Wikipedia talk:Arbitration Committee/Procedures/2014 review

Community appointments of functionaries

Overview

The current method of appointing checkusers and oversighters is bureaucratic, involves secret vetting and little direct community participation, and vests the power of appointment in non-experts. It may be more elegant to run simple elections, particularly because the memory of the last CU/OS elections, held at a time of great turmoil for the project, is long gone.

Proposed changes:

• Elections to CU & OS would be held. ArbCom appointments would be deprecated. These elections could be held at the same time as the ArbCom elections, or held at other times of the year and be co-ordinated by the checkusers and oversighters themselves.
• To comply with the minimal requirements of Foundation policy, ArbCom could certify the results, but would not need to vet or vote on candidates.
• Responsibility for scrutiny of functionaries and the removal of inactive functionaries' permissions could be vested properly in AUSC, a body that uses its powers very infrequently because it needs every decision checked by a distant committee.
• AUSC would be responsible for implementing its own decisions; e.g. if it voted to remove a functionaries' permissions, it would have standing authorisation to direct the stewards do the removal.

ArbCom is responsible for CU & OS elections due to a historical anachronism. It is a distraction to its primary business of arbitrating cases and dealing with sensitive matters, and the community is more than mature enough to appoint its own functionaries. AUSC is more than fit for purpose, and has a long track record of holding functionaries properly to account.

Proposal draft v1

Appointments CheckUsers and Oversighters are appointed at Requests for CheckUser or Oversight permissions (RFCUOS). Non-administrators are eligible. Any candidate may self-nominate at any time of the year except November and December, when community attention is already being given to the annual Arbitration Committee elections. Candidates submit a candidacy by self-nominating at RFCUOS, and a one-week question and answer session follows. Any registered editor in good standing may ask a question of a candidate; candidates are encouraged but not required to respond. Two days for comment will ensue, the Arbitration Committee will be notified by the candidate of the candidacy, then a five-day voting period will commence; registered editors in good standing with at least 500 edits may vote once, in support or opposition of the candidacy. Neutral votes and attaching comments to votes is prohibited. Nominations which attract more than or equal to 70% of the eligible votes submitted within the allotted time will be successful. Nominations will be closed by an arbitrator, then the Arbitration Committee will direct the Meta stewards to assign the candidate the permissions in question. Unsuccessful candidates will be ineligible to restand for six months. Members of the Arbitration Committee elected by the community will continue to be eligible to receive either or both permissions, without undergoing this process, and to hold them for the duration of their term. Arbitrators who are checkusers or oversighters ex officio will lose the permissions on leaving the committee. Auditing Complaints about the use of checkuser and oversight shall be within the jurisdiction of the Audit Subcommittee, to whom this jurisdiction is given by the Arbitration Committee. Responsibility for auditing these tools will also pass fully to the subcommittee. Activity and inactivity statistics will be maintained by the functionaries, and responsibility for enforcing the activity requirements (which are, and shall continue to be, maintained by the Arbitration Committee), will pass fully to the Audit Subcommittee. Members of the Audit Subcommittee will continue to be appointed by the Arbitration Committee, in consultation with the community. Removal Removal of the permissions is with the Audit Subcommittee, who may resolve by majority vote to remove the permissions from any functionary at any time under the auditing procedure. The Arbitration Committee will retain the ability to remove the permissions for conduct unbecoming or for the loss of trust.

AGK [•] 11:10, 24 May 2014 (UTC)

Rationale

The Arbitration Committee currently appoints editors to and audits the use of advanced permissions. Under the Arbitration Policy, the committee may delegate any of its responsibilities to whatever body it thinks fit. At the time of writing, the committee has delegated responsibility for auditing to the Audit Subcommittee (AUSC), but retains responsibility for implementing the decisions of AUSC, and has not delegated the appointment of functionaries at all.

This system is fairly archaic, results in an enormous amount of bureaucracy surrounding a simple process, and blurs the lines between the jurisdictions of ArbCom, AUSC, and the community. The system is no longer fit for purpose. Just as worryingly, the system relies on an attentive, competent Arbitration Committee, despite the fact that most arbitrators have no idea how to use the advanced permissions, and are primarily appointed to write decisions on disputes – not elect editors to specialised roles. Arbitrators are like judges who, in addition to hearing cases, are also required to hire computer programmers. Checkuser and oversight is far outside of ArbCom's area of expertise.

The obvious solution is to:

1. let the community appoint the checkusers and oversighters. There is wisdom in the crowd, particularly as many of this crowd are SPI clerks, frequent users of SPI, or administrators experienced in revision deletion and reporting edits for suppression.
2. let the Audit Subcommittee properly audit the use of checkuser and oversight, and implement their own decisions.

This lets ArbCom focus on drafting decisions and hearing requests and appeals, delivering a more effective arbitration process, and improving the functionary auditing and appointments process.

AGK [•] 13:15, 25 May 2014 (UTC)

As those with long memories will recall, the current system was implemented after the May 2010 election. It was hoped that 6 oversighters and 4 checkusers would be elected. The result was 1 checkuser elected, the 11 other candidates (myself included) did not get sufficient support. Or, more to the point, many of us got plenty of support but being active admins who have had to make tough decisions we also got plenty of opposition. (results here) The problem as I saw it was at least partially that we used secure poll to do the voting. No explanation was needed for opposition or support. None of us knew who opposed us or why they did so. If it is to go back to elections I suggest a different process be used. With ArbCom we determine how many slots there are to fill and then the top vote getters fill those posts, period. I would suggest that functionaries be elected the same way. Beeblebrox (talk) 22:57, 25 May 2014 (UTC)

Beeblebrox, isn't there still a minimum threshold of support for election to ArbCom? Why would it not make sense to have the same for other functionaries? And is the only reason for switching from SecurePoll to public voting the hope that people will be discouraged from opposing? I recall that ArbCom elections don't require explanations for opposition, either, or am I wrong? Nathan ^T 01:04, 29 May 2014 (UTC)

The 50% threshold for arbcom wouldn't work for CU/OS. Global policy needs 70% approval --In actu (Guerillero) | My Talk 14:28, 29 May 2014 (UTC)

Comments on the revised procedures

In essence these new rules appear to stipulate 5 actions every 3 months, and at least 1 response to a community request. Provided I'm interpretting this correctly, that looks fine. The other changes appear to be slightly clearer wording and so on, but relatively little actual change to procedures. PhilKnight (talk) 11:36, 24 May 2014 (UTC)

