User talk:Lar/Accountability

caution about transclusion

hmm... people are transcluding this but I'm thinking of a change.

one change is that to be "in good standing" you have to yourself be in the "category"... those unwilling to face this don't get to call for others to face it. thoughts? ++Lar: t/c 11:02, 10 August 2006 (UTC)

Sounds fair. —Nightstallion (?) 13:03, 12 August 2006 (UTC)

a complaint about transparency

wouldn't accountability need that one can see your actions as an admin, i.e. compile listings with what you did, be it good or bad? Tobias Conradi (Talk) 13:31, 4 June 2007 (UTC)

Massive revision

I did a pretty big revision of this today. This section is for comments on it. I tried to lay out specifically how every part of the process would be executed. If you spot holes I want to hear about them and fix them. If you spot things you don't agree with, not so much. I'm still interested but it's my committment to the community, so I may take them on board or not, as I see fit. ++Lar: t/c 04:07, 15 December 2007 (UTC)

The modified RfC process with majority decision is the key to these recalls, I think. the supermajority RfA standard for exisiting admins is not a good standard. I like your process, although allowing individuals to petition for recall 3 times a year each seems excessive - if you are making that many bad admin actions, there should be no shortage of other people to make the requests instead. See also my brief page on recall. NoSeptember 13:29, 15 December 2007 (UTC)

As much registering an interest in this approach. I may steal Lar's but I like some of the NoSeptember notes as well. --Herby ^{talk thyme} 13:42, 15 December 2007 (UTC)

Thanks for the feedback so far guys. I think NoSeptember has an interesting approach of tying standing to experience with the area(s) at issue. What would you do if the recall request was based on multiple areas of perceived malfeasance? Require standing in all? any one? :) As for my choice of 3 times, it was random but intended to be permissive. 1 and 2 seemed too restrictive to me. ++Lar: t/c 05:00, 16 December 2007 (UTC)

NoSeptember's restrictions seem to be on the "must be a chicken to judge an egg" theme, and would tend to limit requesters to be other administrators. I don't like adding yet more power to admins that ordinary experienced users don't have. If we trust experienced non-administrators to choose adminstrators, bureaucrats, and arbitrators, surely we should trust experienced non-administrators to propose to recall administrators. --AnonEMouse ^(squeak) 18:11, 28 December 2007 (UTC)

I'd have to think about NoS's stuff more but I agree, I don't want to limit to just admins... that's sort of what the WP:AAA proposal does, at least for starters. It deviates wildly after initial certification or whatever, of course...

Hi Lar - You speak of the criteria for the petitioners. May I suggest that you put a time frame on that in such a way that it not only eliminates newbies, but also recognizes changes in behaviour in longer term members (e.g., no significant blocks for disruption in the past year, must have contributed x number of mainspace edits in past six months). Many currently well respected members of the community have had a bit of a negative history early in their career. There are also a fair number of editors who have been on our registers for a long time and may have participated actively in the past but now only on occasion. Thanks. Risker (talk) 17:46, 28 December 2007 (UTC)

Agree. --AnonEMouse ^(squeak) 18:11, 28 December 2007 (UTC)

Yes, let me think about that, I think it has merit. We should encourage reform and improvement so this really is a good point. ++Lar: t/c 00:09, 5 January 2008 (UTC)

Changed it to add "within the last 4 months" on the disruptive behaviour part. This acknowledges that yes, people do change. ++Lar: t/c 21:05, 7 January 2008 (UTC)

Mainspace edits

What's the use of the 100 "substantive article improvements"? Might this be too aggressive of a filter? bibliomaniac15 19:08, 28 December 2007 (UTC)

