Wikipedia talk:Administrators/Administrator Accountability

From Wikipedia, the free encyclopedia
Jump to navigation Jump to search


I may have a reason to be biased, but I don't like this policy. First of all it doesn't seem to really say change anything — it just moves stuff around and reasserts existing beliefs. Second, the first rule is far too strict, as many official Wikipedia policies can be and often are broken, and are intended more as suggestions. If someone unintentionally incorporates too much POV into an article, that's breaking policy. If an administrator deletes a page accidentally while attempting some deft rearranging of articles, the page should be undeleted and the administrator reprimanded, but certainly not de-admined. We need more common sense applied and less blanket statements. Derrick Coetzee 06:56, 26 Aug 2004 (UTC)

<AOL> +sj+ 08:49, 26 Aug 2004 (UTC)
I agree with Derrick. Johnleemk | Talk 18:45, 26 Aug 2004 (UTC)

In principle this is a very good and important proposed policy. Two things should be changed though, as a safeguard against abuse: (1) measure 1 is too broad and should be restricted to those policies that concern admin powers. (2)If the request receives a vote of no confidence, it should not be referred to ArbCom. This automaticity would make abuse of this policy more attractive and lead to unnecessary bureaucracy in cases without merit. Concerns have been raised about "lynch mob" behaviour. IMO this need not be the case: if it becomes a problem, we should raise the number of votes required from simple majority to 2/3 majority, or "consensus" (i.e. 80%). A further safeguard could be introduced, e.g. the number of requests for admin review should be limited to one per month per user (not admin). But the principle behind this proposed policy is important: admins are approved by all Wikipedians to whom they should be accountable, and this power should be retained by Wikipedians and not be transferred to some other authority. At present, when admins are approved, it is an expression of trust that is not based on experience with how they act as admin. That's less than ideal, and this review process would improve the situation. - pir 09:37, 26 Aug 2004 (UTC)

On the need to move RFROAA and RFDA to their own subpages[edit]

I don't think this is necessary; they would be used so little that they can happily serve as part of RFC. I do like the idea of formalising the process, however. I also raise the question: is RFDA even necessary? The Arbcom could impose de-adminship as part of a penalty for really bad misuse of admin priveledges. -- Grunt (talk) 02:50, 2004 Aug 26 (UTC)

I think the review page would be useful; we need to not only be accountable, but to be seen to be accountable. I also think that this would separate out complaints related to admin abilities from general user complaints. Making personal attacks for example, would have nothing to do with admin accountability - it might be part of a RfC, but has no relevance to a person's use of admin abilities. Keeping the pages separate would allow us to concentrate only on those issues related to administration. But I don't see the need for a de-adminship page as well. Any removal of Admin abilities would have to come from the Arbcom, and all they would need is the information on the RFROAA. -- sannse (talk) 11:55, 26 Aug 2004 (UTC)
I'm not saying that the review page won't be useful; RFROAA should just stay a part of RFC. RFROAA would not be used often enough to warrant being its own page. If it ever reaches that point, we're doomed :/ -- Grunt (talk) 14:20, 2004 Aug 26 (UTC)
I suppose my main worry is that the pages would just get lost in the general RfC noise. With a separate page it will be visible and open, and easy to see how much (or how little) of a problem admin actions are. It will also help those who wish to keep a closer eye on things here, without having to wade through pages of RfC. It will be especially useful to keep a separate page if procedures in using the page are significantly different from using RfC -- sannse (talk) 15:06, 26 Aug 2004 (UTC)
I have to agree that I think a separate page would be better. Even if it is not used often, it will serve to show Wikipedians that admins are held accountable. If it is kept in RfC, people might not notice that actual admins are being questioned and actions are being taken to prevent any abuse of power. Skyler 02:53, Aug 27, 2004 (UTC)

subpages for each dispute[edit]

If it is decided to split WP:RFROAA back into a separate page, should it continue to use subpages for each dispute? I believe the style of user conduct disputes and admin conduct disputes (and VFD, for that matter) has its advantages:

  1. It keeps the main page simple and easily readable—an index rather than an entire book. Compare Wikipedia:Requests for comment#General user conduct to a randomly-chosen old revision of Wikipedia:Conflicts between users to see what I mean.
  2. Using subpages keeps the edit history of both the main index and the specific requests usable.
  3. Archiving and linking to subpages is easier; if a link to a prior review is needed there is no need to root through old revisions of a page that, if used as the sole area of debate, will have a massive edit history.
  4. The RfC templates keep the discussion somewhat useful; if used properly they produce (in theory) a straightforward document of the dispute, rather than simply giving flamewars a new arena. Again, compare old revisions of Wikipedia:Conflicts between users to various RfC subpages.

Anyway, format and organization are largely uncontroversial and easy enough to change if they should become so, so they aren't major issues, but it would be nice to decide this in advance. —No-One Jones 04:24, 29 Aug 2004 (UTC)

Lynch Mobs[edit]

I worry that some of these procedures skirt too close to quickpolls again. I think one of the things we learned from quickpolls were that the Wikipedia community can be quite like a lynch mob. As such, I tend to think democracy is not the proper procedure for judicial decisions. Perhaps a better solution would be an expansion of the arbcom that would allow them to move faster - say, double the size without doubling the number of votes needed for decisions. Snowspinner 04:40, Aug 26, 2004 (UTC)

I disagree with your view. Wikipedia gives any user (even any visitor) the power to modify the most important part of the whole project, i.e. the encyclopedia itself. The assumption that most people will exercise this power responsibly and with skill has served us well, and we've developed effective means of counteracting those who abuse that power. I see no reason why the same would not be true for judicial decisions. Wikipedians should have the right to "recall" abusive admins, (in fact since Wikipedians are the victims of abusive admins, they are best to judge who is abusive) and we only need to think of ways to counteract abuse of this right to make it work. - pir 09:55, 26 Aug 2004 (UTC)
Editing articles is very different from policy decisions - Wikipedians may not single-handedly alter policy, which is why we're having this discussion. Furthermore, we don't have a democracy regarding editing articles. We have an anarchy that eventually moves towards consensus, or, at least, towards a default position. I don't think that method is going to work for decisions like this etiher. Snowspinner 15:44, Aug 26, 2004 (UTC)
That's a strawman. I did not say that any Wikipedian should be able to single-handedly alter policy etc. My point is that we need to look at where power lies (in this case, the power to review sysops) and I'm saying power should ideally lie with Wikipedians as a whole, rather than transferred to another authority like ArbCom, as long as this is technically feasable, practical and there are no other serious problems with it. Obviously some things cannot and should not be done by all Wikipedians, like running the servers or accepting donations to Wikipedia. This is not the case for reviewing sysops, which is technically feasable, practical, and there are very good reasons why this should be done by Wikipedians collectively: Wikipedians collectively are much better placed to judge sysops that abuse their powers because they are most at risk, compared to ArbCom which is not necessarily representative (many members are sysops themselves and can defend themselves much better than your "common" Wikipedian). You have not responded to this last point. Please also explain why you think this would not work for decisions like this as you say above, and why safeguards (like more stringent requirements for majority, e.g. 2/3 or 80%) could not make it work. - pir 09:26, 27 Aug 2004 (UTC)
I'm sorry, but I'm just unable to get behind anything that tries to make a distinction between "sysops" and "common" Wikipedians in terms of interests or judgment. You become a sysop because you're a respected and stable member of the community. You become an arbitrator because you're an exceptionally respected and stable member of the community with what the community sees as especially good judgment. I really can't accept a line of argument based on the idea that the arbcom is somehow unqualified to make certain decisions. Snowspinner 13:47, Aug 27, 2004 (UTC)
Of course what you say is true for almost all sysops. The thing is that some people (no matter how respected they are in the first place) have and develop a taste for making use of their power against others. That's only human - I'm not even saying that I'm not like that. So to make it easier for admins not to fall for that, it would be good to make them accountable to all Wikipedians. The way I understand this policy, it is not about having a whole lot of votes (the need for which is only rare at present) - on the contrary: ideally the policy will re-inforce a culture of admins using their power with caution. I'm not saying that ArbCom is unqualified, I'm saying that Wikipedians collectively are better qualified because they are more likely to be the victims of rogue admins. You did not reply to my question why you think it would not work if we have appropriate safeguards. - pir 19:38, 27 Aug 2004 (UTC)
While I'd like to see the Arbcom expanded, there should also be a way to review an admins actions at a lower level than that of the Arbcom. They shouldn't be asked to rule on minor transgressions, any more than they should rule on every incident of vandalism or personal attack. -- sannse (talk) 11:55, 26 Aug 2004 (UTC)
Fair enough. But I think if the transgression is too small for the arbcom to rule on, it's also too small to justify sanction against the administrator for beyond "Hey. Don't do that." Which RfC already provides. Snowspinner 15:44, Aug 26, 2004 (UTC)
I think there is a level or two between "don't do that" and Arbcom, just as there is for vandalism, personal attacks and other editing problems. We also need to be able to see any pattern of minor transgressions from a particular admin - one or two may be dismissible, but if it is a daily event then we have a problem. But the main reason I'd like to see admin actions separated from RfC is visibility. I would like it to be clear that admins have to stick to the rules, and that we take it seriously if they don't. I think a clear and separate procedure will help to demonstrate that - and hopefully to also demonstrate that mostly we do stick to the policies. -- sannse (talk) 23:28, 26 Aug 2004 (UTC)
I totally agree. Visibility. blankfaze | (беседа!) 00:13, 27 Aug 2004 (UTC)

