 16 August 2006 (UTC)
Looks good overall. Is there a place for multiple signatories, in the case of distributed situations where any one user doesn't have the standing to file just based on what has happened to them? -- nae'blis 23:56, 16 August 2006 (UTC)
Could they just file additional evidence/basis under their comment section? Maybe put a delay from filing of 24 hours before ArbCom can accept or reject it to give people adequete time to file their own information? rootology (T) 00:02, 17 August 2006 (UTC)

I have only one word to say in response to this suggestion: Stevertigo.

This is a very good direction to be going - it gives a definitive route that such a proposal will go and a definitive end point, which we're missing at RfC right now. The arbitrators are trusted such that they should be an excellent resource for dealing with these kinds of issues in a positive manner. Good ideas, all around on this suggestion. Tony Fox (arf!) 02:09, 17 August 2006 (UTC)
In answer to a lot of questions on detail, I would really suggest that people watch the Requests for arbitration page closely for a bit. It has its own way of working but the acceptance procedure is very flexible. Once a request for arbitration is started people can add their comments to it, including evidence. Typically it won't be rejected until four arbitrators have voted "reject" and at least seven days have passed since it was first filed. In short, the arbitration system is already well organised and it works well and the fit, it seems to me, is good for this purpose. --Tony Sidaway 02:15, 17 August 2006 (UTC)

Filing format

Filing user: (name/signature)
Administrator being recalled: (name/link to their User space)
Basis of Recall: (description of why the recall is being filed--the verbose section, the "why")
Evidence: (evidence--diffs, citations, policy violations, etc.--to the point, the literal evidence of why it should get a pass from ArbCom to go forward)
Evidence of attempts at dispute resolution:
Administrator response: (response by the admin)
Comments by parties
Arbitrators opinions on hearing this matter.
It would take a few seconds to run up a template for requests for community input on an admin (or some other wording). If we could arrive at the agreement that this would be a good way to proceed, the only thing that needs to be done is to select parameters (the quota the period, and so on). This would be subject to agreement of the Arbitration Committee. I don't see a problem with that; they're likely to be flexible and, I think, only too happy to enable community input to deal with problematic administrators. --Tony Sidaway 22:36, 16 August 2006 (UTC)
Like arbcoms, I'd include a bit for "Evidence of attempts at dispute resolution." --InkSplotch 23:28, 16 August 2006 (UTC)
Added. rootology (T) 23:36, 16 August 2006 (UTC)
Yes, basically I see this as an ordinary arbcom application with the exception that the proposer asks for permission to solicit community input on an administrator with a view to desysopping (or perhaps other remedies). --Tony Sidaway 23:35, 16 August 2006 (UTC)
Now we're getting somewhere. Change the input from strict votes into a discussion (interpreted by ArbCom) and you might even get my support. Fagstein 04:53, 17 August 2006 (UTC)

For the "discussion" are we basically talking it'll be an RfA (Recall RfA), and then to just determine if the votes should be standard format--i.e. Support = Stay On as admin, or Support = Recall/remove? rootology (T) 04:55, 17 August 2006 (UTC)

I think "support" should be to support the recall/removal of the admin. And on another note (sorry to branch off), I think we should notify all arbitrators of this to get their input. I really like this proposal, but it depends too much on the ArbCOm for them not to be notified. For one thing, if arbitrators simply won't desysop anyone based on lack of community confidence, this whole proposal is worthless. Karwynn (talk) 14:21, 17 August 2006 (UTC)

Filing criteria

To keep the filings down to legit levels and dissuade crap, what do you think of a modified 5 month/1000 edit threshold to file?rootology (T) 22:49, 16 August 2006 (UTC)

There are no filing criteria on RFAR. If an application is crap, the arbitrators toss it out. Although Wikipedia policy and practice are only acquired with experience, the arbitration committee is down-to-earth and will not reject a reasonable case simply because the proposer hasn't been around for more than a coupld of days. --Tony Sidaway 22:56, 16 August 2006 (UTC)
That works. rootology (T) 23:03, 16 August 2006 (UTC)
Interesting lesson to be learned from that, too. Karwynn (talk) 14:22, 17 August 2006 (UTC)

Functional blocks

There appear to be several conceptual "chunks" that we agree upon, now we're really thinking about ordering. Ignoring how much more esthetically pleasing Wikipedia talk:Admin recall/DocMkII is, it and Tony's version have the same basic steps. Rather than spending lots of time going down to ever-smaller levels of detail (like the format of filing) can we concentrate on the high-level sections and only start to think about the tactical when we agree on the strategic?
brenneman {L} 00:21, 17 August 2006 (UTC)

Tony's
  1. Filing (one user, to Arbcom )
  2. Arbcom approves
  3. Quasi request for adminship (consensus required to kick out)
  4. B'cat closing
MkII
  1. Filing (multiple users, via request for comment)
  2. Quasi request for adminship (consensus required to keep in)
  3. B'cat closing
  4. Arbcom approves
Current version
  1. Filing (lots of users, via request for adminship)
  2. Quasi request for adminship (consensus required to kick out)

Discussion of functional blocks

Sorry to complicate things - but my MK2 was much simpler (until you added back in elements of Mrk1) it didn't involve crats, votes, or quazi RfAs. If was just a request during an RfC, which if certified by a min of six admins, goes to arbcom. Much simpler. --Doc 00:34, 17 August 2006 (UTC)
I didn't see that. Can't people just look at an RfC and say "hmm, time to go to arbitration"? This is how it's done now. And you can even miss out the RfC stage. Of all desysopped admins, I can't off-hand think of a single one that involved a prior RfC. If there's a problem with an admin, and it's obvious enough to call into question their use of the bit, bring it to arbcom at once. --Tony Sidaway 00:38, 17 August 2006 (UTC)
My "MkII" nomenclature was based on Doc's proposal being "MkI." - brenneman {L} 01:15, 17 August 2006 (UTC)
Regarding RfCs before an admin was desysoped: Wikipedia:Requests for comment/Stevertigo Aren't I Obscure? 02:48, 17 August 2006 (UTC)
I like Tony's the best, "current" second. Taking this to RfC would be unwise. This sort of thing just turns in to a flamewar when filed "against" a high-profile editor of any kind. Plus, active RfC's have been deleted. I think we should keep it to a venue that cannot be removed or subjected to other possible abuse, like described in rootology's original proposal. Karwynn (talk) 14:25, 17 August 2006 (UTC)

Confirm vs. Dead-minn

My very strong preference if that the administrator in question must meet the same hurdle as a new nominee, that something like 75% of people must support their continued adminship. Getting consensus that someone must be removed is a much different (and more difficult) than than determining if they will be allowed to continue. Voting to dead-min is tantamoung to saying "This is a right, shall we take it away," while asking for reconfirmation is more like saying "This is a privledge, shall they keep it." - brenneman {L} 02:34, 17 August 2006 (UTC)