The procedures at WP:AC/PRR are just copies of the current procedures, which are stored at WP:AC/P. The proposed changes are on this page; I have raised some in the section above. AGK [•] 12:10, 24 May 2014 (UTC)

@PhilKnight: I have edited the attached page to make the proposed changes more clear. AGK [•] 12:13, 24 May 2014 (UTC)

• @AGK: Would this change mean that the only vetting would be what the community does during the election or would the AUSC also do some vetting? Callanecc (talk • contribs • logs) 13:41, 24 May 2014 (UTC)

• Sounds interesting and in line with what a number of the community have been asking for. I'm curious, will the election process be secret? (like Arb currently is) or more like RFA (probably not a great idea..) Is there a threshold to pass, or is it based on a fixed number of slots? I looked around but didn't see any details, or perhaps they have yet to be worked out. Dennis Brown | 2¢ | WER 13:52, 24 May 2014 (UTC)

• @Callanecc and Dennis Brown: The arrangements would be whatever is most sensible, but they are still open to discussion, and are best decided at a later point. AGK [•] 14:45, 24 May 2014 (UTC)

• Without seeing the actual planned changes to the text, I don't think it's really useful to comment very greatly. However, if the Arbitration Committee is washing its hands of responsibility for selecting CU and OS, then it needs to divorce itself from the AUSC as well (i.e., it would no longer be an Arbcom subcommittee) and give the AUSC the direct right to confirm and remove CU/OS rights. Alternately, it can fold up the AUSC entirely and refer issues to the Ombud instead. Also, it means that arbitrators should no longer be automatically granted CU/OS upon their appointment. That is, Arbcom can't pick and choose here, it either throws the whole CU/OS issue to the community or it keeps it. But bear in mind that there hasn't been any genuinely significant community interest in taking over the management of CU/OS in several years. What makes Arbcom think that the broader community wants this task? It's hardly the biggest time sink Arbcom has on its plate - that would be ban/block appeals, which I would have thought would be first on your list of tasks to divest. Risker (talk) 15:48, 24 May 2014 (UTC)

• You've been told the thrust of it, so if there are grounds to oppose in principle I'd love to hear them now before more time is wasted. I don't immediately see why AUSC can't, practically, be completely independent, but keep the name of "subcommittee"; that is where its authority comes from, in any case. (As for the Ombudsman, I think it inadequate for our purposes.) Why are you saying it's the current AUSC or nothing?

  The broader community, presumably, wants an effective committee. If that's the case, the committee really needs to drop all this baggage it's carrying. We can do less but better, or more but worse; that's the true dichotomy here. In addition to always wanting an effective, competent committee, the community also already approves of appointing people to "offices" in an open, accountable manner.

  Arbitrators having CU and OS seems like an issue of less importance, but since you raised it: I see no reason why ArbCom can't continue to give arbitrators the permissions. We have them for investigating committee issues as well as to allow us to observe.

  Reform of ban appeals is going on as we speak, but isn't yet ready for public consultation because there are more details to iron out. AGK [•] 16:18, 24 May 2014 (UTC)

I'm okay with th notion that Arbcom wants to get out of this business, provided that the community agrees to modify the arbitration policy, which is pretty much a prerequisite to any of these proposals making sense. It was obvious several years ago when the last CU/OS election was held and only one candidate met the minimal standards, and after which the community has shown almost disturbingly little interest in CU/OS selection, that this isn't on the community's radar as something that needs fixing. And if the committee's purpose in divesting itself of this role is to focus on cases, then there's no rationale for it to grant its own members the permissions. This isn't just an adjustment of Arbcom's practices, it means rewriting the CU and OS policies, and it means developing new processes within the community that the community has shown no interest in developing. I actually disagree with you that AUSC has shown its ability to do anything much in the past few years, because it's been largely inactive. Risker (talk) 17:01, 24 May 2014 (UTC)

I've written some of my initial thoughts on configuring this, but looking at it now it doesn't appear anything here actually involves changes to the arbitration process. This isn't really ArbCom abdicating its responsibility to appoint new checkusers and oversighters, but merely ArbCom drastically reducing its role (and increasing that of its delegates).

By the same thinking, ArbCom hasn't either. Very few functionaries, these days, are "disciplined" because, as you remember, the worst ones were thrown out by the first AUSC. I don't think changes to the CU and OS policies are necessary; nobody's proposing that ArbCom not make the actual appointment (which is a requirement of global policy). AGK [•] 17:11, 24 May 2014 (UTC)

Erm, no, it's not a requirement of global policy for Arbcom to make the appointments if the community runs an election, which is what you are proposing. There's no reason for Arbcom to be involved at all in the appointment process then. Risker (talk) 17:18, 24 May 2014 (UTC)

I don't think we can go down that route, because the community wants independent oversight of ArbCom. That method provides for no independent oversight, so the proposal is a hybrid system: keep the oversight (but passed fully to AUSC), and keep the election appointment system. Here's another way to think about it: this is a return to the old elections system, and a scrapping of the time-wasting stuff that's built up at ArbCom's side, around CU & OS. AGK [•] 17:28, 24 May 2014 (UTC)

If the community is directly responsible for CU/OS selection, leaving Arbcom out of it entirely, then there's no need for "independent oversight of Arbcom" in relation to that matter, and the hypothetical independent oversight can stick to cases and block/ban reviews. Saying that Arbcom must be involved in this process because the community wants independent oversight of Arbcom is nonsensical; the community would have DIRECT oversight of the CU/OS process. And it's not a return to the old election system, which was predicated on candidate pre-vetting by Arbcom, the step that you're trying to deprecate. Risker (talk) 18:18, 24 May 2014 (UTC)

Oops, I meant "independent oversight of CU & OS" (struck above, accordingly). In a fully elected system, the global policy would still allow only the ArbCom to vote on removal, so why are you saying the community would have independent oversight? The proposal is for ArbCom to then delegate responsibility for removal to AUSC, under the "delegation" clause in WP:ARBPOL. AGK [•] 12:47, 25 May 2014 (UTC)

From an outside perspective, it's nice to see enwiki giving directly-elected CU/OSs a try again. Out of ~18 wikis with functionaries, only four use solely arbcom appointment; the rest find that direct elections work well. What is a bit strange here is the bureaucracy that is still in place. The move is explicitly about reducing bureaucracy, yet there is just as much in place afterwords as before.