A few comments[edit]

Measure I: "Any explicit violation of any official Wikipedia policy, specifically policies guiding administrators and the powers they hold"…

Is this clear that it relates only to those policies relating to admin abilities? (and should it? I think it should.)

Measure II: "If the majority votes no confidence, then the action shall be reversed and the matter referred to the Arbitration Committee for a decision as to possible disciplinary action against the admistrator in question. "

I'm not sure that this should happen in every case. For example - in the case of a single misapplied block, perhaps the block should be reversed and it simply noted that the action was wrong in this case. If a pattern of misapplied blocks becomes evident, then that might be a time when further action might be needed.

Measure III: "That Wikipedia:Requests for de-adminship be a single, stand-alone page instead of a redirect to Wikipedia:Requests for comment."

As I said above, I don't think this is necessary. Any de-adminship would be an Arbcom matter, and related to serious, extensive or repeated violations of policy. In cases such as this, the correct place for the complaint would be Requests for Arbitration (with Requests for Review pages listed as evidence)

-- sannse (talk) 11:55, 26 Aug 2004 (UTC)

Measure I appears to be a statement or definition, rather than a measure. (added by Pcb21 13:43, 2004 Aug 26 (UTC); not signed)
I'm thinking thus:
  • Measure I only applies to admin actions. Other bad actions go through the normal channels.
  • The voting procedures described in Measure II might work better if it were more like the votes of the Arbcom; i.e. if it is determined that an action is necessary, actions should be proposed and perhaps voted on.
  • Measure III is subsumed by Measure II as a possible action resulting from an RFROAA.
Comments? -- Grunt (talk) 14:18, 2004 Aug 26 (UTC)
That all sounds good to me. We would have to see whether proposing variations of actions works easily - it's been something that has slowed the Arbcom considerably. -- sannse (talk) 15:06, 26 Aug 2004 (UTC)
My main problem is with the seemingly common viewpoint that Measure I should only apply to admin policies and admin actions. Don't we think that Admins should be responsible and held to a higher standard than the rest of all Wikipedians in order to keep their admin status? The measure states "Any explicit violation of any official Wikipedia policy". Why should an explicit violation committed by an admin go through normal channels? It seems like this would be cause to review their adminship. If we don't hold admins to a higher standard as we do with politicians, than we are cheating ourselves of having completely reponsible admins that adhere to all areas of Wikipedia policy, rather than just admins that do a good job in their admin actions. Skyler 03:11, Aug 27, 2004 (UTC)
My view is that admins are subject to exactly the same rules as any other user. But those are covered by other policies - none of which exclude admins. An example - if an admin were to start vandalising pages, they should be blocked by another admin, just as any other user would be. Now, in reality - that block could easily be removed by the blocked admin, but that would be an abuse of their admin abilities and would come under this rule. So I don't see a need for the original problem to be covered by this policy. However, if an admin reached ArbCom for serious violations of any policy, I would guess that the committee would seriously consider whether they should remove admin rights. -- sannse (talk) 13:45, 27 Aug 2004 (UTC)

Another proposal[edit]

Since the ArbCom is not really representative, sysops should be made accountable to the larger community. We should allow a certain minimum number of people to file a motion of no confidence in a sysop, and then use a secret voting process like for the Arbitration Committee (everyone who is here for 3 months can vote). If the sysop then has less than 60% support, he should be desysopped. Gzornenplatz 12:33, Aug 26, 2004 (UTC)

Perhaps there could be other possible consequences set forth by the original motioners; e.g. "I motion that Grunt be forced to unblock those he has wrongfully blocked and publicly apologise to those people." -- Grunt (talk) 14:13, 2004 Aug 26 (UTC)
Yes, that may be an option, but desysopping must be a real possibility. The current de-facto life tenure for sysops makes some of them behave in an arbitrary and autocratic way. Gzornenplatz 14:23, Aug 26, 2004 (UTC)
If a sysop is being really bad, the community can just propose that they be desysopped. If it passes, so be it. -- Grunt (talk) 14:30, 2004 Aug 26 (UTC)
If you mean it's already possible, I disagree. It's important that the voting is secret, as some people might not dare to openly oppose a sysop for fear that, if the vote fails, they will be even more on that sysop's bad side. Gzornenplatz 14:48, Aug 26, 2004 (UTC)
I meant that desysopping _would be_ a real possibility, not that it's already possible. And it should not be an issue to be on a sysop's "bad side"; after all, if they do anything really bad, just take them up on RFROAA :)
It is an issue if RFROAA doesn't work. Gzornenplatz 15:29, Aug 26, 2004 (UTC)
If it's an out-of-control dispute between two people, it can be taken to the arbcom. -- Grunt (talk) 15:30, 2004 Aug 26 (UTC)
The arbcom doesn't work either. Some cases take forever to come to a decision, others are not even accepted in the first place. Gzornenplatz 15:47, Aug 26, 2004 (UTC)
That's the point. You act as if sysops are mean on purpose, and that they do it on a wide scale. Admins aren't even appointed without a clear consensus — for example, recently Lucky 6.9 wasn't appointed even though there was a majority of support votes. Wikipedian users don't appoint admins unfairly. Why should the process of desysoping be so secretive when the process of appointing sysops is so open? If the admin behaves unfairly toward the user after the process fails, the issue can be brought up again and again. If there's a case, nobody's stopping you from presenting it. We're a self-patrolling community. If the admin plays dirty, the other admins notice. Desysoping is a task that should not be taken lightly, which is why we have arbitration and why it is long. I'd rather have justice done than to turn off a good user accidentally because of a rushed decision. Voters can be unfairly influenced by trolls and sockpuppets, you know. Johnleemk | Talk 11:40, 27 Aug 2004 (UTC)
Most admins are OK, but abuses happen. Part of the problem is that there are still sysops from the early days when standards were much lower, and some of those would probably not be accepted today if they had to apply again. See above for why it should be secret. And self-patrolling doesn't work currently. I can present a case, but nothing happens. The admins are in control and tend to protect each other. Only one or two have ever been desysopped, although there were several who have repeatedly exceeded their authority. This lack of accountability has to produce resentment. Gzornenplatz 12:07, Aug 27, 2004 (UTC)

Measure I has been revised[edit]

I'd like to note that Grunt's proposed change to the wording of Measure I has been implemented. Measure I now limits violations to ONLY policies guiding administrators and the powers they hold. blankfaze | (беседа!) 21:01, 26 Aug 2004 (UTC)

As I have stated above, I disagree with this revision. I believe admins should be held to a higher standard. Skyler 03:16, Aug 27, 2004 (UTC)
It may not be exactly what you wanted, but I'm trying to find a middle ground in hopes of appealing to the most people. I'd really like this proposal to pass :-). blankfaze | (беседа!) 03:25, 27 Aug 2004 (UTC)

Measure II has been revised[edit]