Aaron, somehow my brain glazed over and misread this. I think this is a bad idea, and will lead the policy to never get passed. Even my original policy draft(s) never called for this--it was always 75% plusish had to say "deadmin him", that in of itself seemed to have little practical resistance. rootology (T) 16:57, 17 August 2006 (UTC)
That's impracticable. Few really effective admins could withstand a determined campaign by the disaffected trolls and miscreants who would turn up to vote. In my opinion we should be asking whether the community has lost confidence in an administrator. It's in the nature of these things that it will attract groupies and trolls with an axe to grind, so there's no sense in encouraging them. --Tony Sidaway 02:52, 17 August 2006 (UTC)
That's your opinion based (perhaps) upon the feedback that you receive on your adminstrative and editorial actions. I do unpopular things and editors often have a go at me, and find that in almost every case when I take the time to engage the person respectfully the problem goes away. Unless your contention is that greater than 25% of the wikipdian community consists of "trolls and miscreants" and that following that the b'cat would zombie-like count the numbers and that the commitee apparentlly asleep at the wheel would certify? --Aaron Brenneman 03:25, 17 August 2006 (UTC)
That's a bit harsh, isn't it? --InkSplotch 03:32, 17 August 2006 (UTC)
The Arbitration Committee has historically used RFA affirmation votes only in situations where they have found administrators guilty of actual abuse unworthy of an administrator. It follows that a desysopping vote, which is only based on prima facie allegations of abuse which the committee declines to arbitrate (signalling that they don't see a major conduct issue) should require strong community consensus. The Committee always has the power to summarily order the removal of a sysop bit if there is a major problem. --Tony Sidaway 18:52, 17 August 2006 (UTC)
  • I didn't mean it to be "harsh" and if it came out that way I apologise without reservation. Tony and I have quite different expectations regarding the nature of the community, and (without casting aspersions) as he's on the receiveing end of a great deal more heat than I am regarding conduct it might not be unreasonable for him to feel that way. For example, Tony and I have both had the same number of Arbitration rulings handed out, but he's had four (six?) requests for comment. Is there some way for me to say that without it sounding like a dig? Because it's not meant to be. --Aaron Brenneman 03:45, 17 August 2006 (UTC)
I don't think so. Don't worry too much, you certainly won't be the last to say something along these lines. Any policy or process to de-sysop someone will be meet with resistance, and criticisim of the process by an admin will invariably lead to comments like this. Whether meant as an insult or not, I don't think it'll ever help matters along. Any admin discussing this issue will have a different viewpoint from a standard editor. Whether they have good reason (or we just expect them) to be concerned, everyone's motives are ultimatly their own. I have faith that every admin, experienced or not, contentious or quiet, will put the encyclopedia before themselves in this. Wow, I ramble a lot, don't I? --InkSplotch 04:04, 17 August 2006 (UTC)
Aaron, whether or not 25 percent of the Wikipedian community is trolls and miscreants is not relevant. What is relevant is the fact that nowhere near 1 percent of the Wikipedia community ever votes on anything - and with a little concerted talk-page and off-wiki campaigning, it would be ridiculously easy for an organized group to generate enough votes to derail just about any such renomination campaign for admins who have done anything remotely controversial. Especially given that to begin with, the people motivated to vote on these proposals will be the people who want the administrator gone. FCYTravis 06:14, 17 August 2006 (UTC)
Voting to remain would set a higher bar than voting to remove, but one with the potential for abuse. However, I think if the arbcom comes before the vote, and has the option to reject/halt proceedings without having to opt for a vote or arb case, it might be viable. --InkSplotch 03:36, 17 August 2006 (UTC)
I think that if an admin is worthy of the mop and bucket, they should be able to muster support from people they know in the recertification, trolls be damned. I say 75% support means you keep the bit. Otherwise I think you'll run into a LOT of "But I thought we were supporting his de-adminship", and the results get all muddled and thrown into doubt for the closing 'crat to sort out. -- nae'blis 04:04, 17 August 2006 (UTC)
A couple points to keep in mind when deciding which way to go:
  1. Every case where the ArbCom has told someone to be re-confirmed for adminship has failed.
  2. Ta bu shi da yu, possibly the most popular admin on Wikipedia, acquired twelve "oppose" votes on his request for re-adminship, and only two of them referred to his vandalism of Dalek as the reason for opposition.
--Carnildo 04:51, 17 August 2006 (UTC)
That's at Wikipedia:Requests for adminship/Ta bu shi da yu 2 and it's worth looking at Wikipedia:Requests for adminship/Sean Black 2 as well. --Aaron Brenneman 05:30, 17 August 2006 (UTC)
See also Wikipedia:Requests for adminship/Ta bu shi da yu 3, where Ta bu eventually reclaimed the bit. Also, it's not fair to count Ta bu's second RFA as a failure, since it was voluntarily withdrawn at a ~88% approval rate. In addition, Wikipedia:Requests_for_de-adminship says that User:Guanaco eventually regained the bit (on the 4th try); the fact that he later lost it again isn't really relevant to your assertion that "Every case where the ArbCom has told someone to be re-confirmed for adminship has failed." unless you meant it strictly to refer to the first attempt, which seems distorted. -- nae'blis 15:01, 17 August 2006 (UTC)
I will not support any proposal in which an admin being recalled must garner a 75-percent supermajority to retain their bit. It is far, far too easy to votestack against anyone who is at all controversial - and furthermore, it presumes that the administrator is guilty and should have it removed unless he can prove otherwise. The burden of proof must be on those who seek to remove the sysop bit, to demonstrate that the administrator has committed actions which merit desysopping. FCYTravis 06:02, 17 August 2006 (UTC)
75% means you only need 25% to de-admin. That's ridiculous. Most active bold admins have stepped on a few toes, combine that with the small amount of people who think there's a whole lot of admin abuse and would like to see admins desysopped on principle, it would be ridiculously easy for a very small percentage of the community (given how few vote) to block a re-comfirmation. We've also some admins who specialise in things like copyviolating image deletions - and they'd get voted down by people whoose beef is really with policy not the person. Sure, an admin who'd genuinely lost the support of a quarter of the community probably should loose their bit, but not an admin who'd been opposed by only one in four of those motivated enough to vote. It should take at least half of those motivated enough to vote expressing opposition before we consider de-admining. But personally, I'd rather leavve it to our elected arbs - their decision is far more likely to represent the will of the community than any straw poll. --Doc 08:47, 17 August 2006 (UTC)
FCYT & Doc have convinced me. I'm going to have to go with 75% to de-admin. --InkSplotch 13:30, 17 August 2006 (UTC)
Meh. If someone gives up the bit voluntarily, we don't say "okay, 75% have to agree you didn't mean it", but I won't stand in the way if that's the general consensus. I just think that putting this on the same page as regular RFAs means we'll have to come up with a whole new set of terminology (Retain or Demote, perhaps?) for a process where 75% must !vote for de-adminning someone. -- nae'blis 14:24, 17 August 2006 (UTC)
Ok, if this were a democracy I'd go with a supermajority to desysop, but this isn't. It may be best for Wikipedia to have it the other way around, so long as the B'cat and Arbs review it. I'm not sure though. I'd almost say compromise at 50/50, but that somehow seems silly. Karwynn (talk) 14:28, 17 August 2006 (UTC)
Actually when someone gives up the sysop bit voluntarily (in the absence of coercion by arbitrators and whatnot), I don't see a problem with a 25% requirement to get the bit back. Why not? This is presumably someone who worked well as an admin and voluntarily surrendered the bit, quite often for reasons associated with work commitments and so on. --Tony Sidaway 19:02, 17 August 2006 (UTC)
Here's the deal. If an admin has to garner a 75 percent supermajority to retain his bit, that essentially means that the process has already deemed him guilty of whatever he's been accused of. If that's the case, then I will vehemently object to this proposal, because that changes the entire meaning of the B'cat and four-arb approval. Until now, my understanding has been that the Arb review is to prevent any clearly outrageous recall proposals from going through, but not making a full-blown investigative ArbCom case out of the question. They are like the grand jury, and should there be a reasonable amount of evidence suggesting that the community needs to look at the facts and examine this administrator's actions, they will approve the question and put it to a vote. But if the admin has to garner a supermajority to retain his bit, that means that the four-arbcom vetting process has become a finder of fact, and their approval of the proposal means that the admin WILL be desysopped unless the admin can prove otherwise to 75 percent of those who bother to vote. That completely reverses the burden of proof and is completely unacceptable. FCYTravis 16:16, 17 August 2006 (UTC)
Trav, it's never, ever, not for a moment been a case in the proposal that they need need to get a 75%+ majority to KEEP their bit. It's been the other way around from Day 1: 75%+ majority has to say "You're not an admin any longer." rootology (T) 16:23, 17 August 2006 (UTC)
Unfortunately, see above - Aaron Brenneman has proposed that the admin in question needs to get a 75 percent majority to keep their bit. FCYTravis 16:25, 17 August 2006 (UTC)
Somehow I misintepreted that--my intention was never to make it easy to punt an admin, just fair and practically possible if the majority of the community felt it should be that way.... rootology (T) 16:57, 17 August 2006 (UTC)
I agree, FCYTravis, but with some reservation based on the fact that this isn't a trial, it's a process to decide what's best for Wikipedia, and that may not necessarily be what's "fair". Again, I agree with you, but it is worth extra consideration outside the "courtroom" aspect of it. Karwynn (talk) 17:32, 17 August 2006 (UTC)
Then if you believe it's about deciding what's best for Wikipedia, you will agree with me that a mere "vote" does not always take into account what is best for Wikipedia. We have some policies that had to be instituted by fiat, such as WP:BLP - because what is popular is not always right, and doing what is right is not always popular. See the battles over fair use images for an example. Wikipedia has essentially asked its administrators to perform some of the duties of a traffic cop, and nobody would call traffic cops popular. Until, that is, you've had a major pile up and you need someone to sort everything out and get the freeway running again. FCYTravis 18:10, 17 August 2006 (UTC)
That's the point of the b'cat going over the Recall RfA same as a regular RfA. It's only as much a vote as regular RfA is (which as pointed out by others, it's not). I think my initial policy wording scared off a lot of people because of that. rootology (T) 18:15, 17 August 2006 (UTC)
If it's intended to be a process to decide what's best for Wikipedia, it should be an arbitration case. That's the Arbitration Committee's sole job. They can voluntarily subcontract part of that, but in general they're the final arbiters. --Tony Sidaway 19:05, 17 August 2006 (UTC)