A very simple way to do it, which is done on the majority of other projects, would be as follows: A user submits a request for CU/OS, a week is allotted for voting (though a time for questions is a good idea too!), at which point if the user has 25 supporting votes and at least 80% support, a bureaucrat closes the request and asks a steward to implement the consensus. Monitoring of the existing CU/OSers is done by the Ombudsman Commission. Removal for inactivity is done alongside existing activity review procedures.

Of course enwiki doesn't need to follow what is commonly done, but this complicated process of arbcom ceremonially appointing the candidates after an election is strange, as is the continued existence of the AUSC. If anything, oversight would be maximized by allowing the OC (which is responsible for the oversight of functionaries everywhere else) to do its job here as well. That way, there would be true oversight of what is going on, not review by users who might have a conflict of interest as they also hold local rights. There are also no projects which combine elected CU/OS and appointed by arbcom, though this isn't necessarily a bad thing to have. If arbcom needs the rights, they meet the global policy requirements to have them.

One other minor point: users with CU/OS rights no longer need to relinquish them to serve on the OC. Ajraddatz (Talk) 21:42, 24 May 2014 (UTC)

• The AUSC is perhaps the most useless body we currently have on en.wiki and, in my opinion, it should be dismantled. Aside from its uselessness, it's also historically been difficult to fill all seats with competent people, due to the paucity of candidates. So, no, I'll oppose any reform which does not include the dismantling of AUSC.

  Quite frankly, the current situation concerning advanced permissions is not problematic and should have been left well enough alone; if there's something we should be reviewing, then that is BASC. Salvio ^{Let's talk about it!} 09:17, 25 May 2014 (UTC)

• Do you consider AUSC to be useless in theory, in practice, or both? In theory, I think there is value to an independent auditor system. In practice, I see an AUSC that has little to do in a functionary system containing little actual abuse or misuse; that it isn't busy doesn't mean it isn't needed.

  I do not consider the fact that BASC also needs reviewed a credible argument against reform, and in any event if you wish to reform ban appeals why not do so yourself? AGK [•] 12:47, 25 May 2014 (UTC)

• Do you consider AUSC to be useless in theory, in practice, or both? Probably both: a body which, on average, investigates fewer than a dozen cases a year is superfluous. And considering the time it takes us to appoint its members, it's also economically inefficient. It's bureaucracy for bureaucracy's sake and should be discontinued. ArbCom can handle those cases just fine.

  I do not consider the fact that BASC also needs reviewed a credible argument against reform, neither do I. On the other hand, I do consider the fact that the current rules concerning advanced permissions do not need reviewing a credible argument against reform. After all, it is up to you to demonstrate that the rules currently in force should be changed; and when one of your arguments in favour of reform is that the new rules would be more "elegant", well I'm utterly unpersuaded.

  You're proposing to change a system because it's old, which is not a reason. Another argument in favour of reform, in your opinion, is that most arbitrators have no idea how to use the advanced permissions. Assuming for the sake of the argument that you're right (and I'm not conceding that), how do you propose to remedy that? Of course, by giving the power to elect functionaries to the community, which is made up of people who may have even less of an idea how to use advanced permissions, and who may, on occasion, have access to less information about the candidates than ArbCom do. Then you go on to say that the current system results in an enormous amount of bureaucracy, which is an overstatement. The point, in my opinion, is different; the question we should be asking ourselves is: since bureaucracy is inevitable, is the amount of bureaucracy surrounding the current procedure to appoint functionaries efficient? In my opinion, it is and, so, there is no need for change. The current system, in my opinion, strikes a good balance between the need to allow the members of the community to voice their opinions concerning the candidates, the need to have people who know how to use advanced permissions vet candidates (to prevent incompetent candidates from being elected/appointed) and procedural efficiency.

  and in any event if you wish to reform ban appeals why not do so yourself? Because it would be counterproductive to run the two discussions concurrently, as you yourself conceded on the mailing list. Salvio ^{Let's talk about it!} 16:02, 25 May 2014 (UTC)

Thanks. To my mind, there are several bases for change:

1. ArbCom is growing increasingly ineffective. Unbundling the responsibilities the community can clearly handle would, I propose, allow us to do a better job of the jobs that only we can do.
2. You miscalculate the bureaucracy associated with these appointments, possibly because (IIRC) you haven't had the misfortunate of co-ordinating a round of them. It drags on for weeks, while arbitrators are burned out or holidaying over summer. More to the point, delegating it to the community, in addition to the benefits I describe in point one, would mean lots of people could vet the candidates: many hands make light work. (We could retain a veto over the appointment if private evidence made it clear the candidate is totally unsuited.)
3. The process isn't very fair at all. The ridiculous "vetting" process denies people the chance to even stand, and candidates were being shot down in recent years on shaky grounds and in the years before for ludicrous, overtly political reasons. In what other Wikipedia process are people allowed into important positions based on the say-so of a bunch of people who, contrary to your veiled assertion, have no idea how to use the tools (as they regularly admit), let alone know who else could use them?
4. Interest in the process is declining, possibly because the community know the real power of appointment lies with us. Administrators are elected in the most hated venue on Wikipedia, yet even they are typically screened by around a hundred people. Compare oversighters, a tiny office with huge power, and only about 20 non-arbitrators bothered to participate last time. While this doesn't necessarily justify my proposed solution, it does seem to discredit the status quo.