I personally like the idea, if it means what I think it means. What I hope it means is real article editing (TM and R), something more than vandalism reversion or spelling correction edits. For example, for the last few days, I've been looking at my watchlist and it's been full of a fellow editor using Wikipedia:AutoWikiBrowser to change - to — and similar things in quite a few articles on my watchlist; he does several a minute, and while I haven't counted, I suspect he would meet "500 article contributions" every 1 or at most 2 days. I suspect "recent changes" vandal fighters rack up similar numbers. Not to gainsay spelling correction or vandal fighting, they're useful or even invaluable contributions, after all, but what I hope we're really trying to get at is that the editor be at least a little "experienced" in what we are actually supposed to do around here, write an encyclopedia, and "500 article contributions/100 substantive article improvements" is sort of a measure of article editing experience. --AnonEMouse ^(squeak) 22:15, 28 December 2007 (UTC)

As someone who has racked up a few edits vandal fighting across wikis I find this interesting. While agreeing with AnonEMouse I can see the clause effectively preventing me from being part of the process? Lar's comments would be interesting here. --Herby ^{talk thyme} 11:30, 30 December 2007 (UTC)

I'm very open to suggestions here. What I want to prevent is some ED Lulzer coming in and using AWB or other automation to do 500 short to long dash edits just to qualify, but also I don't want to rule out long term serious vandal fighters. I have a lot of sympathy with the notion that if you're not contributing somehow that your words well OUGHT to carry less weight (I won't go so far as to say unless you have one FA, I won't listen at all)... ideas? I don't want this to be TOO complicated. maybe qualify under A or B or C ish kind of thinking? ++Lar: t/c 00:11, 5 January 2008 (UTC)

Maybe make it at least 1000+ edits (not hard), at least 500 mainspace, and at least two of the people endorsing recall need to have joined no later than a year after you did (19:08, 8 June 2005 it looks like for you), so two of them would need to have been at it since 8 June 2006. That would eliminate any games, wouldn't it? Lawrence Cohen 00:17, 5 January 2008 (UTC)

question

I followed this back from Mercury's new RFA. What would happen if an admin posted something on a recall page such as, "As of January 1 200whatever, I am open to recall per these standards. If I attempt to remove myself from this committed obligation, I pre-authorize the first Steward to see or be notified of this message with confirmation that I have removed myself from my recall commitment, to desysop me immediately, and I wave the right to get my admin tools back then by anything other than another full RFA." And if it then included some relevant diff linking to that statement. Would something like that even be acceptable? Lawrence Cohen 09:33, 30 December 2007 (UTC)

I find this interesting. My matrix contains an "inactivity clause" (something I feel rather strongly about). However when I put it in I did wonder whether I would ever "withdraw" it and whether I could or should put in a "this is binding" kind of rider - same sort of principle as I see it? Cheers --Herby ^{talk thyme} 11:27, 30 December 2007 (UTC)

Herby: I think ultimately, this all is about your word. To yourself, and to the community. It's not about rules and laws and enforcement. Is your word good? Saying "I really mean it" doesn't mean you actually mean it any more than not saying it. ++Lar: t/c 00:18, 5 January 2008 (UTC)

I thought about the idea of it either when Durova or Mercury first began their processes, and I asked then on one of their pages, but in all the chaos surrounding each I think it fell under the radar. If an admin did something like that, would it be completely binding, if wording to the effect of

"If I try to undo this, remove this, or anything else, desysop me later per this edit; I understand the consequences of it, even if I try to deny this later."

The real question I suppose is what would happen if the admin did try to deny it later--would that early authorization be compulsory, if they said it was? Would a Steward act on it? Lawrence Cohen 15:55, 30 December 2007 (UTC)

Hmmm. Would "my account was compromised when it made that edit" fall under the heading of "deny"? --Thinboy00 @182, i.e. 03:22, 4 January 2008 (UTC)

That would be a pretty thin argument, especially if the person made the posting during their RFA, perhaps? Lawrence Cohen 22:19, 4 January 2008 (UTC)