Arbcom first or last

Placing the Arbitration committee first as the gatekeeper is in effect no change from the current system. By placing them at the end it maintains a safety valve, put still puts the effective control in the hands of the community. - brenneman {L} 02:43, 17 August 2006 (UTC)

Putting arbcom at the start ensures that frivolous requests, and cases needing a more subtle investigation, are dealt with appropriately. --Tony Sidaway 02:54, 17 August 2006 (UTC)
He beat me to it, but yes, putting ArbCom before let's them decide on either the RfA-type process, or a more discrete arbitration case. --InkSplotch 03:15, 17 August 2006 (UTC)
That's a bit of a straw man. Nothing here would stop arbitration from happening if it emerges that serious misconduct had occured. --Aaron Brenneman 03:20, 17 August 2006 (UTC)
I didn't intend it as a straw-man. I would think ArbCom would want first crack at a case before a drawn-out discussion and popular (consensus) vote had taken place. It's not an option I'd expect they'd exercise often, but if they do I expect it'd be for situations that, as I said, are best handled discretly. --InkSplotch 03:30, 17 August 2006 (UTC)
Bah, that phrase has too many negative connatations. Red herring? Anyway, an obvious extension of that logic would require that a normal request for comment pass before the committee, just in case it was something that they wanted to handle "discretly." - brenneman {L} 03:37, 17 August 2006 (UTC)
I disagree. RfC's do not produce enforcable results. Policies do not change, people do lose their sysop bit, no one is blocked or banned. I think de-adminning someone is more on par with arbitration than an RfC. Also, on further reflection, I think it proper to let ArbCom know of such an important issue earlier in the overall process, rather than later. --InkSplotch 03:43, 17 August 2006 (UTC)
I definitely agree that there needs to be a 'gatekeeper' up front to weed out frivolous requests - though I had previously suggested giving that role to the bureaucrats rather than the ArbCom. I think ArbCom could work if it was on the basis of 'is there any reasonable likelihood of the community de-sysoping this person' rather than 'do we feel this person should be de-sysoped'. I'd object if ArbCom routinely denied requests where it was clear that a sizable portion of the community wanted the person out because to me the only point of this is to restore admin accountability to the community instead of admin accountability to the ArbCom. Other than this ArbCom vs bureaucrats issue Tony's concept seems largely the same as what I proposed (normal RFA with slightly different questions and possibly lower acceptance thresh-hold - though I would like to see that raised over time) and thus is fine with me. --CBD 14:28, 17 August 2006 (UTC)
Have it be able to step in either time. How about, the filing can be reviewed to see if it's pure trollery or if it has some possible merit. Then, at the end of the process, if the ArbCom feels that the "voting" has not been compromised by trolls/sockpuppets, they approve it? THe only say they would have is in trolling and sockpuppeting, and they could step in at any time. Honestly though, this just seems like more work for the ArbCom. Which is something I thought we were trying to sidestep :-( Karwynn (talk) 14:31, 17 August 2006 (UTC)
If Arbcom acts up front, it could become routine for the to reject requests, but it's just a possibility. If they act at the end, after the voting, I think there is more likly to be cries of "foul!" from the peanut galleries. I see having them act up front as doing more to preserve community control over the desysoping process. There would be nothing to stop someone from filing an arb case after the fact to claim an invalid vote. --InkSplotch 17:06, 17 August 2006 (UTC)
While I still prefer the ultimate idea of the community (users/admins) tossing frivilous rejections by a certification, that doesn't appear to be happening. Given that, I'd prefer that ArbCom go up front, and if 4 ArbComs agree to take on the case, it goes to an RfA that requires roughly 70%-75%+ support for the admin to be deadminned, based on b'cat interpretation, etc.rootology (T) 17:12, 17 August 2006 (UTC)
Same here. I'd prefer the community having the say-so in frivolousness, but if people refuse to trust those who disagree with them, the policy won't pass. I think ArbCom is, if anything, trustworthy, so it should still work out. In fact, it might even be a relief - it could stem the tide of troll name-calling that is bound to happen (speaking of which, I hope you don't mind me stealing your userbox.) Karwynn (talk) 17:36, 17 August 2006 (UTC)
It's really about having a gatekeeper. The Arbitration Committee members can make up their own minds on the basis of what they think is best for Wikipedia. If they think that community input would benefit the encyclopedia they will support it, if not they won't. Of course being arbitrators they can't be told how to vote, and they can only be overruled by the Foundation or Jimbo. This is a good thing. It's why we can trust them. --Tony Sidaway 18:44, 17 August 2006 (UTC)

Archive suggestion (done!)