AGK [•] 17:12, 25 May 2014 (UTC)

That's mostly a rationale to remove this entirely from Arbcom's scope, AGK. The process you're proposing still leaves all the power in Arbcom's hands, it just farms out the work elsewhere. If you do not think there is value to the vetting process, which is actually the only current step that is genuinely arbcom-dependent, then the entire argument of having Arbcom manage the process is moot. However - and this is the big "however" - there's no indication anywhere that the community has any interest in managing these affairs either. And there's something particularly absurd about expecting Checkusers and Oversighters to run elections: they didn't sign up for such bureaucratic tasks, and the reason for running elections is that they're overworked. Risker (talk) 18:22, 25 May 2014 (UTC)

Incidentally, the global CU/OS policies say that Arbcom "can" remove permissions, not that it is the only acceptable method for removing permissions. Most projects that have Arbcoms are not responsible for CU/OS in any form, and their Arbitration Committees have no responsibility or authority in relation to those permissions. Risker (talk) 18:37, 25 May 2014 (UTC)

I proposed: [Jurisdiction over] removal of the permissions is with the Audit Subcommittee. This actually means we are divesting power; why do you think otherwise? The only reason for retaining AUSC is so the community still has indirect oversight over CU and OS. AGK [•] 22:08, 25 May 2014 (UTC)

Because the committee will still be making appointments. Risker (talk) 22:41, 25 May 2014 (UTC)

Only ceremonially. AGK [•] 19:44, 26 May 2014 (UTC)

The perception of AUSC as being useless might be attributed to the very fact that there are few high-profile cases of abuse. Just as we don't disband disaster management organizations because there are no disasters for the year, we shouldn't disband AUSC just because there are few incidents. I think that AUSC is still necessary as a check-and-balance. I think the avenue should explore is to devolve the role to wider community, becoming completely separate (I also believe a similar treatment to BASC would take a lot of workload off ArbCom), or passed on to the Ombudsman commission altogether. - Mailer Diablo 20:47, 25 May 2014 (UTC)

But we do disband committees whose work is completely overlapped by other committees, in order to reduce bureaucracy and ensure that lines of authority and responsibility are clear. Keep in mind that the AUSC could look at a case, make one decision, and subsequently be overruled by the Ombud commission. It's a duplication, and double jeopardy for CU/OS. Risker (talk) 22:41, 25 May 2014 (UTC)

The O.C. is utterly detached from Enwiki, whereas AUSC's members are appointed from this community. I do not consider the two at all interchangeable. In the entire existence of AUSC, there has been, I think, one case that was heard by both bodies; and there has never been a case where the bodies have arrived at different decisions. AGK [•] 19:44, 26 May 2014 (UTC)

Do we have any evidence, be it statistical or anecdotal from sitting or former ArbCom members, that the CU/OS appointments are an inordinate drain on their time, and removing this responsibility from them will lead to markedly greater efficiency in their other duties? If not, I'm not sure adding a new bureaucracy and a new election system will solve more problems than it causes. Elections on EnWiki already have flaws (just look at the past, oh say, five years of talk page history on RfA/B). Furthermore, the checkuser policy says "On wikis with an Arbitration Committee (ArbCom) whose members have been elected with the support of at least 25-30 members of the local community, CheckUsers can be appointed by the Arbitrators only." So ArbCom will have to "rubber stamp" any decision anyway. In that case, wouldn't it be better for ArbCom to continue to vet, as the responsibility lies with them? I'm sure they would be more comfortable making the decisions—which they must, outside a change at the foundation level—if they do their due diligence; which is what we have now. So absent a marked expectation in ArbCom efficieny, I see the costs outweighing any benefits of this or related proposals. -- Avi (talk) 23:53, 25 May 2014 (UTC)

Yes, I know most of the current arbitrators who have had any sort of extensive involvement in this process consider it a significant distraction. The process has to be run twice a year (AUSC then CU & OS), and each arbitrator has to familiarise themselves with every candidate (because if the arbitrators don't, nobody will – only 14 of us vote when there is no community election). AGK [•] 19:29, 26 May 2014 (UTC)

I agree from what I saw it was time-consuming. I agree if the "vetting" is treated as not vetting but voting (which I was concerned some arbs were doing when I was on the committee) it can make this like having to pass twice. I don't think this can be thrown to the community in its entirety, especially without a process. And what is the basis for "ArbCom is growing increasingly ineffective."? statement? Cas Liber (talk · contribs) 20:20, 26 May 2014 (UTC)

Just to clarify, the CheckUser policy does not require projects with an arbcom to appoint CU/OS... it's an option. Hence the word "can" in the sentence, as well as projects who have arbcoms that meet the 25 vote standard but still separately elect CU/OS (example: dewiki). So there is no need for arbcom to appoint CU/OS if a vote is being held. Ajraddatz (Talk) 17:26, 28 May 2014 (UTC)

Here's some points that I thought of, looking through this. (While I am a steward and that certainly affects my opinions, I don't act on enwiki requests such as granting CU and OS, as this is my homewiki; also, while I have run for CU on this wiki in the past, I am not planning to run for CU/OS here anytime in the near future - being a steward takes up enough time as it is!)

• 5 days seems very short; in comparison, on most other wikis (Commons, Wikidata, dewiki) the minimum length is 2 weeks. We want as many people to participate as possible, and giving it even less time than RFA is not a good idea.
• The global policy requires that (neutral) notifications to the community be put in prominent places; again, we want to get as many people to participate as possible.
• As has been said above, SecurePoll would make it very difficult to get 70%, which is required by the global policy. Yes, ArbCom could then go back and magically appoint that person, but that comes off to me like gaming the system.
• Again, per global policy, removal requests must come from either arbitrators or the community through a vote; AUSC is neither, so ArbCom would have to make the request.
• As has been said above, this would allow for non-admins, though whether non-admin CU/OS are a good idea is debatable.
• I assume that CU/OS applicants would have to disclose their alternate accounts to ArbCom... right?
• Has there been any thought given to passing parts of the election-running to the bureaucrats? Most other wikis do this, and the bureaucrats take the requests to Meta. It would also provide for some separation of powers. Which leads to my next point...
• Having bureaucrat discretion would provide for protection against canvassing, such as what happened at commons:Commons:Checkusers/Requests/INeverCry. There is a strong sentiment on dewiki that enwiki CUs run too many checks, and they have been known to conduct canvassing campaigns in the past, such as at Commons.
• Finally, I'm not sure if having individual nominations is the best idea, because of the difficulty of gauging need, and the overhead required to run each election and promote it - yearly elections might still have some promise. It would also make the elections less painful, so that one person is not stuck in the spotlight. --Rschen7754 01:24, 19 June 2014 (UTC)

Further to, does the community have access to reports or statistics on the current workload or backlog for oversight and checkusers? I think that information, if available, would assist in the decision on whether a significant overhaul should be invested in. Generally, anything that would increase participation from the general community on the appointments seems favourable even if bringing it closer to WP:RFA is a road many do not want to walk. Mkdw^talk 22:43, 11 July 2014 (UTC)