In response to concerns of Sannse and others, Measure II no longer recommends referring every action to the Arbcom. Now, administrators would only be referred to the Arbcom for disciplinary action after *3* majority votes of no confidence. blankfaze | (беседа!) 21:14, 26 Aug 2004 (UTC)

We're thinking that the three count would apply towards individual actions; i.e. if a really bad admin did a lot of really bad things at once, each really bad thing would count as an individual action and they would be sent to arbcom right away (after vote of no confidence passed); only one vote would probably be required for the entire batch of related actions, though. -- Grunt (talk) 21:15, 2004 Aug 26 (UTC)
    • Right. But then if an admin say, made a questionable block. That one-time occurrence would probably not merit immediate Arbcom action. So we're hoping this change will find a middle ground. blankfaze | (беседа!) 21:27, 26 Aug 2004 (UTC)
I think there's a serious problem with Measure II the way it is now: trolls could just continuously request these votes. A limit needs to be put on how often one can do that. Or one could request that a vote is only taken when the request is supported by one or two Wikipedians. It is important that these votes remain exceptional rather than every-day. pir 21:00, 27 Aug 2004 (UTC)
Someone mentioned somewhere that each user could call a maximum of one vote per month. I like that idea; if it was a serious problem, someone else could just bring it up. -- Grunt (talk) 22:26, 2004 Aug 27 (UTC)
Adding it. blankfaze | (беседа!) 22:45, 27 Aug 2004 (UTC)

My own concern[edit]

I am very active in maintaining Wikipedia:pages needing translation into English. There are a lot of times when this results in my speedy-deleting garbage articles in foreign languages. When I'm in doubt, I put them on Vfd, but it's a judgement call every time. So far the clear consensus is that I'm erring on the side of caution. Effectively this seems to say that if someone feels I make one bad judgement call in the other direction, I have to go through some kind of process related not to reversing my action (which I would always willingly do for any reasonable complaint) but over my qualifications to be an admin. I assume dozens of other admins are in parallel situations on other parts of Wikipedia. Our role requires us to take actions. Sometimes we will be wrong. It seems to me that as long as we unresistingly reverse ourselves when we are wrong and we are not often wrong, that is all that can be asked of a human being.

I will admit that I haven't read this closely. I will do so only if it is actually solidified and brought up for a vote. Until then, frankly, I'm just hoping it will go away. In about a year on wikipedia, I have not witnessed a single case of an administrator doing anything significantly inappropriate in his or her role as an administrator (some are hell as editors, but they still will be if they are no longer admins). I think that regardless of probable good intent, this is (1) an attempt to solve a problem that, to the best of my knowledge, does not exist and (2) one more tool that will allow trolls to play with process and tie up the people who are actually trying to do real work. -- Jmabel 00:25, Aug 27, 2004 (UTC)

  • Whilst I of course respect your opinion, I disagree. I think that Wikipedia needs to have this policy or at least something similar. Right now, administrators are rarely, if EVER, sanctioned, chastised, anything! There's barely even systems set up. We need checks and balances! I'm tired of hearing users complain about cabals and how administrators can do pretty much anything they want. It's true! If I wanted to, I could go around doing about 1000 stupid or destructive things before anyone would talk about desysopping me. We need a system in place, even if it's not a huge problem, if ONLY to be fair. blankfaze | (беседа!) 00:59, 27 Aug 2004 (UTC)
I agree with you both. We have to be careful that every minor judgement call doesn't become a cry of "policy violation!" We all make these decisions each time we speedy delete or block someone for simple vandalism - we need not to be afraid to make those judgments, and not to be risking days of hassle and accusations every time we do. But on the other hand we also need to change the current climate where there is little or no analysis of admin actions, and anyone who dares to question a decision is automatically presumed to be trolling or worse. Or (if that's a bad analysis of the situation) where that is perceived to be the current climate. What we need from this policy is a balanced middle way, a sensible policy that makes us accountable but does not restrict our daily, legitimate work. Jmabel, please do read this carefully and help to find a way to find this balance - the more of us that do so, the more likely we are to come up with a policy that will actually work. -- sannse (talk) 13:46, 27 Aug 2004 (UTC)

Premature deadline[edit]

This is a serious issue to set an arbitrary deadline, especially during summertime, time of vacations. IMO the deadline must be selected (moreover, discussed and voted) only after an initial discussion fades out. Mikkalai 08:56, 27 Aug 2004 (UTC)

What is more, I suggest that a deadline for the vote must be set only after significant evidence will be gathered of the visibility of this proposal. E.g., at least two dozen of users comment on this page. Otherwise it might well be a bolshevik-type putsch by a handful of bureaucracy zealots. Mikkalai 09:08, 27 Aug 2004 (UTC)

Although accusations of bolshevism are a bit exaggerated, I'd support an extension of the deadline, and more debate about the proposed change before it goes to vote. pir 11:50, 27 Aug 2004 (UTC)
Only a bit. It is a good historical example when a determined minority in quest for everyone else's good can screw up everybody else big time. Mikkalai 18:13, 27 Aug 2004 (UTC)
I agree. This discussion has been publicised on the Village Pump less than 48 hours before the close of discussion is scheduled. Added to only a weeks' voting time this is wholly inadequate at the peak vaction time. Here in the UK we have a national holiday this weekend and about a quarter of the population is reckoned to be away from home at least part of the time! I suggest at least an additional 2 weeks for discussion to allow adequate publicity. -- Arwel 21:22, 27 Aug 2004 (UTC)

This is sheer hysteria. There's ample precedent for the person making a policy proposal to set the voting timetable, and this voting timetable is a pretty standard one. Snowspinner 20:40, Aug 27, 2004 (UTC)

I don't think the underlying concern is "sheer hysteria", although such exaggerations are not surprising in response to over-the-top rhetoric about bolshevism and zealotry. I would say that the proponents of the policy are trying in good faith to allow some time for discussion before the "vote", and also working to advertise the discussion appropriately.
"about bolshevism" is not rhetoric; it is a historical example, known to all world, of twisting opponents' arms and screwing up a generally good idea. I could have used another example, but it would be less recognizable. Mikkalai 04:58, 29 Aug 2004 (UTC)
The problem is that we can't act too quickly for people to respond, but if we dally too long the ideas will be ignored and grow stale. I would say that the actual timetable used, five days for discussion and modifications before starting, is just a little short for a proposal of such significance. So I agree with Pir's sentiment, and would suggest that we extend discussion for a few days, and start the survey/poll on September 1st instead, to conclude on the 8th. --Michael Snow 21:22, 27 Aug 2004 (UTC)

mikkalai's objections[edit]


Watch for Parkinson's Law. Mikkalai 08:56, 27 Aug 2004 (UTC)

The amount of additional work must be compared with the importance of holding admins to account. pir 11:54, 27 Aug 2004 (UTC)
Did they go amok yet? Is it really time to start witchhunt? And who will hold to account those who want to hold admins to account? Mikkalai 17:53, 27 Aug 2004 (UTC)

After reviewing Wikipedia:Administrators, I cannot help but notice that the only abusive things admins can do what others can't is protection and deletion of pages, and blocking users. To keep an eye on these actions and to persecute this kind of abuse, the current practice is more than enough. I don't hear from "developers" and "board of directors" that they are tired of doing so. Mikkalai 18:27, 27 Aug 2004 (UTC)

Measure 1[edit]

Measure 1 is tautology: "explicit violation is abuse". Mikkalai 08:56, 27 Aug 2004 (UTC)

That misses the point: the crucial thing is precisely determining what is abuse and what is not. This policy is about whether this decision should be made by Wikipedians or not. pir 11:56, 27 Aug 2004 (UTC)
That's exacly my point:
  • The term "explicit violation": how does it serve to clarify the notion of "abuse"? Explicit is basically "readily observable". Why subtle, creeping, clandestine viaolations are not abuse? IMO they are even more abusive. And IMO this very proposal, if accepted will be an abuse.
  • Why "explicit violation" itself cannot be the base of pesecution? I have an impresion that measure 1 (and this whole numbering thing) was introduced simply to look the proposal more "solid" (two items is something not serious. 'Three' is the minimal bureaucratic number.) Mikkalai 17:49, 27 Aug 2004 (UTC)
  • Finally, if a violation is "explicit", why vote? Mikkalai 18:16, 27 Aug 2004 (UTC)
The vote is not to determine whether or not an abuse was explicit - it is to determine if there was an abuse at all. blankfaze | (беседа!) 21:02, 27 Aug 2004 (UTC)
See? You yourself are confused with your own language: The "explict" was "violation", not the "abuse". And you still didn't answer my question: what is the difference between abuse and violation in this context, and why measure 1 is not tautology. And why "clandestine violation" is not abuse. Instead, you prefer to spend your time on personal threats. Mikkalai 21:26, 27 Aug 2004 (UTC)