If we're starting to get somewhere again, I'd like to archive the bulk of what came before The Gray BoxTM and perhaps put both DocMkII and Tony's Hybrid in a "two proposals enter, one process leaves" sort of face-off. (Actually, what I hope will happen is we'll get back to basics and pull things of use from both, and come up with something to bring to the community). I think at this point we're confining ourselves too much to the talk page, and most people haven't realized the sea change that's going on behind WP:RECALL. -- nae'blis 01:05, 17 August 2006 (UTC)

I think this is a good idea. Archive, then hammer out something to replace whats on the main page now, then report to the Village Pump? rootology (T) 01:08, 17 August 2006 (UTC)
I support a summary and refactoring, with the caveat that much ofthe top of this page can simply be deleted as it's already in Archive2. Also, even threads that are active can be archived/summarised as long as everyone understands that's what is going to happen. - brenneman {L} 01:15, 17 August 2006 (UTC)
Active threads may still have active discussion. WHy would you cut that off? Karwynn (talk) 14:12, 17 August 2006 (UTC)
Would you let someone besides Aaron archive it, then? The entire first half of this page, with the exception of the straw poll which should ALSO be archived (as it shows strong opposition to the idea that's being/been revised out of existence), no thread before the "White washing" header has a comment from 16 August or 17 August. Many of the ones below that are also about the old proposal or could be summarized as well, and we'd end up with a much cleaner discussion going forward. Plase assume good faith, Karwynn. If you want, I'll archive the obviously dead threads back into the archive pages (where their edit history still exists) later today. -- nae'blis 14:22, 17 August 2006 (UTC)
Who said anything about bad faith? I'm just saying it was bad judgement to archive active discussions. I could care less who does it, and I must have been to harsh for you to think I do. I just don't want active discussion archived by anyone. That was my only concern. Archive dead discussion all you want. Karwynn (talk) 14:33, 17 August 2006 (UTC)
Okay, it looked like above the comment about "Why would you cut that off?" shows a misunderstanding between the two of you on what Aaron's intent was (I believe he was being unclear in what he's trying to do). I was just trying to provide a third way... I've got a meeting in a bit, but I'll try to get to the archival soon if no one else does first. -- nae'blis 15:05, 17 August 2006 (UTC)

I've gone and done it, up to Zoe's statement about archiving which was nearly 48 hours ago using copy/paste archiving. It is a good idea. rootology (T) 16:03, 17 August 2006 (UTC)

And I took out the straw poll (since it applied to the now archived original proposal, at Wikipedia:Admin recall. I'm in the process of setting up the main page to be more reflective of what we've done here. -- nae'blis 18:55, 17 August 2006 (UTC)

....Aaaaaand DONE! See what you think of the proposal page now, and especially whether or not we can archive any more sections out of this talk page. Really, I think #Functional blocks is the oldest section we need, including its subsections about confirmation vs. de-adminn. I'd be willing to support making everything above that Archive 3, or even part of Archive 2. Maybe leave the filing format section, but really that's for a template somewhere, not the overarching policy suggestion here. -- nae'blis 20:53, 17 August 2006 (UTC)

Good job! As to archiving active threads, it helps to understand this if you get the idea that signed comments are "owned" out of the way. The value in an active discussion is not usually whomever said whatever, but the the ideas that are exchanged. Also, "archive" doesn't just mean "dump in the obliette," it should also incude a careful and sensitive summary of what's gone before. If an active thread (meaning one that people are editing) is not adding any extra ideas that it's not such a bad thing to roll it up into a little ball under "FAQ." That would still be an excellent idea to have in a big box at the top of this page, by the way. - brenneman {L} 02:43, 18 August 2006 (UTC)

Points of debate

I think there are only a couple of really significant areas of disagreement currently;

Gatekeeper: What should we do to 'weed out' frivolous de-admin requests?

A - Have an RfC and go to 're-RFA' if some number of users/admins call for it.
B - Have ArbCom review requests and approve a 're-RfA' if appropriate.
Karwynn (talk) 17:47, 17 August 2006 (UTC)
C - Have bureaucrats review requests and approve a 're-RFA' if appropriate.
D - Have ArbCom review the 're-RFA' after the fact and only desysop if they approve it.
E - Don't have a gatekeeper - just let the community reject frivolous requests.
F - Other

Approval: Roughly what degree of consensus support should an existing admin need to remain an admin?

A - 75% to 80% - Same as getting adminship in the first place.
B - 50% - Admins should be supported by roughly half the community to remain admins.
C - 20% to 25% - Only de-admin if the community overwhelmingly supports it.
Karwynn (talk) 17:47, 17 August 2006 (UTC)
D - Other

Are there any other major disputes? I think most people agree that the primary format should be very similar to a current 'request for adminship'. My current preference would be to try it with 'no gatekeeper' (E) and then either have bureaucrats (C) or ArbCom (B) become the gatekeeper if we find that alot of frivolous requests are being filed. On percentage I'd prefer 75%+ (A), but could live with something as low as 50% (B). If we did start out at 50% I'd like to see it periodically reviewed to slowly raise the standards of admin approval. I suspect that the only people who wouldn't get at least 50% approval are those who are routinely de-adminned by the ArbCom anyway. --CBD 17:03, 17 August 2006 (UTC)

My preference is Gatekeeper E (with an oversight in the form of certification thresholds), and Approval C. I am set on Approval C, but will not object to a B. A is too much. For the gatekeeper, while I support E, I will have no objection to an implementation that has either Gatekeeper A or Gatekeeper B. Gatekeeper D is not acceptable to me. rootology (T) 17:28, 17 August 2006 (UTC)
My ideals: Gatekeeper E, Approval A (at about 70%, not 80). This is not a fair trial, it's a vote of confidence and should have the same standards as the original RfA. Benefit of the doubt does not always help Wikipedia. Additionally, the community can and should be trusted to weed out trolling, not with admin tools, but by ignoring them as WP:TROLL says anyhow.
My "real world" support: Gatekeeper B, with ArbCom only stepping in in cases of "votes" compromised by trolling, socking or frankly, idiocy; Approval C. Approval B doesn't really make sense - either they need a consensus to be desysopped, or a consensus to be re-sysopped.
The reason my real world supports are different is that I don't think my ideals will get this supported by the community anyhow. People who are interested in shutting this down don't trust the community to weed out trolling, so ArbCom stepping in will be required, if nothing else, to get this to be real policy. Like I said, it may even weed out all the name-calling done by certain editors willing to disrupt these and other dispute resolution methods. My approval support is also based on getting this passed. I regret that this may make it ineffective though, I'd prefer something along the lines of 30 or 33 percent.
I am willing to support a policy consisting of Gatekeeper B and Approval C, which I believe is a reasonable and workable compromise. I still don't think it's absolutely necessary, but if it will serve to increase community trust in the administrator oversight process and provide a backstop in cases of egregious misconduct (I am no Pollyanna, clearly there are cases of admins abusing their powers) then I believe that on balance, it will benefit the encyclopedia. FCYTravis 07:14, 18 August 2006 (UTC)

Terminology: or, WTF is a "recall"?