@Mkdw: The AUSC maintains stats of logged actions at WP:AUSC/STATS. In terms of backlogs there isn't really a way to find out as OS requests are made offwiki, you can get some ideas of the CU backlog by looking at WP:SPI, but that doesn't take into account offwiki requests (including WP:UTRS and WP:ACC). Callanecc (talk • contribs • logs) 14:55, 5 August 2014 (UTC)

Comments re AUSC

I can't support any new policy that keeps this: "Community appointees to the AUSC are discouraged from routine or regular use of either tool; however, they are permitted to use the tools in order to develop a sufficient skill level to adequately assess the actions of Checkusers and Oversighters, and may assist in addressing time-sensitive situations, or serious backlogs". The AUSC is a joke and you should let us do work instead of making us sit on our hands waiting for our one case every 3 months. It doesn't make us impartial; it makes us incompetent to do anything but agree with the arbs on the committee --Guerillero | My Talk 21:52, 24 May 2014 (UTC)

This. The oversight team is suffering from a serious lack of manpower (see the bottom of Wikipedia talk:Oversight for evidence of this), and it really doesn't help that Arbcom plans to keep making a significant portion of our active functionaries (which is distinct from "all our functionaries" - many fewer are regularly active than have the bits) unable to help. In addition, as long as Arbcom mandate says that AUSC members cannot do work as functionaries, I suspect Arbcom will have an extremely hard time recruiting experienced functionaries - which should make up at least part of each AUSC - to do the job. Why would those of us who are active hand in our ability to do the work that needs doing in exchange for a the world's worst-paying sinecure? A fluffernutter is a sandwich! (talk) 22:13, 24 May 2014 (UTC)

I would support removing this part from the policy. I don't think it makes sense to restrict AUSC members from doing much-needed work, and I don't see why their activity as functionaries would affect their ability to investigate complaints. GorillaWarfare (talk) 23:35, 24 May 2014 (UTC)

If we were to take this off, we will need to replace this with a requirement for the involved AUSC member to recuse themselves from the investigation if any one of them become the subject of a complaint in the course of his normal CU/OS duties. - Mailer Diablo 21:07, 25 May 2014 (UTC)

@Guerillero: I am amenable to removing that requirement, but it doesn't seem to have much relevance to the proposals being debated. I would welcome your thoughts on the issues raised above, on the understanding that we can always reconsider the AUSC activity requirements a little later. AGK [•] 12:38, 25 May 2014 (UTC)

No, this is my one top issue with the AUSC. I honestly feel that if I don't press it as much as I can it will be forgotten about. And, to be honest, I think your proposals are a problem looking for a solution. If we are talking about big, major, changes to the AUSC then my proposal/plan would look like this:

1. ArbCom divests all powered having to do with functionaries to the Audit Committee
2. AuditCom is made up of 2 committee members and 4 community members the role of the liaison is deprecated
   • The community members are decided upon by a direct election
   • Arbcom gets to decide how their members are decided upon
3. AuditCom has original jurisdiction for all cases of CU/OS tool misuse and abuse on en.wikipedia
   • Much like Jimmy's role with ArbCom, ArbCom has appellate jurisdiction only in cases where AuditCom decides to pull someone's tools
4. AuditCom can make actual decisions and not be a moot court like the current AUSC is
5. The Audit Committee has the ability to set its own procedures
6. ArbCom can not close, modify otherwise meddle in the affairs of AuditCom

The other thing I would support would be to have arbcom absorb the AUSC--Guerillero | My Talk 17:07, 25 May 2014 (UTC)

Surely the two proposals can be debated concurrently…

It appears to me that our proposals overlap, and in fact are nearly identical, so I'm afraid I don't understand quite why you seem to be opposing! AGK [•] 17:16, 25 May 2014 (UTC)

I like the idea of a devolution. I am pretty sure ArbCom wants some workload off them as they have a massive chunk to handle. Also, this would make power less centralized at ArbCom, and I see separation of powers as a good thing. - Mailer Diablo 21:02, 25 May 2014 (UTC)

• I think that it's time to reduce the unnecessary and duplicative bureaucracy. The AUSC's current role was always duplicative of the m:Ombudsman commission, which was rather moribund in 2008-09, but is now a thriving and active - and perhaps most important, completely impartial - committee. There has not been a removal of tools "for cause" in several years, only removals for inactivity, and that can be handled by writing activity standards into the CU and OS policies and enforcing in a no-exceptions manner that any member of the community could post with links to the published activity report and the applicable policy.

  There are few enough interested and qualified candidates for Checkuser and Oversighter roles. It is a waste of resources to have those people doing bureaucratic tasks, which in this case would still amount to almost no work at all. Granting more authority doesn't change the fact that there aren't enough cases to justify having the AUSC in the first place. Let's just turn the current community AUSC members into CU/OS, refer any complaints to the Ombud commission, and move on. Risker (talk) 18:08, 25 May 2014 (UTC)

The one positive side-effect of appointing new AUSC is that many of them would eventually stay on to take on the normal role of CU/OS, thereby increasing the CU/OS population indirectly. Many of these appointees are new to CU/OS privileges, and it is always a good idea to bring in new blood whenever possible so as to bring in a fresh pair of eyes and to prevent being perceived a running some form of Cabal. In a way, AUSC does keep the existing CU/OS on the toes. - Mailer Diablo 20:55, 25 May 2014 (UTC)

Well, in recent years, a term on AUSC (which always has few applicants) has been seen as a more effective means of getting CU/OS permissions than applying for CU/OS directly. We've mostly added a "get the tools easily" loophole here; there's never been an occasion where AUSC members have been turned down for CU/OS after their term. Risker (talk) 21:27, 25 May 2014 (UTC)

If we're serious about turning AUSC into an independent body, we could always allow them to use the tools during the subcommittee term, but bar them from standing until after they leave AUSC. It'd be easily done. AGK [•] 22:10, 25 May 2014 (UTC)

• Risker, the O.C. is not an appropriate substitute for AUSC because it may not enforce (the stricter) local policies. We also have no idea who will end up on it in the future. And please read the concrete proposal I subsequently made, which does not require the CU/OSers to do anything differently. AGK [•] 21:41, 25 May 2014 (UTC)

You think that it is appropriate to have a shorter voting/commenting period for checkusers and oversighters than for administrators or bureaucrats? And we don't have stricter local policies; in fact, this project is by far the most liberal in interpreting the global CU and OS policies. Risker (talk) 21:52, 25 May 2014 (UTC)

Um, the voting and comments period is seven days, which is the length of time an RFA or RFB would run for. Adding the Q&A period makes it far longer. Plus, as I've said already, none of this is in any sense finalised.