Measure 2[edit]

  • For reasons of visibility, multiplication of pages should be discouraged. It is a general policy of wikipedia that pages are split when the size becomes unmanageable. If we expect such growth of conflicts with admins, then something is inherently wrong, and no bureaucratic "Measures" will help. Mikkalai 08:56, 27 Aug 2004 (UTC)
We should not forget that the possibility of admin review votes will have a big effect on admin behaviour. The aim is not to hold lots of votes. The aim is to introduce safeguards against rogue admins and thereby influence their behaviour. pir 12:00, 27 Aug 2004 (UTC)
Precisely, pir. I think I shall add that statement in the introduction to the proposal. blankfaze | (беседа!) 21:08, 27 Aug 2004 (UTC)
What is "brute-force" about voting? - pir 12:00, 27 Aug 2004 (UTC)
Consider the difference between voting and mediation. Mikkalai 18:01, 27 Aug 2004 (UTC)
The whole history of democracy shows that only majors matter of policy are decided by all-people voting, referendum, etc. For everyday issues, power must be delegated and someone must always be pesonally responsible for decisions. Soviets is the best example of what will happen if for every issue the voting is required, with responsibility for bad decisions meing diluted, hidden, or manipulated. Mikkalai 18:01, 27 Aug 2004 (UTC)
Why does power have to be delegated when we have the practical means to retain it? De-admin requests will not be an everyday event. Soviets did not fail because they were democratic at the very beginning, they failed because they were taken over by the Bolsheviks and became totalitarian. I do not think that we can learn much by comparing Wikipedia to imperial Russia in 1917, but if anything your analogy supports my argument: we need safeguards to prevent abuse of power by admins. Suppose some totalitarian organisation, like Scientology or the North Korean government (the Dear Leader is known to use the Internet a lot!), decide to inflitrate Wikipedia. They give a twenty or fifty paid people the full-time job of contributing to Wikipedia. After several months, they'll be very highly respected and popular so they'll become admins without any problem. They'll also get their people in ArbCom. Once they've been installed they change their tune and take over. How would you prevent that if there's no way to de-sysop abusive admins via a popular vote? - pir 19:17, 27 Aug 2004 (UTC)
The policy must clearly state that it is applicable only if attempts to resolve the conflict amicably were performed. If we call for blacklisting of admins for each and every misjudge (declared to be abuse, by this poposal), can of worms will be opened. Mikkalai 06:30, 29 Aug 2004 (UTC)

Measure 3[edit]

  • "...Thus Wikipedia would have a simpler, more democratic process". A simpler, yes, a more democratic: no. When only 0.003% of wikipedians vote, this is a mockery of democracy and an abuse of the term. Mikkalai 08:56, 27 Aug 2004 (UTC)
I partly disagree. I have no idea where your figure comes from, but from my experience votes are quite representative. People who decide not to participate in votes probably have good reasons not to: they may be happy with the outcome of the votes ; they may not feel sufficiently informed to take part ; they may not feel sufficiently involved in Wikipedia to have an opinion. As long as every Wikipedian has the right to participate in the vote, and as long as these votes are visible and well advertised, there should be no problem. - pir 12:04, 27 Aug 2004 (UTC)
That's the point: "who decide not to participate". The problem is that people DO NOT decide not to participate: the people simply don't know what's happening. They simply have no time to keep their eye at politics. They are busy contributing (in their spare time, free of charge, mind you). And the history of bolshevism shows that a minority can really screw up any good idea, if they concentrate their efforts against the rest. Mikkalai 18:06, 27 Aug 2004 (UTC)
This should not be a problem if votes are well advertised. Abusive admins are a nuisance and a concern for all Wikipedians, so I really believe that a large (or at least representative) proportion of people will participate. In addition, and as I tried to explain above, a minority that concentrate their efforts to screw up things is far more dangerous if they are already in a position where they have more power. - pir 19:22, 27 Aug 2004 (UTC)

The Introduction has been revised[edit]

A section has been added to the introduction of the proposal. It reads: "The aim of this proposal is not to increase bureaucracy nor to persecute admins. Rather, this proposal aims to introduce safeguards against possible rogue and abusive admins and thereby influence their behaviour, and to formalise the currently murky and obscure process of holding admins accountable for their actions." blankfaze | (беседа!) 21:27, 27 Aug 2004 (UTC)

The discussion period has been extended[edit]

Due to concerns from multiple users, the period allotted for discussion has been extended by 48 hours. Voting will now begin at 00:00 UTC, 31 August 2004 and last until 00:00 UTC, 7 September 2004. blankfaze | (беседа!) 21:35, 27 Aug 2004 (UTC)

Great, although I would still definitely urge that it be extended at least one more day, specifically due to the point mentioned above that 30 August is a holiday in the UK. This would allow people who may be unable to participate in the discussion over the weekend to return from their holiday and still give them one day for any final input they have before voting starts. --Michael Snow 21:43, 27 Aug 2004 (UTC)
Dude, today is the first I've seen this proposed policy. How about extending the discussion period a but longer, say, until 7 September? -- orthogonal 00:35, 28 Aug 2004 (UTC)
The policy was posted on Goings-on and on the mailing list. There's not really much more to do in the way of publicity for a proposed policy. Perhaps if more would-be concerned Wiki-citizens were to follow the places where policy proposals are mentioned? Snowspinner 00:41, Aug 28, 2004 (UTC)
And the link to the place this was held is...? I could want to make a policy a two too. A fine crop recently. Mikkalai 00:49, 28 Aug 2004 (UTC)

A quick peek into who the proponents of the new policy are[edit]

I confess I decided to run a small experiment to get an idea how these measures are going to be implemented. I made a harmless change in the article: I replaced "Discussion on this proposal shall last" by "The authors of this proposal suggest that the discussion on this proposal shall last". Notice that the article requests: "Please do not edit this proposal", and I DID NOT edited the proposal iteself (Me being good at red tape :-). After being reverted, I re-reverted with the explanation for those who cannot read what they write themselves: "I did not edit the proposal, complying with the preamble. I edited the preamble. Why revert? I made the statement more precise, especially since it is disputable."

After that I've got the following threat.

Please cease your unwanted edits to Wikipedia:Administrators/Administrator Accountability Policy. It is my decision as to how long the discussion will last, as I am the proposer. It states clearly in the text "Please do not edit this proposal..." Anyhow, if you insist on continuing to make this edit, I will be forced to block you from editing for "disruptive actions" as per Wikipedia:Blocking policy. Thanks. blankfaze | (беседа!) 20:56, 27 Aug 2004 (UTC)

This guy evidently didn't bother to peruse the Wikipedia:Blocking_policy#Disruption before starting to throw threats onto me. I say, if such trigger-happy people will formulate the policy in wikipedea, and what is more, proceed to implement it, then you are in big trouble.Mikkalai 21:50, 27 Aug 2004 (UTC)