A simple observation: the term "recall" should preferably be avoided in any policy that is enacted; this term is by no means universal (it is essentially unknown in Britain, for example), and as such it is likely to be incomprehensible or even misleading to many Wikipedians. If no alternative term can be decided upon, then at the very least the term "recall" must be explicitly defined before its first use. — Haeleth Talk 17:07, 17 August 2006 (UTC)

My use of Recall RfA was based on the dictionary, and the general idea/theme of a recall election, that is:
1. To ask or order to return: recalled all workers who had been laid off.
2. To summon back to awareness of or concern with the subject or situation at hand.
3. To remember; recollect. See Synonyms at remember.
4. To cancel, take back, or revoke.
5. To bring back; restore.
6. To request return (of a product) to the manufacturer, as for necessary repairs or adjustments.
So that, in essence, the admin in question is being "recalled" to the RfA process for community reevaluation. Plus, using WP:RECALL is a bit bold, to imply the importance/severity of the matter at hand, and my calling it a "Recall RfA" was literally to differentiate it from regular RfA. I've tried to be very consistent in referring to "Recall/recall" as the overall process--based on various versions and drafts, from notification, filing, approval/certification, the RfA itself, to a b'cat either hitting the desysop button of simply archiving the whole unsuccessful operation. "Recall RfA" itself I've always tried to refer to as the 'final phase', the literal inverse RfA where the community makes it's opinions heard.
So! "Recall" = the whoe process, end-to-end; "Recall RfA" = the literal inverse RfA/final phase. rootology (T) 17:17, 17 August 2006 (UTC)
Hmmmm. It would appear that motion of no confidence is the term in many places where recall elections are not used, but that implies a level of parliamentary bureaucracy that I'm not comfortable with. I'll be sure to link to both articles as "real world" examples of similar procedures, though, in my revised introduction. -- nae'blis 19:33, 17 August 2006 (UTC)

Active vs. Inactive admins

I think this should take into account the amount of activity an admin did in the last X period of time. An admin that was close to inactive should have a lot less leeway than an admin who is very active. If an admin is on wikipedia working 16 hours a day then they're going to make more mistakes than someone who has just watches 10 or so of their favorite articles. I also think this proceedure should be able to de-sysop someone for purely being inactive. I mean really inactive. You should see WP:PRIMO. If an admin goes inactive for a long time, then either one of their kids finds the password and goes "oh joy what fun!" or the account gets hacked into by someone. Also we can relate to what FCYTravis said about some admin work will cause more criticism. I think if an admin does tons of this they need leeway. But if an admin does nothing but control their favorite articles, then they need some real scrutiny. Anomo 19:01, 17 August 2006 (UTC)

In any given month, perhaps only 10% of all Wikipedia admins are active. To claim that if an admin goes inactive for a long time, then either one of their kids finds the password and goes "oh joy what fun!" or the account gets hacked into by someone." is to misrepresent the facts. --Tony Sidaway 19:12, 17 August 2006 (UTC)
The second part of this is a non-starter. If you've got evidence that a particular account has been compromised, then bring it up and the situation will be dealt with VERY quickly. Until then, admin bits on inactive accounts aren't likely to be cause for de-sysoping, although any process that is developed could potentially take that into account (which seems to be your first point). Hacking generally requires an emergency response in any case; you don't want to be waiting days or weeks for a result. -- nae'blis 19:32, 17 August 2006 (UTC)
The suggestion that inactive admins should be desysopped is routinely made, and just as frequently rejected. --Doc 19:55, 17 August 2006 (UTC)
Actually the percentage is closer to 80% if you include admins who make an edit. If you only include admins that do at least one admin-specific thing, the percentage is closer to 60%. The data don't tell much more of a story than that. There are usually a dozen or so admins that numerically have done most of the blocks, deletes, and protects, but that's in part because they're making a twitch game out of RC instead of taking on tasks that require a good deal more investigation and other legwork. The Uninvited Co., Inc. 20:39, 17 August 2006 (UTC)

It's not one month, it's one year and I think after six months someone should look into it. I haven't seen any compromised, but the possibility is a reason not to let inactive admin accounts go on forever. I also said my reason for scrutiny of ones barely active is someone really active is likely to make a mistake somewhere and get a complaint so small things build up and the admin gets lots of criticism whereas someone who is almost inactive can do bad things and not get noticed because basically nobody sees it. Anomo 20:56, 17 August 2006 (UTC)

We've never had a compromised admin account. If we did, there's a limmited amout of damage it could do before desyopping would occur as an emergency measure. We don't need to generate solutions to non-existing problems. --Doc 21:18, 17 August 2006 (UTC)

Another proposal

- Have admins apply each year to have their admin status renewed. - If they still have consensus among other admins that they should continue to be an admin, then their status in renewed for another year. - If they fail to get consensus, they go through another RFA.

Some points:

  • I believe this avoids the situation where "trolls" or "POV pushers" could go after an admin for their actions.
  • While there is no support for automatically desysopping inactive admins, I think we can agree that all admins should still want to be admins. By having them reapply it makes their intention to remain an admin clear, while not reapplying would be a sign that they do not want to remain as an admin.
  • I envision that the vast majority of admins will basically have their status renewed without the need for votes or comment, it will only be the rare few where another admin feels the need to oppose renewal that will go through a more rigorous process. 12.75.3.141 21:03, 17 August 2006 (UTC)
This had been proposed before and failed. Anomo 21:04, 17 August 2006 (UTC)
No good. There are 1,000 admins. That's 20 confirmations a week, and it will grow. Unworkable. --Doc 21:16, 17 August 2006 (UTC)
I don't see why... 99% of these should require neither comment nor action. 12.75.0.173 21:50, 17 August 2006 (UTC)
I'm not sure what process you're envisioning that would require "neither comment nor action", and yet determine that the adminstrator still had the confidence of their peers. Can you elaborate? -- nae'blis 22:04, 17 August 2006 (UTC)
Well, I'm assuming that there would be no objections raised to the vast majority of admins reapplying. In the absense of any objection, there needn't be other comments or actions. Just list your name -> no objections raised -> continue being an admin. I don't see how this would become unworkable. 12.75.0.119 22:20, 17 August 2006 (UTC)
Considering that there are always one or two no votes on all RfAs just because people think "there are enough admins" or the requirements for adminship are too loose, there will surely be at least one or two votes for removal just because. User:Zoe|(talk) 23:39, 17 August 2006 (UTC)
I think you missed the part where I said that only other admins would initially vote. Hopefully admins can be trusted to be reasonable. 12.75.4.127 23:47, 17 August 2006 (UTC)
I missed the part where you were limiting the vote to admins. User:Zoe|(talk) 00:00, 18 August 2006 (UTC)
Not suggesting that I like the idea, but Meta does it this way. - brenneman {L} 02:29, 18 August 2006 (UTC)
The problem is that that doesn't scale, Aaron. As mentioned, en.Wikipedia would be looking at 20 confirmations per week, not 5 per month - and that number would only continue to grow as we add more admins. FCYTravis 02:33, 18 August 2006 (UTC)
Can people please actually read the proposal again, including the part that I've bolded (twice now). The bar is set very high for any opposition, and it will be a very rare occurrence to have any disapproval. There will be no problem dealing with 20/40/60 a week, as 99% of these will just go by with no need for any action. 12.75.3.180 02:46, 18 August 2006 (UTC)
If you think a majority (or even a significant minority) of admins are going to be opposed by other admins as being unfit for adminship, then we have more serious problems then can be dealt with here. 12.75.3.180
  • FCYT - don't pull me into this one! ^_^ I was not making "support" noises, just adding another information point.
    brenneman {L} 02:51, 18 August 2006 (UTC)
  • This is a red-herring. We've debated this before. It didn't fly then, it won't now. If we let this run, this debate will engulf the rest of the discussion. Suggest it is either dropped, or someone goes and revives the old discussion elsewhere. --Doc 08:46, 18 August 2006 (UTC)
  • Been there, done that, didn't work. - Mailer Diablo 09:11, 18 August 2006 (UTC)