Lawrence Cohen: I see where you're coming from with that clause. The idea has merit but I personally would not want to be held to this recall thing forever. I think it is a great idea but I could see some combination of circumstances possibly arising (that I can't name right now for lack of imagination) that would make even ME want to withdraw. I say I won't withdraw during recall or to avoid recall... not never. Someone else could make such a promise I suppose. One practical problem, though, (which you are already alluding to) is that in my considered judgement (as a newly elected steward), the clause is unenforceable (just as the mechanics of WP:AAA are) because it requires judgement on the part of a steward of something other than "did this discussion reach consensus that represents the will of the community" or "did this process adhere to stated policy"... something no voluntary thing ever can. Especially if it involves promises made at some point in the past. ++Lar: t/c 00:18, 5 January 2008 (UTC)

Hi Lar, I hadn't thought of it quite from that angle. I was thinking more along the lines of, why wouldn't a voluntary resignation of adminship ever be honored, and I couldn't imagine a case where it wouldn't be. It was basically an idea for a crazy person to put themselves permanently on the hook for that: "If I ever take myself out of recall, desysop me if I stand by that decision." So Stewards or the community in general wouldn't honor a self-made request like that? I honestly never thought of it being the case of it not being binding because the community decided it wouldn't be, since the community can't stop someone from laying down adminship (or so I thought!). Lawrence Cohen 00:36, 5 January 2008 (UTC)

As I see this, it comes down to when do stewards remove permissions... there are three specific cases:

1. Someone reports that the community, following their defined and previously consensed on (hence "official") process (such as Commons:Commons:Administrators/De-adminship) has decided an admin should be a non admin.
2. A member of a duly constituted arbcom from a wiki that has one reports that the arbcom has decided an admin should be a non admin.
3. A person reports that they are an admin who wants to be a non admin.

None of those require anything like the amount of judgement a steward would have to apply to this situation ("well he said it back then and now he says he doesn't mean it but someone else says he said it back then so that counts more than what he says now"...). I won't do steward things here as it's a home wiki but I expect any other steward would not be willing to do that. (see Effeietsander's comments at Wikipedia_talk:Admin_Accountability_Alliance#Community_decision where I asked him to comment) 01:20, 5 January 2008 (UTC)

Thanks, this is all fascinating. I never really imagined that a person couldn't make a binding decision, ahead of time, towards themselves in this manner. The third option you list, though--the person who is an admin asking for their permissions to be removed--so, in practice, this is only granted in general when the person actually makes a proactive request. If I were to write, here:

"I commit myself to the recall process as detailed at User:Lawrence Cohen/recall at the date and time of this posting. If I attempt to remove, cancel, or substantially change that process, I pre-authorize any Steward notified of those actions to summarily desysop me. Lawrence Cohen 22:21, 7 January 2008 (UTC)"

Nudges

Some great discussion here, and I didn't notice right away. Sorry about that... if anyone ever thinks I'm not responding here feel free to nudge me on my talk after a day or two goes by :) ++Lar: t/c 00:19, 5 January 2008 (UTC)

Keeping a sense of proportion about this

I added a LOLcat to the page partly to remind myself that despite the maze of ifs, ands, wheretefores, the parties of the second part and so forth... that this is serious yes, but it doesn't have to be humorless. Tell me if you think that's detracting or distracting :) ++Lar: t/c 00:23, 5 January 2008 (UTC)

A stultifying read

Oh, my head, my beautiful head in ruins! I now remember why I've never asked what would be the specific process of admin recall: I hoped someone else would create criteria and methods for me. I read through your process and I blank out. Truth, I can't seem to follow yours from beginning to end at the moment. I absolutely agree that admins in this category need a delineated process to provide for an actual implementation and application of recall when necessary. And that it needs to be done and in place prior to problems with the admin. However, I think a uniform process for all admins would be preferable to everyone creating their own. I'm put in mind of those software EULAs or credit card agreements so complicated very few people read them thoroughly. Perhaps I missed it but is there a reason why a relatively firm set of criteria and a distinct process hasn't been ironed out for everyone who puts themselves in this category? Even a couple or three standards could be used, say "Admins open to recall using the "Lar Standard" criteria" or somesuch. "...using the Firsfron Standard". Yes, I know: if I want to use a simpler standard than this one, I'm welcome to do so. It just seems so... messy. I guess this is feedback. And, no, I don't have ideas for improvement. Helpful aren't I? Cheers, Pigman☿ 05:11, 7 January 2008 (UTC)