No, it isn't. It's only liberal by comparison to other projects' policy. The distinction is important. AGK [•] 22:15, 25 May 2014 (UTC)

"Its only liberal by comparison to other projects' policy" - yes, that's exactly my point. You've suggested we've got stricter interpretation and I am contradicting you. Who else would I compare our project to but other projects? Incidentally, there's nothing in the Ombudsman policy that prevents it from enforcing stricter local policies, and I understand that it is normal for them to do so. Remember that the vast majority of projects that have checkusers have nowhere else to take concerns, and some local policies actively separate Arbcom from any authority or responsibility in relation to CU/OS. Risker (talk) 22:57, 25 May 2014 (UTC)

I said (the stricter) local policies; that is, our local policy is stricter than the global one. As you seem to agree, I do not see the contradiction.

I have no interest in inviting an unelected Ombudsman, which will be comprised in the future by unknown people from Enwiki and people who have never edited here, to audit CU & OS on behalf of the community. Keeping ArbCom's ability to audit and revoke CU & OS (or, in our case, to allow its delegates to do so) is only possible if ArbCom continues to ceremonially appoint the functionaries. AGK [•] 19:35, 26 May 2014 (UTC)

Well, I'm saying that I have no problem with referring to a completely independent body for review; in fact, I considerably prefer it, having sat on the AUSC for about 30% of the last five years. I think that our local AUSC, when it was getting lots of requests in 2009 and 2010, was largely ineffective because of the inactivity of community members - we had to resort to alternates two years in a row because of this - and now it doesn't receive enough requests to justify its existence. I'm also NOT saying that enwiki has a stricter local policy. In fact I am saying that enwiki has a far more liberal local policy, and possibly the most liberal interpretation of the global policy, of just about all projects. We *do* have a much larger CU cadre per number of edits made of all the larger projects. We *do* conduct more checks per number of edits made than any other large project. We also are the project that attracts the most spambot edits, the most socking vandalism, and we are the community most likely to request checkuser (via SPI, which doesn't even exist on most large projects); there's a reason why we do checks more frequently. (There are also philosophical differences between the projects, with dewiki having full flagged revisions on everything for example, and others just indef-block any suspected sock accounts.) None of this tells me that Arbcom needs to be involved at all; the only value that Arbcom adds to the process is in vetting candidates, and that's the step you want to eliminate. If Arbcom leaves the scene, it still has the ability to sanction any editor as appropriate to the circumstances of the case (i.e., there's nothing that stops Arbcom from removing CU/OS because of evidence provided in a case), and there can be automatic enforcement of activity rules based on the currently published activity. (It was intended that AUSC come up with that data, but it's pretty much always sitting CU/OS who do so.) Risker (talk) 04:00, 19 June 2014 (UTC)

• To quote "Candidates submit a candidacy by self-nominating at RFCUOS, and a one-week question and answer session follows. Any registered editor in good standing may ask a question of a candidate; candidates are encouraged but not required to respond. Two days for comment will ensue, the Arbitration Committee will be notified by the candidate of the candidacy, then a five-day voting period will commence; registered editors in good standing with at least 500 edits may vote once, in support or opposition of the candidacy. Neutral votes and attaching comments to votes is prohibited." So...
  • Self nominations at any time, whether or not there is a need for further checkusers/oversighters.
  • No co-ordination to actively seek out candidates at times when there is a strong need.
  • Only five voting days. The question and comment period is irrelevant, and you give no reason for not permitting them to happen concurrently with voting. Five voting days is insufficient for any post, and particularly for one that involves access to private information.
  • Why does CU/OS have a higher editing requirement than voting for Arbcom? Risker (talk) 22:57, 25 May 2014 (UTC)

Thanks for your comments and for looking at the details of the proposal. Don't forget that the proposal is a very early draft, so criticising the minutiae – rather than the thrust, which is what I meant when I said "can you read the proposal?" – may not be a great investment of your time. AGK [•] 19:35, 26 May 2014 (UTC)

• Given Guerillero's position, which seems totally reasonable, perhaps the AUSC is better populated with people who do not wish to spend a lot of time on Checkuser and Oversighting, but are prepared to put considerable effort into irregular cases, and to produce better audit information for the community. All the best: Rich Farmbrough, 19:32, 27 May 2014 (UTC).
  
  • I wish there were cases to put time into, Rich. --Guerillero | My Talk 02:43, 29 May 2014 (UTC)

Administrator prerequisite

The draft says "Non-administrators are eligible". My understanding from Philippe is that WMF requires that users have passed an RFA or similar process like an Arbitration Committee election to receive access to the CheckUser and Oversight tools. Would the community process outlined in this draft be acceptable to WMF for a non-administrator to become a CheckUser or Oversighter? --Pine^✉ 22:59, 25 May 2014 (UTC)

I would assume so, but WMF would have to ratify any final proposal, obviously. NE Ent 23:05, 25 May 2014 (UTC)

I think that this would be good to confirm now, so that people don't get their hopes up. However, there is the fact that at least for oversight it is impossible to properly do without full admin permissions. In particular, it is not possible for a non-admin oversighter to do suppression-deletion, because it requires access to admin tools. Risker (talk) 23:21, 25 May 2014 (UTC)

Would it be reasonable to do wider recruiting in the existing admin corps, especially for Oversight? Most administrators with good track records with revdel are probably suitable as oversighters. However, I think there might be a problem with giving many more admins access to functionary email lists with confidential information that can be leaked. There may be a way to unbundle the suppression tool from the functionary email lists, in which case I think the suppression tool and access to the oversight queue could be trusted to many more people who have already passed RFA. Personally I would be willing to serve as an oversighter, but I'm in no hurry to get hazed at RFA and answer lots of questions about subjects that I have little interest in like AFD just so I can get access to the oversight tool. --Pine^✉ 23:36, 25 May 2014 (UTC)

Either oversighters are functionaries or they aren't. I think it's rather insulting to suggest that someone trusted enough to remove private information from visibility is not trusted enough to participate in discussions as a functionary; the role is based on trust to maintain the confidentiality of private information. As already noted, one can't really be an effective oversighter without having the delete button. Risker (talk) 00:42, 26 May 2014 (UTC)