Defending one's life

Where do any of your proposals allow for the admin so targeted to respond to the accusations? User:Zoe|(talk) 23:38, 17 August 2006 (UTC)

Well, my thought was that if the admin can convince arbcom that either there is 'no case to answer', or that he's 'got the message, and will comply in future' then arbcom wouldn't desysop him (or throw him to the lynch mob under the other proposal).--Doc 23:51, 17 August 2006 (UTC)
But, where? User:Zoe|(talk) 23:59, 17 August 2006 (UTC)
I assume it would be under Statement by admin before the ArbCom under Tony's system, *and* their statement(s) in the Recall RFA in any system. Under the others, I don't think there's any stricture against the admin responding, just like in a normal RFA. Anyone who is going to be "annoyed" by that was probably set on voting against them having the bit, anyway. -- nae'blis 00:15, 18 August 2006 (UTC)
It's been a set part of every "filing" format since version 1.0, Admin Response, or some variant wording thereof. rootology (T) 05:15, 18 August 2006 (UTC)
I think this is a case for arbitrator discretion. Arbitrators are intelligent and can make up their own minds. The administrator must be notified (as all parties must be notified in any application for arbitration) and is able to respond by writing on the arbitration page, his talk page (if blocked) or in email to an active arbitrator or clerk. The recall vote would of course afford the administrator a further chance to defend his actions as an admin. --Tony Sidaway 00:24, 18 August 2006 (UTC)

Isn't being an admin supposed to be a selfless position to serve others without any pay? Why would losing it be compared to losing one's life? Anomo 02:12, 18 August 2006 (UTC)

That attitude honestly mystifies me too, but hey. rootology (T) 05:15, 18 August 2006 (UTC)
Oh, please. Are you people just looking for chances to attack other people? It was hyperbole, of course. User:Zoe|(talk) 22:58, 18 August 2006 (UTC)

By the way, *nice* name for a section. I need to see that film again. rootology (T) 05:15, 18 August 2006 (UTC)

Status now

  • Gatekeeper: B - Have ArbCom review requests and approve a 're-RfA' if appropriate.
  • Approval: C - 20% to 25% - Only de-admin if the community overwhelmingly supports it.

So, are we in agreement then that this is most prefered and desired? User files on the ArbCom page, standard 4+ approval to accept, and it becomes a Recall RfA (where if the community overwhelming concensus is to remove, desysop occurs after b'cat review), or at ArbCom discretion it becomes a proper ArbCom review instead? rootology (T) 20:00, 18 August 2006 (UTC)

So, basically, you're proposing the current system, except for the fact that the ArbCom will simply refuse to place people for confirmation votes because of what happened with Stevertigo. You might as well just drop it; the ArbCom already has the power to desysop people and this proposal adds nothing to that. Kelly Martin (talk) 21:19, 18 August 2006 (UTC)
Uh... reading that ArbCom case they decided to let the community decide, and they did. What exactly was bad there? rootology (T) 21:23, 18 August 2006 (UTC)
Actually, the community rebelled - declined to decide - and told arbcom to do its job. --Doc 21:33, 18 August 2006 (UTC)
Uh, where is this? I'm not seeing it from scanning the ArbCom for this or the RfA. All I see is "The community has made it pretty clear that they wish us to deal with the matter. Therefore this case will be reopened." from Teresa, but unless I'm issing the community making it abudantly clear in the RfA, where did this happen? rootology (T) 21:48, 18 August 2006 (UTC)
I think you're seeing it. Arbcom reopened the case, because the community were not happy that arbcom had sent the matter to RfA. I'm not at all sure that their was any support for a reconfirmation process in that case.--Doc 23:08, 18 August 2006 (UTC)
The actual vote is at Wikipedia:Requests for adminship/Stevertigo1 and speaks for itself. There was a widespread feeling that the AC was shirking its responsibility to deal with the matter themselves. That was in the early days of the AC and they weren't as sure of themselves as they are today. The Uninvited Co., Inc. 23:43, 18 August 2006 (UTC)
Only nine months ago, actually. The Arbitration Committee had been running for nearly two years. Other administrators, such as Guanaco, had already been sent for reconfirmation. The difference was that in the meantime RFA had become a much more rowdy, less reasonable place, and many of us had lost faith in its ability to reflect community consensus. --Tony Sidaway 23:52, 18 August 2006 (UTC)
I'd be fine with requiring admins to maintain 75% support and I think there's little point in going beneath 50%. Any vote which would be going 70-30 for deadminning someone which would then end with her not being deadminned would probably be seen as silly. Haukur 21:44, 18 August 2006 (UTC)
I think if you look to the prior discussions, you will find that requiring admins to obtain 75% is unacceptable to many of us. Indeed, you will never get a consensus for that demand. --Doc 23:08, 18 August 2006 (UTC)
In general, admins end up making enemies in the course of enforcing decisions and also tend to incur the enmity of those users who dislike admins as a group. A look at any of the RfB results can show how this happens. The Uninvited Co., Inc. 23:43, 18 August 2006 (UTC)
So, are we in agreement then that this is most prefered and desired?. Um, no. We are not even in agreement that there is a need for a change in process. User:Zoe|(talk) 22:59, 18 August 2006 (UTC)
Yes, there is no agreement here. --Doc 23:08, 18 August 2006 (UTC)
Ditto. Though I applaud the efforts to compromise on this policy, it's still not better than leaving it to ArbCom as it is now. Fagstein 06:48, 19 August 2006 (UTC)

Protection Policy

Should a strong reason for Admin recall not be Admins who violate the Protection Policy ? The policy clearly states Admins should not protect pages in which they are involved; violating such a clear rule like this should, in my opinion, be a reason for Administrator Recall.

This is because page protections are designed to be a "cool-down" period for pages undergoing fierce debate - therefore Admins who do so double the problem and harm Wikipedia as a whole.