Okay, maybe everybody needs to take a nice big time-out here (24 hours at least, though I'm not threatening to block anybody). So right, I don't agree that Blankfaze would have been justified in blocking Mikkalai based on my reading of the policy, but at the same time I think he was acting in good faith and focused on trying to stop a rapidly developing edit war (the warning came after Snowspinner's second revert to the page). And Mikkalai is certainly not helping the situation by taking such a hostile attitude, as reflected in the posts on this page. As for the "experiment", perhaps Mikkalai would do well to consider some of the good advice at Wikipedia:Don't disrupt Wikipedia to illustrate a point. --Michael Snow 22:01, 27 Aug 2004 (UTC)
I thought didn't. I used a loophole in guy's text, and in a minor way. And your advice is very good. And it again illustrates my point that we'd better make existing rules visible, rather than invent new ones. Mikkalai 23:29, 27 Aug 2004 (UTC)
About making existing rules visible, rather than inventing new ones - I've long been of the opinion that we have way too many policy and other pages in the project namespace, and am always a little disturbed when people create even more. It's a form of instruction creep, and also makes it more and more impossible for anyone to actually learn and absorb our policies. Having fewer policy pages would make the remaining ones more visible and their contents more familiar to everyone. That's not really related to this proposal, but I'd like people to be a little more aware of the problem. --Michael Snow 23:57, 27 Aug 2004 (UTC)
Hear, hear! Geogre 23:44, 28 Aug 2004 (UTC)
Well, it is directly related to this proposal, which explicitely muliplies the number of policy-enforcing pages (Measures 2 and 3). Mikkalai 00:03, 28 Aug 2004 (UTC)
Well, as I see it, his edits were disruptive, as they might mislead a reader of the proposal into thinking that for some reason voting would start at any time other than the one specified. I do not believe his edit was necessary or even in good faith, for all he would have to do is raise the issue on the talk page and wait for me to respond. Whilst perhaps Mikkalai was being bold, I think that in instances such as this, where there are explicit instructions not to edit the page, that discussion is the wiser route. Anyhow, like Michael said, my warning was more of an attempt to curve an edit war before blocking became necessary, rather than to make a serious threat. blankfaze | (беседа!) 22:15, 27 Aug 2004 (UTC)
"explicit instructions not to edit the page" Yeah. If it were so, I wouldn't. It was "do not edit the proposal". And I didn't. My whole point is that such policy cannot be written in haste and put to vote discussed only by a handful of zealots. Mikkalai 23:29, 27 Aug 2004 (UTC)
blankfaze, who is entitled to give "explicit instructions not to edit [a] page", and under what circumstances? As far as I was aware, anyone could edit anything. Thanks. -- orthogonal 04:00, 28 Aug 2004 (UTC)
Well I've taken the time to work up this proposal, and I'd like to manage it myself. So I politely asked that others not edit it. I don't think that's too much to ask, really. blankfaze | (беседа!) 05:29, 28 Aug 2004 (UTC)
blankfaze, you politely asked, and then you threatened to use your admin powers to block him despite his having done nothing more than edit a page. Yes, you made the proposal, and yes, things are so politicized around here lately that -- despite the wiki spirit of community editing -- I personally wouldn't have edited it.
Your argument that Mikkalai's "edits were disruptive, as they might mislead a reader of the proposal into thinking that for some reason voting would start at any time other than the one specified. I do not believe his edit was necessary or even in good faith, for all he would have to do is raise the issue on the talk page and wait for me to respond" is entirely specious. Not only was no one going to be misled -- as all he was doing was adding a suggestion, but the argument that he should wait for you to respond pre-supposes that you and only you have a right to set the time of the vote. Again, given the politicization, I'd likely have deferred to the author's preferred time, but no one is obligated to do so.
But the page is not your property, and all Mikkalai wanted to do was suggest extending the editing period -- something you yourself later decided to do. Given that you are a sysop, the onus is on you to provide a better, more collegial example -- but instead you decided to take the trust the community reposes in you and use it to be heavy-handed and threatening -- and to a newbie, of all people. Not to put too fine a point on it, you were being a bully while sitting safely behind your computer screen, and expecting to get away with it because you're a sysop.
You claim, above, that "my warning was more of an attempt to curve [sic] an edit war before blocking became necessary, rather than to make a serious threat." That is disingenuous at best: You weren't trying to curb an edit war; as it was your own edit your were trying to maintain, you could have avoid the edit war that simply by compromising.
But on your own talk page, you admit that (emphasis orthogonal's) "I was sort of hoping to scare him into stopping his sillyness [sic]." You expected to be "scary", but you now claim that while you hoped to be scary, what you did wasn't a "serious threat"?
How is threatening to (indefinitely) block a user from editing not a "serious threat"? Not only is it a serious threat, the only WP sanction less severe than banning, it's also a threat that can only be made by a sysop. Rather than avoid being a judge in your own cause by finding another user or sysop to mediate, you threatened to take away a newbie's entire ability to edit over your own edit war.
And you've avoided answering my initial question: who is entitled to give instructions not to edit a page? You state you gave "explicit instructions not to edit", but I don't see where you get the authority to give "explicit instructions" to anyone. Or do you claim that as another perquisite if being a sysop.
You acted precipitously, you in your own words tried to "scare" a fellow user, you were a bully, and you abused your powers as a sysop. And now, rather than admit your mistake like a grown-up, you persist in characterizing Mikkalai's attempt to add a clause -- a clause merely turning a hard time limit into a suggestion -- to your proposal's preamble as "disruptive behavior".
You owe Wikipedia an explanation of just where you think you get the right to "explicitly instruct" the rest of us, and an explanation of how you justify threatening a user who was merely following the wiki spirit of editing boldly. And you owe Mikkalai, and Wikipedia an apology. -- orthogonal 07:40, 28 Aug 2004 (UTC)
Well, I think you're taking this entirely too seriously. He was not editing in good faith. He was trying to be an asshole. But, I don't have time to sit here and argue with you. I owe apologies to no one. So, file an RfC on me if you wish - otherwise, drop this, please. blankfaze | (беседа!) 15:13, 28 Aug 2004 (UTC)
blankfaze, I notice that you are a strong proponent of wikipedia: Blocking policy/Personal attacks, which would allow users making personal attacks to be blocked at the discretion of a sysop. Question: does calling another user "an asshole" constitute a personal attack, and if the proposed blocking policy passes, would you block a user for it? If so, for how long would you block him? -- orthogonal 20:33, 28 Aug 2004 (UTC)
Ortho, he didn't call me asshole; he said I was acting as asshole. (This should be read that "I thought he was trying to be an asshole".) I suggest yto read some law texts on slander. And by the way, this shows another danger of the recent flood of policies. I wouldn't entrust a fate of any admin to a (someone else's words) "lynch mob": how many lawyers are out there that have spare time to consider accusations with proper diligency?.
blankfaze: FUI, I was trying to be smartass, not asshole. And I was editing in good faith: I intended to stress that the date was set arbitrarily by the author of the proposal, rather than discussed and agreed upon. And I was careful enough not to edit the Proposal itself, to comply with the request of an autocratic author. And here's your problem that you'd rather not talk to people, but ban-ban-ban! them. (no man, no problem; Joseph Stalin's attitude)Mikkalai 21:11, 28 Aug 2004 (UTC)

You made a change that was against the traditional way of doing things, and justified it with a hysterical post about how the people who were proposing hte policy were Bolsheviks and how their policy constituted abuse. Since then, you have called blankfaze a zealot, a tsar, and said "good fuck to you." Sadly, the disruptive users policy, you are right, does not apply to you in this case - however, you are well past the limit of what is acceptable on Wikipedia, and I strongly advise you to back off. Snowspinner 00:05, Aug 28, 2004 (UTC)