I agree. The sensitive stuff Oversighters see is mostly in the diffs they suppress, not in the emails they read on the functionaries list; and I would be uncomfortable creating "Oversighter lites". AGK [•] 19:47, 26 May 2014 (UTC)

@Risker: no insult was intended of existing oversighters. I was asking about how to lower the risk of increasing the pool of oversighters to many more people. @AGK: ok thank you. --Pine^✉ 06:57, 27 May 2014 (UTC)

The WMF has said before that we believe that Arbcom members who are elected have passed an RFA or RFA-identical process sufficiently to grant rights. Provided that this election is adequately advertised, adequately policed and patrolled, and that it is a sufficiently broad vetting, it is my belief (though unconfirmed as of yet) that the legal team would also believe it to be RFA or RFA-identical.  :) Philippe Beaudette, Wikimedia Foundation (talk) 23:57, 25 May 2014 (UTC)

@Philippe (WMF): about Risker's point above about oversighters requiring revdel, I thought that granting someone access to OS automatically granted them access to revdel. Would a non-admin oversighter also have access to revdel? I'm not saying that this is a good idea from a policy point of view because I think it makes sense to have someone with revdel experience become an oversighter. --Pine^✉ 07:01, 27 May 2014 (UTC)

Off the top of my head, I'm not sure, but I think I know someone who'll know, and therefore save me the trouble of checking (I'm very lazy, you see...) @Jalexander:? Philippe Beaudette, Wikimedia Foundation (talk) 08:32, 27 May 2014 (UTC)

@Pine and Risker: Based on Special:ListGroupRights, Oversighters can delete revisions (both page histories and log entries, granted by deleterevision & deletelogentry rights). From what I can see the only things they can't do are delete pages (delete right; which I don't see much of a need for), undelete pages (undelete; such restoring prior to supression), as well as delete (flow-delete) and hide Flow topics and posts (flow-hide). I'm not sure how much affect those two rights for Flow will have as I'm not familiar with how it'll work in practice. Callanecc (talk • contribs • logs) 11:09, 27 May 2014 (UTC)

It is something done very commonly, Callanecc, probably about 25% of my suppressions are delete-suppressions, particularly for pages involving personal information of minors where there is specifically no desire to have the page remain in existence. Risker (talk) 11:36, 27 May 2014 (UTC)

@Philippe (WMF) and Pine: As Callanecc said above it does indeed grant revision delete as well automatically (it is a separate right but is currently bundled in the local group). It would essentially have to be, without it they are unable to access the revision delete interface at all and so would be unable to suppress any already deleted revision so stuck to only suppressing log entries (and users if and only if they were able to block). Jalexander--WMF 11:55, 27 May 2014 (UTC)

@Jalexander: could the page deletion, page undeletion, flow deletion and flow hide rights be bundled with oversight? If so I think oversight could be usefully granted without admin rights. --Pine^✉ 06:59, 28 May 2014 (UTC)

@Pine: Yup, in theory those (or any rights) could be added. Just on wiki consensus/decision and then a bug filed. Assuming all of the other requirements are met legal happy etc, but as Philippe said above I think those are surmountable with RFA type elections. Another thing to remember is that the global CU and OS policies require at least 20 support votes I believe and those would apply (but if we're doing an RFA type election on enWiki I doubt that would be an issue). Jalexander--WMF 08:03, 28 May 2014 (UTC)

Thanks. @Callanecc: and @Risker: see above. As I said before I think it makes sense for someone to have revdel experience before becoming an oversighter, but we don't require revdel experience before someone becomes an admin, so perhaps the experience is beneficial but not necessary. The community could decide on a per-nomination basis. I expect most people who pass will be administrators but there may be exceptions, and if Arbcom thinks the community can be trusted to background check the applicants with at least the same level of scrutiny that Arbcom does while creating more public transparency in the voting process, and with a requirement that there be 25 support votes, I can support to AGK's proposal for this process for appointing oversighters. If it doesn't work we can revert to today's system. --Pine^✉ 08:13, 28 May 2014 (UTC)

Well, I for one would be opposed to granting essentially one of the most powerful tools in the admin toolbox to users who haven't gone through RFA, if for no other reason than that it creates a situation ripe for problems. For example, can an oversighter revision-delete something? Under what circumstances? What about deleting content? Under what circumstances? I can't intellectually justify allowing non-admins to revision-delete or page-delete, oversighter or not, but that's exactly what they'll be able to do. Risker (talk) 02:12, 9 June 2014 (UTC)

• @Risker: It seems to me that this problem is the lesser of two evils and that any non-admin who passes would be expected not to use revdel for purposes other than oversight. If we don't allow non-admins to run for oversight, how do you suggest we address the problem of the shortage of oversighters? --Pine^✉ 07:31, 20 June 2014 (UTC)

What shortage of oversighters? There is no shortage of oversighters, and there hasn't been for years. There's poor organization amongst oversighters to work together to develop some kind of schedule and assume responsibility for responding to requests. There's also unreasonable community requests - I've had people go ballistic when they don't get their request addressed within half an hour, when the [unwritten] standard is "within 24 hours", and at that we're still far more efficient than the vast majority of projects. But there's no shortage of oversighters - even if none of the arbitrators ever uses their tools. Risker (talk) 11:32, 20 June 2014 (UTC)

• @Risker:, how would you respond to Fluffernutter's comment above? She wrote "The oversight team is suffering from a serious lack of manpower (see the bottom of Wikipedia talk:Oversight for evidence of this)..." --Pine^✉ 07:22, 21 June 2014 (UTC)

Pine, I'd say that really the issue is more in figuring out a better way to ensure that the OS role is "staffed". On any given month, there are about 400 suppressions done by about 20 people. That only works out to an average of 20 suppressions per active oversighter per month. We could also look at setting up a calendar where each oversighter "signs up" for a handful of 12-hour shifts per month, during which they'll be responsible for clearing OTRS requests for oversight. Remember, the standard isn't "suppress as soon as the email arrives", it is "addressed within 24 hours". Less than half of all OS actions are taken based on OTRS requests; many of them come through IRC, through PMs to oversighters, mainly relating to excessive personal information by self-identified minors. That is by far the largest category of suppression requests. I suppose the other issue is that some oversighters respond to the OTRS queue extremely rarely, and often the same people don't use IRC, so I have no idea how they receive their suppression requests. Risker (talk) 07:57, 21 June 2014 (UTC)

I think it is an excellent principle that adminship not be a prerequisite for anything. I have no issue with an "RFA type" requirement, nor with identifying to the foundation - or indeed public identification, and apart from the last I doubt the community would either. All the best: Rich Farmbrough, 19:26, 27 May 2014 (UTC).