Anthony 13:18, 19 August 2006 (UTC)

Egregiously violating the protection policy is a good reason for arbitration. --Tony Sidaway 13:33, 19 August 2006 (UTC)
Agree with Tony. Admins also involved in disputes of any sort on a given page shouldn't protect them, nor should pages that are protected be edited in general. rootology (T) 17:10, 19 August 2006 (UTC)
That isn't what our protection policy says. --Tony Sidaway 17:44, 19 August 2006 (UTC)
I was paraphrasing. The policy says, "Admins should not protect pages in which they are involved," and "Administrators should be cautious about editing protected pages and do so in accordance with consensus and any specific guidelines on the subject. In most cases, administrators should first raise the issue on the relevant talk page, unless a case of obvious trolling and/or revert warring, or blatantly unsuitable content." I was oversimplifying. They should (based on this) get someone else to protect if they are involved in a dispute or involved with the page, inlcluding "making substantive edits to the page (fixing vandalism does not count), or expressing opinions about the article on the talk page before the protection." rootology (T) 17:49, 19 August 2006 (UTC)
Administrators are frequently called upon to perform actions concerning users, articles and other pages, and some editors may disagree with their actions. Thus they become drawn into disputes. "In which they are involved" should be interpreted very strictly to mean something like "in which they have made recent content edits or expressed strong opinions on content on the talk page". An overbroad interpretation would make it impossible for anyone who has monitored a page to perform any protection action on it. --Tony Sidaway 17:44, 21 August 2006 (UTC)

Ranks above administrator

I have removed the following section because it doesn't make any sense.

Arbitrators or any current/future "higher than administrator" users (currently bureaucrats) may be called in this fashion for administrative recall, but their removal should it reach completion must be approved by Jimbo, or an official action as determined by and seen fit by them. Recall of other administrators, including anyone those with additional privileges/duties such as CheckUser and Oversight, will be as per this policy, and the final RfA decision will be binding with respect to their sysop privileges only.

There are no ranks on Wikipedia. --Tony Sidaway 17:40, 21 August 2006 (UTC)

I know, and I kept fiddling with that section to get it to make sense to me. There are, however, privileges and duties that don't have anything to do directly with the sysop bit, which is what this proposal is about. It's probably worth mentioning somehow that it wouldn't strip ArbCom status, for example. -- nae'blis 18:06, 21 August 2006 (UTC)

Muddle

If there is not screaming I'm going to smoosh everything on the main page into one proposal and we can work from that baseline.

I don't think I'd be too far wrong in that some of the people contributing to this page feel that no amount of tweaking will ever make this workable. On the other hand, there exists a cadre who believes that it will be acceptable to the larger community. What's the best way to proceed, attempting to respect objectors while still moving forward towards a viewing by a larger audience?

brenneman {L} 07:03, 25 August 2006 (UTC)

From experience of other policy proposals. When you've got as much initial opposition as you've got here, and as much basic disagreement between proponents, then you are dead in the water. New policies only seem to get consensus when, at the outset, most people are saying 'hey, that's great, why didn't we think of that before'. You can proceed to a poll if you want, but I'd say it would be a WP:SNOW. But maybe it's the best way of drawing a line under this. --Doc 08:05, 25 August 2006 (UTC)
I don't think it would help to "smoosh" these very different proposals into one, but do give it a go if you like. In my own view, the recent desysopping of dbiv demonstrates that the arbitration committee is willing and able to perform its function in this regard. --Tony Sidaway 22:21, 25 August 2006 (UTC)


I just found this today. Sysop Accountability Policy Anomo 04:52, 27 August 2006 (UTC)

Some of the old CSD-expansion proposals had a "if X is approved, this one is trumped". I'm not sure if that would help achieve parallel proposals here without a smooshihg, which I agree wouldn't be likely to be helpful. -- nae'blis 22:32, 25 August 2006 (UTC)

Admins who flout policy

There needs to be a policy for recalling admins who arbitrarily follow Wikipedia's Neutral point of view, Verifiability, Reliable source rules when it suits them and who resort to cabal tactics to create an artificial consensus. I have seen admins who rule against content when it opposes their political opinions, yet insist upon imposing POV material when it agrees with them Unless the community does something about this, it will continue to reflect badly upon Wikipedia.--68.45.161.241 21:41, 5 September 2006 (UTC)

Jimbo Wales says ignoring the rules is policy. Anomo 21:43, 5 September 2006 (UTC)
No editor can get away with the tactics described above for long. Having a sysop bit doesn't make one immune from Arbitration. "Ignore all rules" doesn't mean "Ignore policy". --Tony Sidaway 05:00, 6 September 2006 (UTC)
That confused me to say "Ignore all rules" doesn't mean "Ignore policy". I'm confused. Anomo 10:14, 6 September 2006 (UTC)
Exactly what it says. Ignoring rules doesn't excuse inserting biased content , for instance. Policy is policy. --Tony Sidaway 11:12, 6 September 2006 (UTC)
It has one sentence, "If the rules prevent you from improving or maintaining Wikipedia's quality, ignore them." And the word "the rules" links to "Wikipedia:Policies and guidelines". Anomo 12:03, 6 September 2006 (UTC)
Not all policy is written. Not all written policy is accurate. --Tony Sidaway 16:16, 6 September 2006 (UTC)
Do you have an example of the former? --badlydrawnjeff talk 18:14, 6 September 2006 (UTC)
It's probably not wise to risk getting bogged down in details here. The fact that we frequently have to update written policy to bring it up to snuff with current practice is enough to establish the truth of my claim. --Tony Sidaway 18:23, 6 September 2006 (UTC)
I don't see the harm in examining such "unwritten" policies. After all, you've provided it as an example. Rewriting policy to update with practice isn't really an example, it's rather an example of the fifth pillar, and how we change things so our actions match our policies, instead of acting rogue. --badlydrawnjeff talk 18:26, 6 September 2006 (UTC)
1.5 years ago, we didn't have CSD-A7 [1], but people still used the same basic idea to speedy things sometimes. They did so out of process, but they were still helping the encyclopedia. Many policies are de-facto policies before they're codified. --Interiot 18:30, 6 September 2006 (UTC)
The blocking policy, for instance, has been updated over time as blocking practises have evolved and the older encapsulation became obsolete. This is an excellent illustration of my earlier statement: Not all policy is written. Not all written policy is accurate. --Tony Sidaway 18:33, 6 September 2006 (UTC)
And 1.5 years later, A7 is still fairly controversial. The issue, of course, is "de-facto" v. "codified." That doesn't technically exist. Just because a slight consensus of people ended up codifying what many people were doing, but shouldn't have been, isn't evidence of an unwritten policy - it's evidence of people doing the wrong thing that was later codified as being okay. --badlydrawnjeff talk 18:38, 6 September 2006 (UTC)
All Wikipedia policy is de facto and subject to evolution. The codification frequently lags behind that evolution. To say something is "wrong" because it isn't written down is to miss the point of having an adaptive policy for administering a living, growing and highly successful wiki. --Tony Sidaway 19:07, 6 September 2006 (UTC)
Well, no. Heh. That's all I can really say about it. You know full well that it's not how it actually works. --badlydrawnjeff talk 19:21, 6 September 2006 (UTC)
So why do people get blocked for improving/maintaining Wikipedia's policy? --badlydrawnjeff talk 12:05, 6 September 2006 (UTC)
Because they don't know how to Ignore all rules and they mistake their disruptive actions for improvement of the encyclopedia. --Tony Sidaway 16:15, 6 September 2006 (UTC)
And there's the rub! As the "policy" doesn't really state 'how," as much as giving a subjective rationale. One person's improvement is another disruption, after all. You've effectively proven my point here. --badlydrawnjeff talk 18:14, 6 September 2006 (UTC)
Not really. Ignore all rules is one of those instances where the proof of the pudding is in the eating. Some people successfully apply it all the time with no problems; others struggle and unsuccessfully invoke it to justify their disruption. --Tony Sidaway 18:28, 6 September 2006 (UTC)
Then you're not really reading it. Or worse. When has it ever been invoked "correctly?" --badlydrawnjeff talk 18:38, 6 September 2006 (UTC)
If it's invoked as an excuse for an action that failed, that is a good sign that it has been invoked incorrectly. It may or may not ever be invoked correctly; that's immaterial as long as it is successfully applied. --Tony Sidaway 19:11, 6 September 2006 (UTC)
That might be the most circular logic I've read about it yet. It's becoming obvious why Administrator recall will never work - not because the community will never agree to it, but because the powers that be have so successfully hedged themselves against any legitimate criticism. --badlydrawnjeff talk 19:21, 6 September 2006 (UTC)
Ignore all rules ensures effectively that what you call the powers that be, the community and the encyclopedia share a common goal and are effectively as one. Those who are without cannot use the one to attack either of the others. --Tony Sidaway 02:38, 7 September 2006 (UTC)