I agree it's stultifying! I had a simpler one from when I first wrote it till about 2 months ago, but one that I thought was too ambiguous after what happened during Durova's and Mercury's recalls. So I started trying to nail everything down. Every time I said "and after that we have an RfC" or whatever I realised I had to explain what exactly that meant. By the time I got done, it was big. But I don't think it has any loose ends at all that I could find. Other people have simpler ones. I just worry that when someone gets to the "and now it's RfC time" no one will know exactly what that means. But I agree simpler would be more desirable. I think a few other ones are turning out to be used by more than one person as well, so that's a promising sign. Having one standard one takes things away from the "and it's voluntary" style Aaron and I were going for. Having a few standards might be good though. ++Lar: t/c 21:02, 7 January 2008 (UTC)

"Re-confirmation" and discretion

This page describes a process in which the recall subject may "choose to take the matter to ArbCom". When the arbitration case ends, or if ArbCom declines to take the case, "that concludes the matter as far as [the subject is] concerned". Alternatively, the subject may seek community approval at a modified RfC at which "no consensus == no change", and only users "qualified to vote in an ArbCom election" will count toward the total.

In my opinion, these scenarios are not fully in keeping with the spirit of the recall commitment. Wikipedia:Administrators open to recall says that category members are 'willing to stand for "re-confirmation" of adminship if a sufficient number of editors in good standing request it.' The term "re-confirmation" is most naturally interpreted to mean a repetition of the original confirmation at RfA: having stood for confirmation of adminship before, the member is willing to do so again. Accordingly, although the venue for re-confirmation is currently "an open question, to be decided by the entrant", it seems to me that no matter what venue is chosen, the standard for re-confirmation should be the same as for initial confirmation. Then we can most justifiably call the result a re-confirmation.

To be clear: the recall process proceeds in two stages, petition and re-confirmation. In the first stage, petition, the recall subject is granted discretion over "The number of editors, their standing in this project, the good faith of the request, etc." This stage acts as a sieve to filter out insignificant requests, so that the community is not burdened with re-confirmations whose success is foregone. But once this initial hurdle has been passed and the re-confirmation stage begun, I would expect the standard, meaning particularly (1) the eligible participants, and (2) the threshold, to be that of RfA.

Specifically, regardless of the venue chosen for re-confirmation (e.g. RfA, modified RfC), I would expect, just as at RfA, that (1) any Wikipedian with an account is welcome to support or oppose, and (2) an RfA-level of support is required to retain adminship, and no consensus means no confirmation. Obviously, taking the matter to ArbCom, or undergoing an RfC where "no consensus == no change" and only users qualified to vote in an ArbCom election are counted, would not meet these expectations. While I acknowledge the discretion granted to category members regarding stage-1 criteria and stage-2 venue, I think the recall commitment is best interpreted to include a non-discretionary stage-2 standard tied to that of the original confirmation at RfA. Tim Smith (talk) 22:08, 7 January 2008 (UTC)

I'm not sure I agree with all that, but I have also been worried about that ArbCom thing. See, I suspect there is a non-negligible chance ArbCom may decide not to take the case, stating "this is not what ArbCom is for", in which case, what, the admin stays? At least personally, when I signed up, for me, it meant I would step down if the community no longer had confidence in me. While the ArbCom is a great place to decide if I misbehaved enough for me to take the mop away from me by force, I'm not sure if they're the best way to gauge community trust. For me, if six respected editors complain about me on my talk page, if I don't agree, I'll go to something like RFC or RFA, not ArbCom. --AnonEMouse ^(squeak) 22:17, 7 January 2008 (UTC)