Advise taken. But what 'traditional way' are you talking about? I was on vacation for three months, now I'm back and lo! there lots of new traditions around and still more under development! Of course I was a bit pissed off, and I apologize. BTW, there is nothing personal here. I never ever blocked anyone non-anon and never reverted more than twice per issue. So I am not afraid I will be de-admined or something. And I am bit worried of your polite wording "sadly". So you are a trigger-happy one, too? Given an opportunity, you'd be more than happy to kick my smart ass, yes? Are you working on a new policy to have this opportunity? You are saying I am hysterical. Smooth operators are way more dangerous. Mikkalai 21:11, 28 Aug 2004 (UTC)
I too have to question the idea of this being the "traditional" way of formulating policies. The traditional way is the wiki way... community editing of a proposal until a consensus version is found. The idea of an "author" of a policy is a recent development, and I'm not yet convinced it has a clear advantage over older methods - although I do appreciate why Blankfaze chose to use the more formal procedure, and I'm interested to see if it works. I think Mikkalai's "experiment" was ill advised - not because it involved editing the proposal page (although I would have personally chosen to discuss the issue rather than edit if I had decided to challenge the procedure) but because it was an edit designed to provoke a reaction, which I don't think was helpful. But I think the results of those edits reinforce the need for some sort of policy rather than otherwise. I feel there is a tendency at the moment for "arguments from authority" by admins, which really worries me. It's not so much any gross abuse of admin powers that concerns me, but rather the small day-to-day pressure to conform to what a small group sees as "correct behaviour"... the edits in good faith that are dismissed as vandalism and the valid disagreements that are automatically considered trolling simply because they go against the opinion of the in-crowd. I don't know if this policy is the right way to tackle these problems - perhaps there isn't a way to do so - I'm too discouraged right now to know -- sannse (talk) 22:18, 28 Aug 2004 (UTC)
"it was an edit designed to provoke a reaction" Allow me to disagree here with the interpretation. It was an edit targetting what was worst in the page: an arbitrary deadline, not someone's ego. I disagree with the proposal itself, but I was ready to discuss it, and started doing so in the talk page, with little response. Next thing I noticed is that the proposal is of limited visibility and the deadline is unrealistic, therefore I targeted the deadline, so that people smarter (and more polite) than me can join the discussion. As I admitted, it was especially interesting for me to see what kind of reaction it will cause, and it is only natural. Each time you edit a text written by a person who disagrees with you, it definitely "provokes a reaction". Also, and again, I DID NOT edit the proposal! Mikkalai 00:13, 29 Aug 2004 (UTC)
I understand that you did not edit the proposal (as I clarified in my last edit with "proposal page"). My interpretation of this as an edit designed to provoke a reaction was based on your remark... "I confess I decided to run a small experiment to get an idea how these measures are going to be implemented". Perhaps it would be fairer to say it was an edit designed to see if there would be a reaction. Either way - I don't think this is important now. I'm more interested in understanding your view-point on the idea of a policy, whether you believe one is necessary and (if one is needed) what form should it take. -- sannse (talk) 20:58, 29 Aug 2004 (UTC)

Measure II has been once more revised[edit]

The following statement has been added to Measure II:

In order to prevent misuse of the system, users would be disallowed from listing an administrator on Wikipedia:Requests for review of administrative actions more than one time in any 30-day period.

I'd like to see some discussion on this... on the statement itself and then the time period. Is 30 days acceptable? Or would some of you prefer a shorter time? blankfaze | (беседа!) 22:50, 27 Aug 2004 (UTC)

I personally hate fixing arbitrary numbers into policy (page size in kb, number of voters, length of vote, minimum of edits to become a sysop etc etc). I blathered on about why I think this at Wikipedia:How to create policy. Pcb21| Pete 23:10, 27 Aug 2004 (UTC)
I can see your point, but what alternative do you suggest? There needs to be a safeguard against abuse of the policy. - pir 09:47, 28 Aug 2004 (UTC)

The proposal is meaningless. If a bad guy is pesistently abusive, there is no need each time list him anew. It is only enough to add more evidence into the initial record of abuse. On the other hand, if someone repeatedly marks omeone else as abusive without serious, this is clearly a "disruptive action" and must be handled as such without any new policies. Mikkalai 23:38, 27 Aug 2004 (UTC)

By the way, how about "forgotten and fogiven" threshold? (I mean erasing from direct public records after some time; clearly you cannot erase from history.) Mikkalai 23:38, 27 Aug 2004 (UTC)

Why the policy is being edited despite the request against this?[edit]

Is there a rule anywhere that a person who starts a page may act as supreme authority? The pase says "please do not edit this proposal". Why it is edited by the initial author while forbidden for the rest of us? Mikkalai 23:38, 27 Aug 2004 (UTC)

I suggest to take an example fom a generous schedule used at Wikipedia:Dealing with disruptive or antisocial editors

Similarly to this example, IMO no votes should be declared started without a "cool-off" period after the last change of the proposal. Mikkalai 01:13, 28 Aug 2004 (UTC)

unintentional (?) hilarity[edit]

Does anyone else find it terribly amusing that the author of a policy designed to deal with abuse of sysop powers threatened a blatant abuse of sysop powers in order to preserve a preferred version of that policy? Was the comedy deliberate?

More seriously: Should this policy address the problem, now demonstrated, of admins who use their sysop powers to threaten or intimidate other users? Admins are not imbued with any special authority, yet right here we see admins trying to use their technical capabilities to whip other users into line—a line (i.e. please do not edit this proposal and so forth) which they had no authority to draw. (They can ask, yes; anyone is free to request that their proposed policies remain unedited, and anyone else is free to ignore that request.) Should this sort of thuggery be penalized as severely as if the threats had actually been carried out? Should some lesser sanction be applied? Or should these abuses go unpunished? —No-One Jones 06:42, 28 Aug 2004 (UTC)

  • Haha. True. 'Tis a bit comedic now that I think about it. But the guy was trying to be an asshole, so I have no qualms about it really. I WOULD LIKE TO MANAGE THIS POLICY WHICH I HAVEN TAKEN THE TIME TO WRITE UP. Why is that such a crazy demand? Anyhow. Thuggery? That's really not what I was doing, or at least intending. But go ahead and blow it all up out of proportion if it gives you something to do. blankfaze | (беседа!) 15:19, 28 Aug 2004 (UTC)
  • Please explain how changing Discussion on this proposal shall last with The authors of this proposal suggest that the discussion on this proposal shall last amounts to being an asshole. I can agree only to be "pain in the ass" (I know I sometimes am.) As a demonstration of good will I can explain my wording act as tsar: (1) declared attempt to grab ownership (2) forbidding modifications by other persons but for himself and "tructed few" (3) ignoring actual criticism.
  • I aware of the policy of avoiding personal attacks. But we are talking not about a topical article here. A policy is implemented by people, not by computers. Therefore a significant consideration should be given to the personality issue. More concern should be given NOT to how good people will use it, but how bad gyus will. I wrote repeatedly and am not tired to repeat again (well, there even is a say about a fly in the ointment). IMO the ones who tend to introduce stringent policies usually don't have enough patience and/or desire to resolve issues amicably. They prefer to grab a big stick and kick your wrongdoing ass. Naturally, the fisrt step is to manufacture this big stick... I looked into the history of blocking of bad admins. Well. Some are gone, some are doing well. Wikipedia lives. Now here come new people and try to turn it into police state. Mikkalai 16:40, 28 Aug 2004 (UTC)

Having reviewed some history of sysop abuses, I wonder if this policy is even necessary. Someone remarked on the mailing list, correctly, that abuses of admin power are very easy to spot. Public attention is an effective antidote to abuses of power, at least in an open society, and I think this holds true for Wikipedia: for every sysop willing to misuse his or her powers, there are others ready to raise Cain about the abuse—and to undo it. Some users are ready to question even those actions that later turned out to be well-justified, as the JRR Trollkien block wars demonstrated. This social pressure is a check, albeit unofficial, on the actions of would-be dictators; reasonable people (and the sysop selection process is pretty good about weeding out the unreasonable) aren't going to continue acting against strong community disagreement. If they do, well, there's always arbitration.

There have been and will be extreme cases; 168... is the only one I recall, and s/he was a usually-rational person acting irrationally as a deliberate act of protest. However, even a rogue sysop can't do much permanent damage (maybe image deletions, but I think developers can undo those), and extreme cases can be dealt with in extreme fashion: a straw poll desysopped 168, and emergency appeals to the arbitration committee, or to Jimbo, could be used against any future rampages. —No-One Jones 19:29, 28 Aug 2004 (UTC)

in this case, I believe the decision to unsysop 168 might have been wise at the moment it was done. The fact he stayed unsysoped, that it took 2 or 3 months for the arbcom to take care of the case, and more than 6 months for Lir to be banned, turned it into a sort of a joke. I am not sure the emergency was so high that it was worth it.SweetLittleFluffyThing


What I really find worrisome is what I touched on above, at the end of all the taunting. I am concerned about sysops who merely threaten to use their powers in order to win disputes, but without following through and thus exposing themselves to scrutiny. Sysops and others familiar with policy know that they can ignore hollow threats, but a newer user might not, and unless someone spots an abusive sysop right away, it could have a real and uncorrected chilling effect on debate.