Ignore all rules needs a lot of documentation to it rather than a sentence to make it clear. Anomo 19:07, 6 September 2006 (UTC)

If we added rules to the sentence they would be ignored, and would almost certainly be wrong. --Tony Sidaway 19:11, 6 September 2006 (UTC)
What's the difference between policy and rules on wikipedia? Anomo 01:00, 7 September 2006 (UTC)
Isn't it obvious? Policy is applied. Rules are ignored. --Tony Sidaway 02:35, 7 September 2006 (UTC)
So everything that has a policy page is a policy. Then the "rules" are...? Essay pages? Then there's WP:3RR that says it's both policy and a rule????? Wikipedia has "policies" and "guidelines." Maybe guidelines are the rules. Wikipedia:List_of_guidelines and included in them are WP:POINT, WP:BITE, and WP:UP -- all of these are treated as policy. So WP:IAR means WP:POINT, WP:BITE, and WP:UP can be ignored if it's to make the encyclopedia better? There are also processes, which I do see the IAR applied to with WP:SNOW. Anomo 04:25, 7 September 2006 (UTC)
Good thing wikilawyering is a no-no, because this whole discussion would be reduced via reductio ad absurdum to something akin to Captain Queeg's falling apart on the stand under cross-examination. On the other hand, I suppose that following the line of illogical above it's not really a no-no, it's just a rule so apparently it can be ignored. Unless it's a policy. Which even though written is also de facto which is, of course, a contradiction in terms. And this, WP:SNOW, is a beauty, "If an issue doesn't have a snowball's chance in hell of getting an unexpected outcome from a certain process, then there is no need to run it through that process." If the outcome is unexpected, i.e., outside the parameters for establishing a probability function then can we really say it doesn't have a chance? I'd love to see the odds worked up for that one. Ooh, wait, I know, we can invoke Heisenberg here, as the uncertainty principle takes into account unncertainty as a factor, so no outcome is unexpected as it's within the probability bound. Well, except that since uncertainty is expected, it's included in the probability bound, so that can't be it. I guess Wiki ain't quantum merchanics, after all.
Of course, if "unexpected" were changed to "desired" there might be a point to the rule/policy, de facto and yet written -- which of course, even though it must be followed can also be ignored upon the application of a subjective analysis of benefit. Or something like that. --Jim62sch 12:00, 9 September 2006 (UTC)
I congratulate my learned colleague on his mastery of the Chewbacca defense. --Tony Sidaway 03:06, 10 September 2006 (UTC)
Ah. Of course, were one to notice how closely it parallels the logic (illogic?) used in this section as a whole, one would recognise the ironic and satirical tone. Alas, I fear I shall find no fans of Juvenal and Horace here. •Jim62sch• 20:51, 11 September 2006 (UTC)
WP:SNOW is written up, though and it's used to close AFDs early so it has an organization to it. But the rest of WP:IAR isn't. A lot of edit warriors believe they're improving the encyclopedia by making it lean this way or that. I mean if someone believes that wikipedia would be better by doing this or that to an article even though everyone agrees and so that person edit wars to change it, that's IAR really. Anomo 13:44, 9 September 2006 (UTC)


...I'm taking that as a motion to close. KillerChihuahua?!? 12:41, 9 September 2006 (UTC)

Can I second my own motion?  ;) --Jim62sch 13:18, 9 September 2006 (UTC)

Tyranny of the masses

All the recall proposals I've seen have the flaw that thery are subject to tyranny of the masses where an admin is unpopular, but has done nothing that is actually wrong.

Removing an administrator should require actual abuse by that administrator, either long term or short term. We do need to look at what constitutes enough abuse to desysop, and probably at guidelines for emergency desysopping at some point, but any recall process that takes opinion into the slightest of consideration is a problem. - Stephanie Daugherty (Triona) - Talk - Comment - 01:46, 10 September 2006 (UTC)

Where's the pressing need for this process/policy?

WP:RFC, WP:RFAR and unilateral actions by Jimbo and the arbcom already provide methods for dealing with out-of-line admins. Notable offenders are constantly dealt with through these processes. Is there such a backlog at these existing processes of surly, reprobate admins awaiting community censure that a new process and policy, fraught with yet to unidentified risks and uncertainties, is required? No.

Despite the single proposed safeguard of a "gatekeeper function" and the mentioned but not elucidated prohibitions on "repeated recalls," grudge holders will still find ways to game this system to settle old scores, just as they have every other system here; and a system for which no real and clear need has been established. I see no need for this process, and so reason to support it. FeloniousMonk 04:00, 11 September 2006 (UTC)

In fact, wasn't someone just desysoped for abuse of Wiki policy? Seems to me the system works. Of course, I don't hold any grudges against any admins, so I might not be the best judge of this. •Jim62sch• 20:59, 11 September 2006 (UTC)
The one rare case doesn't explain the dozen that slip by. --badlydrawnjeff talk 22:04, 11 September 2006 (UTC)
Really? Specifically, which dozen would that be? I still have seen no evidence presented that there is a genuine need for this. FeloniousMonk 21:20, 13 September 2006 (UTC)
And tip my hand? Never. If it isn't obvious, though, it raises a number of other questions. --badlydrawnjeff talk 21:29, 13 September 2006 (UTC)
Which particular case are we talking about? I know of at least two very recent desysoppings, so they're not exactly rare. --Tony Sidaway 02:48, 12 September 2006 (UTC)
I must've missed one, then. So then it does appear that the system is working. (The one I was refering to was the admin who offered to post deleted (oersonal?) info at another websdite -- I forget the admin's name, sorry). •Jim62sch• 09:58, 13 September 2006 (UTC)