So my question is, should threats to misuse one's admin powers be treated as equivalent to real abuses? (This is my view.) Should they be considered a misdeed, but less severe than an actual abuse of power? Should they be disregarded entirely, leaving sysops free to throw their weight around as they please? Or something else? —No-One Jones 03:28, 29 Aug 2004 (UTC)

I would say that such threats could be percieved as personal attacks and dealt with appropriately. Threatening to use admin powers is not the same as actually using those admin powers. -- Grunt (talk) 18:03, 2004 Aug 29 (UTC)
But it is an abuse of those powers in that you have to have those powers in order to make the threat. I think this relates directly to admin abilities and so is a separate issue from a general threat. -- sannse (talk) 20:58, 29 Aug 2004 (UTC)

Another extension...[edit]

As I have become aware that 30 August is a holiday in the UK, I have once again delayed the start of voting, this time by another 24 hours. This will allow people who may be unable to participate in the discussion over the weekend to return from their holiday and still have one day for any final input they have before voting starts. Voting will now begin at 00:00 UTC, 1 September 2004. blankfaze | (беседа!) 07:31, 29 Aug 2004 (UTC)

I shall only warn you against a potential problem with this proposal. By some points, it reminds the system we use on the fr wikipedia. We are plagued by votes made by sockpuppets. Most of the votes were ineffective to take any action, so the quorum was not even reached for the last vote (people realise their vote changes nothing).

Other than that, I find it curious that only one editor is authorized to edit the proposal (not sure this is a brainstorming) and that in the end, users are only given the choice between supporting or rejecting the entirety of the proposal. This reduces chances to reach the best proposal. Imho.


  • The only reason that I don't want to open up editing on the proposal (even though i realise it is un-wiki) is because i'm worried that it'll turn into a joke, lir's Sysop Accountability thing. blankfaze | (беседа!) 21:21, 30 Aug 2004 (UTC)
    • That may be wise :-) SweetLittleFluffyThing
    • Just for the record -- and I suspect I'll win no love for this, but as it's true it needs to said, and besides perhaps it's an additional reason for blankfaze's proposal, weren't most of the users who turned Lir's Sysop Accountability thing into a joke now themsleves sysops? (Yes, Lir included a vote option that could be judged provocative, (but not a joke) I know -- but most of the damage was done by sysops. zocky offered a pretty balanced deconstruction here: Wikipedia:Votes for deletion/Sysop Accountability Policy) -- orthogonal 05:54, 31 Aug 2004 (UTC)
  • FOR THE RECORD, I am open to ALL suggestions as to the wording and content of the proposal, and I have in fact already implemented numerous changes suggested by other users. blankfaze | (беседа!) 21:23, 30 Aug 2004 (UTC)
I noticed that Grunt just made an edit to the policy. I want to comment on it, but I don't know if I should consider his edit "official" or not. If you'll tell us who exactly is allowed to make edits, it might be easier. Thanks. -- orthogonal 22:41, 30 Aug 2004 (UTC)

Revisiting Measure II[edit]

"In order to prevent misuse of the system, users would be disallowed from listing an administrator on Wikipedia:Requests for review of administrative actions more than one time in any 30-day period."

My understanding of the discussion above was that we were considering limiting User X from making more than one listing in a 30-day period. However, as written this guideline can be read to prevent any user from listing Admin Y more than once in a 30-day period - in fact, I submit that this is the most grammatically sensible reading of the statement. In other words, after having been listed once, Admin Y is then free to abuse admin privileges with impunity for 30 days. It should be self-evident that this is not acceptable. --Michael Snow 18:44, 30 Aug 2004 (UTC)

I've reworded Measure II to read "...misuse of the system, any individual user would be disallowed...". -- Grunt 🇪🇺 20:38, 2004 Aug 30 (UTC)
Yes, I see that I worded that rather vaguely. I really don't mind if people edit the proposal to fix wording or to clarify things. I just want my idea to remain intact and I don't want this to turn into a big joke. blankfaze | (беседа!) 21:24, 30 Aug 2004 (UTC)
So a sysop can abuse user X numerous times in a month, but the user abused can only list one? Should the user list the first such abuse, or wait to save his one shot for potentially greater abuse coming later? doesn't this essentially give a sysop a "get a out of jail free" card after the first listed abuse? -- orthogonal 22:44, 30 Aug 2004 (UTC)
Not at all. If the sysop is really seriously abusing a user, then anyone else is always free to list it, and would, I'm sure. While this system may not be ideal, it's a necessary compromise. Else we'll have all sorts of meanspirited or trolly users listing admins 200 times a month just to be assholes. blankfaze | (беседа!) 00:01, 31 Aug 2004 (UTC)
Yes, that's a fair point, but it seems there must be a better way to do it. How's once week sound to you? At most it means 40 more listings per troll per year. ;) -- orthogonal 00:26, 31 Aug 2004 (UTC)
I dunno... that could turn into a really annoying situation if some persistant red faction type came along with a grudge. How about two weeks? I think that would be fair. blankfaze | (беседа!) 00:44, 31 Aug 2004 (UTC)
Hmm. At what point have the "trolls" won if we bend our policies and perhaps allow rogue sysops to go about abuse with impunity, just because of the possibility that a "red faction type" might do something? You yourself are willing to take the chance that sysops might abuse |Blocking_policy/Personal_attacks by blocking innocent users for up to several days, but you're not willing to chance some "red faction type" adding additional lines to a list? It's just words on a list in this case. -- orthogonal 00:57, 31 Aug 2004 (UTC)
Hmm. I dunno. I just think one week might be too short a term... I don't see what you mean about bending policies... I think that 2 weeks is a very acceptable time period. Much better than a month, at least from your point of view... Anyhow, if you accept this compromise, I shall edit the proposal as such... But if not, I guess I'll leave it be as no one else has expressed concern about this issue, really. blankfaze | (беседа!) 01:51, 31 Aug 2004 (UTC)
I don't want to be disingenuous. 2 weeks is better, but it's not good enough (for me), so telling you to include it as if it would satisfy me wouldn't really be honest. -- orthogonal 01:57, 31 Aug 2004 (UTC)
I suggest to modify:
"...misuse of the system, any individual user would be disallowed..." with something along the lines: "...unless a new and significant evidence is presented that the abuse did not cease." or something, without any deadlines mentioned. Please rephase it better. Me no lawyer. Mikkalai 01:28, 31 Aug 2004 (UTC)
That idea would be nice, except that poses the question, "Who gets to decide whether or not new and significant evidence has been presented... Who gets to decide what is significant..." and all sorts of other mess... :-/ blankfaze | (беседа!) 01:51, 31 Aug 2004 (UTC)
Unfortunately, at some point or other we have to rely on common sense and general rule about trolling. Setting arbitrary numerical constraints has its own drawbacks already discussed, I believe (but don't know where). For example, numerical constraits may be easily circumvened by a vicious clique. Anyway, any legal system has its imperfections and has to evolve. IMO we (you and me) seem to differ only in one basic issue: you seem with to start immediately with some strictly defined rules, while I prefer to give initially some slack and set more stringent rules when an actual necessity arises. I may be "hysterical" sometimes, but I do believe that people here on wikipedia are inherently good. Of course, if at some moment cosa nostra decides that wikipedia suits their needs, IMO any number of policies would not help anyway. Therefore I consider "angry" policies as an insult to the majority. Even apparently "vicious" guys around here IMO have good intentions, and it is amazing how most heated discussions eventually arrived to some point of balance without necessity to apply some kind of force. Mikkalai 04:01, 31 Aug 2004 (UTC)
I find myself sympathizing with both blankfaze and Mikkalai here. Like blankfaze, I'd like enough objectivity such that relying on subjective interpretations is minimized, and a hard number would do that. But Mikkalai's appeal to the good intentions of most users is also appealing.
And I note that this exchange is indeed evidence for Mikkalai's argument: a day or so ago the two of your had a rather heated exchange over editing the proposal, blankfaze justifiably anxious to maintain the proposal's integrity, Mikkalai justifiably desiring to exercise any user's right to make changes, that ended up with threats of blocking. But today the two of you are not only amiably discussing it, together you're making it a better proposal than either of you would perhaps have done alone.
This proposal (correct me if I'm misunderstanding) already relies on the entire user base of wikipedia to decide whether a sysop's actions were abusive or not, by a straight up-or-down vote. Since that already unavoidably introduces the subjectivity of every single voter, blankfaze's desire to reduce subjectivity seems even more to justified. But since we do have to introduce the subjectivity of voting, and since the voters will already be deciding if sysop's actions were correct, let them also certify if the complaint was correct.
This does complicate the vote, but not too much. Where before we have "concur" or "no confidence", now we have "concur with sysop; allegation unjustified", "concur with sysop; allegation justified", "and no confidence in sysop; allegation justified". Only the second of these alternatives, "concur with sysop; allegation justified" is an addition, and it deserves some explanation.
Basically, it means the sysop was justified, but that the voter feels it was a close enough thing or a gray enough area that the complaintant was justified in asking the sysop's action to be reviewed (the contrary, "no confidence in sysop; allegation unjustified", makes no sense, as it could only be used to punish the sysop for something other than the allegation, something we wish to avoid). If the allegation is held to be justified by a majority, regardless of whether the sysop is upheld or no confidence is given, the complaintant may make another allegation before the expiry period (which is one month, or two weeks, or whatever you decide on). If the complaint is found unjustified, the complaintant may not make another complaint against that sysop until the period expires.
This would stop "trolling" complaints just as quickly as having a set expiry period, as unjustified complaints would be decided to be such on the first such complaint, forestalling any others. But justified complaints -- even if it's such a gray area that the sysop is also upheld -- in other words, non-"trolling" complaints of all types, could be made as often as necessary. I'm sure there are flaws in this proposal, and that both of you, blankfaze and Mikkalai, as well as other users, can find ways to improve it, but I think it forms at least a useful basis for further discussion. -- orthogonal 04:33, 31 Aug 2004 (UTC)
I can see the benefits in your idea; However, I'm worried that it would make the whole thing too complicated, when my primary intent was to simplify the process... I'd like to keep the system as basic as possible. I kind-of agree with Grunt's comment below in that if a sysop is really abusing his powers, that others will notice and list it. I'd really like to avoid making this proposal any more complex, if at all possible. Anyhow, I'm really open to any and all suggestions on this... surely there has to be some compromise we can all settle on to get this proposal passed. I really do think this (or something similar at least) is necessary. blankfaze | (беседа!) 04:45, 31 Aug 2004 (UTC)

Length of time before a user may re-list an admin[edit]

In reference to the section:

"In order to prevent misuse of the system, any individual user would be disallowed from listing an administrator on Wikipedia:Requests for review of administrative actions more than one time in any 30-day period."

What is the most acceptable time period?

30 days (as is)

  1. If the sysop is a serious issue, someone else will list them. -- Grunt 🇪🇺 02:14, 2004 Aug 31 (UTC)
  2. pir 08:58, 31 Aug 2004 (UTC)

One week


  • Immediately. No limit. If someone abuses this, the consequences are obvious: like the boy who cried wolf, the voters (who under your proposal have to "concur" or show "no confidence" in the sysop's action) will automatically vote concur. So abuse really ends up only harming the abuser; he'll lose the protection of this policy.
(Indeed, we might need some protection against the immediate dismissal of what might well be real grievances by users tagged, fairly or not, with the label "troll".) -- orthogonal 02:19, 31 Aug 2004 (UTC)
I see your reasoning, but without a time limit, what is to stop a crazed user from listing someone 10 times a day? That would clog up the system, make a farce of it. There must be some sense of order, aye? blankfaze | (беседа!) 04:49, 31 Aug 2004 (UTC)
A crazed user can do this anywhere. By definition, he is troll. I doubt he will be in a position to find a "new, reasonably justified complaint" 10 times a day. Hence he will be a sitting duck for the rest of us to block him. Mikkalai 04:53, 31 Aug 2004 (UTC)
(I've taken the liberty of indenting Mikkalai's comment, above, to avoid interpolating my comment between his and blankfaze's. If this was wrong, please feel free to change it back and call me a dumbass.;-) ).
As I understand it, we don't have any technical way of preventing anyone from doing this, and if the user in question is "crazed", they'll likely ignore the policy requirements anyway. So a policy won't protect us, and we'd have to treat the crazed user as a vandal. I mean, the policy can't protect us against a vandal bot or a power outage at the server farm or a DDOSing or the civil war that "time traveler" guy says happens (happened?) in 2005. But my question really is, how many crazed users do we have, and don't we have other ways to deal with any that exist?
blankfaze is approaching writing this policy like a programmer, and as a programmer myself, I definitely understand that particular motivation. But at some point we have say, well, there are certain exceptions we can't handle, and we just have to gracefully terminate and let the operating system deal with it. But as in programming, let's not expend too much sweat on rare corner-cases, and instead (my opinion is) let's concentrate on making this thing work under most usual conditions. And christ I'm too verbose. -- orthogonal 05:26, 31 Aug 2004 (UTC)
Of course a policy cannot affect external events like power outages. However it should internally prevent itself from being abused, otherwise it will get rejected and rightly so. - pir 08:58, 31 Aug 2004 (UTC)
If we find that 30 days is insufficient to deter rogue sysops, we can always amend the policy later. -- Grunt 🇪🇺 14:51, 2004 Aug 31 (UTC)
True. But it might be easier to tell if the system is being abused, rather than if bad sysops are being deterred. If so, why not start with no expiry period and increase it when and if needed? Right now, I fear that it gives bad sysops a 29 day "get out of jail free" card. -- orthogonal 15:50, 31 Aug 2004 (UTC)
I still think that sysops bad enough to warrant more than one warning in a thirty day period probably warrant the attention of more than one user; therefore the sysop can still be written up within those thirty days (just not by the same person). -- Grunt 🇪🇺 17:45, 2004 Aug 31 (UTC)
OK. Will this be because the bad sysop does things to these other users too? If the bad sysop is just going after one person or a few, how will other people know enough to list him? And if I list against a rogue on behalf of another user, if the sysop is really bad, he may retaliate against me. So I might be reluctant to list in the first place, from fear of potential retaliation, and even more reluctant to list and give up my chance to list and thus protect myself if he does in fact come after me. -- orthogonal 17:52, 31 Aug 2004 (UTC)
The people who like to list sysops in such places have armies of sock puppets at their command. They will list sysops early and often. uc 23:31, 1 Sep 2004 (UTC)
Excellent! I can think of few better ways of identifying and eliminating sock puppets. This alone is reason to pass the proposal. -- orthogonal 23:37, 1 Sep 2004 (UTC)
  • Immediately. What is to stop a crazed user from listing someone 10 times a day? The block button. anthony (see warning) 02:56, 1 Sep 2004 (UTC)

Voting begins[edit]

Voting has begun, and will continue until 00:00 UTC 2004-09-08. -- Grunt 🇪🇺 00:40, 2004 Sep 1 (UTC)

Since the voting is so tight I think the deadline should be extended. Eric B. and Rakim 10:00, 6 Sep 2004 (UTC)

It doesn't seem tight at all. Right now it's:
  • Measure I - (18/26)
  • Measure II - (10/3/21)
  • Measure III - (8/0/28)
Seems pretty clear it's thumbs down on all measures. Fuzheado | Talk 10:11, 6 Sep 2004 (UTC)
Nah, it's not really tight at all. All measures will fail. But thanks for you hope! blankfaze | (беседа!) 17:28, 6 Sep 2004 (UTC)
I'd like to thank Blankfaze, however, for initiating this, even though I was against the exact wording of the measures. It is a useful exercise to try to sort out what is acceptable administrator behavior, as it has clearly "crept" into areas which have caused problems. Fuzheado | Talk 01:46, 7 Sep 2004 (UTC)
Thanks. I figured there would be considerable opposition to this. I'd still like to see something similar implemented in the future, though. blankfaze | (беседа!) 04:02, 7 Sep 2004 (UTC)
blankfaze, while I don't think the proposal is perfect, I think it's an excellent start, and I look forward to working with you to develop a policy that can be generally accepted. -- orthogonal 04:13, 7 Sep 2004 (UTC)