CheckUsers

My sense is that administrator experience is not necessary for someone to be a proficient CheckUser, so I support this proposal for appointing CUs under the same conditions as I discussed for oversighters. --Pine^✉ 08:43, 28 May 2014 (UTC)

Main reason is that they wouldn't be able to block accounts or IP addresses/ranges which is a reasonably big part of the job nor would they be able to delete pages created by socks. I can't really see a great likelihood that there would be consensus to give the block or delete (plus associated rights for both) to non-admins, but that's just my feeling. In terms of experience, there are very few non-admin users who have broad experience in investigating sockpuppetry and then making decisions based on that, primarily because the decision generally needs the admin tools (mainly block). Callanecc (talk • contribs • logs) 12:27, 28 May 2014 (UTC)

Hi Callanecc, the same argument could be made for oversighters, that oversighters who lack the ability to block users for posting private information would be impractical. While CUs and OSes without the ability to block users would require them to communicate with admins to create blocks, indirect blocks already happen when arbitration sanctions are enacted or when a non-CU admin makes a block at SPI based on CU confirmation. I agree that having CUs and OSes without the ability to block would be an inconvenience in some ways but with more people working the CU and OS queues the response times should be faster on average. --Pine^✉ 07:13, 29 May 2014 (UTC)

@Pine: The difference is blocks are a regular part of a CheckUser's job description. A non-admin CU would only be able to perform some functions of the job. Primarily, posting results to SPI pages or making decisions in ACC requests. Whilst this would be helpful it still means that after doing a check they would need to find an admin CU to carry out the result (blocking an IP or range). Callanecc (talk • contribs • logs) 06:22, 31 May 2014 (UTC)

@Callanecc: I think some help is better than none, and the community should be allowed to decide on a case by case basis. Also, non-admins who successfully become CUs and demonstrate proficiency as a CU will build credibility for an RFA. I agree that having admin-CUs and admin-OSes is best and most users who pass this process will likely already have passed RFA, but I think a lack of admin tools shouldn't prevent someone from helping using the tools that they get if the community is willing to give access to those tools. --Pine^✉ 06:27, 31 May 2014 (UTC)

I don't think that it's actually all that helpful, though. For example, I won't block IPs for other checkusers unless I've verified the results myself - after all, it's my username on the block, and I have to justify it. That means double work (and twice as much privacy invasion, for those who feel there are too many checks as it is). For those who believe that non-admins should be perfectly fine with this task, think about whether you'd be willing to have the block right added to the CU group of rights. If the community is okay with that - and I wouldn't bet on it either way - then at least they'll actually be able to do the job. Risker (talk) 04:09, 19 June 2014 (UTC)

• @Risker: I have thought about that issue, and I think you're right that in the case of blocking the person who performs the block should perform a check before doing so. However, non-admin CUs could still usefully use the CU tool and close cases where there is not adequate evidence to support a checkuser block in their opinions. --Pine^✉ 07:31, 20 June 2014 (UTC)

• If there isn't evidence to support a block, the Checkuser tools shouldn't be used at all. The only reason they exist is to provide technical information in limited cases of editor abuse. Abuse is always blockable. Under most normal circumstances, there is no reason to use the CU tools unless a block is one option on the table. While a CU will often just make the technical link and the admin has to take their word for that portion, just as often they have to privately block IPs or ranges without connecting it publicly to that sock (WP:OUT concerns). They would have to get another CU to look at the case, then make the private blocks, doubling the man hours to get the job done. A CU without admin tools is more of a burden and hindrance than a helper. Dennis Brown | 2¢ | WER 13:24, 21 June 2014 (UTC)

Review of admin tools

Right - I've spent several years under the presumption that the committee was the place for review of tool use by admins if any editor was concerned by an admin's use of same (or conduct etc.) was and that this bypassed (or trumped) option of Last Resort. Turns out not to be the case. I think it'd be good for the committee to assist in an algorithm here as I think the current means of admin review are otherwise inadequate if the committee is not part of the process.

• Hence I think it would be prudent if the committee were to come to a consensus about whether the committee has a special role in review of admin tools.

• If it decides it does not, then I think an opinion on review of tool use is needed, but the committee's opting out would be a good igniting point for the review.

• Thus, if the committee steps back, the community can discuss on whether current means are adequate or whether we revisit one of several proposals.

Thoughts? Cas Liber (talk · contribs) 13:05, 26 May 2014 (UTC)

Is the 'algorithm' not fully documented at WP:GBU? It was never the case that concerns about sysop actions went straight to arbitration. AGK [•] 19:40, 26 May 2014 (UTC)

The idea behind the failed policy initiative WP:Request for Admin Sanctions was because WP:AN/WP:ANI are adequate for single issues such as block review or WP:INVOLVED concerns, but not for dealing with long term problems. Additionally, WP:RFC/U is toothless. Of course, there are many other proposals, most of which seem to deal with desysoping only, but regardless, there does seem to be a void between WP:AN and WP:ARB. Dennis Brown | 2¢ | WER 20:02, 26 May 2014 (UTC)

(edit conflict) Amusing. I've been here seven years and never seen that acronym. In fact, I don't recall seeing the little segments it links to, but I must have as it has been there several years. This was even when I was on the committee and we've been discussing this very issue. I recommend a succinct one-line statement identifying the pathway be put in the Removal of permissions as a footnote/addendum. I must add I agree with Dennis in the issues with AN, and the whole structure of RfC/U being inept to deal with the discussion (or many discussions on users) Cas Liber (talk · contribs) 20:08, 26 May 2014 (UTC)

Ah, yes, the algorithm is ineffective when the sysop does not abide by the decision reached in the earlier stages, but I believe that in turn is grounds for an arbitration case anyway. I suspect "reverse RFA" desysopping processes have never came to fruition because there is a clear process on Wikipedia for dealing with abusive sysops. In other words, I don't think this is a process that needs reformed; the community knows that ArbCom will deal with bad sysops. AGK [•] 20:42, 26 May 2014 (UTC)

From my experience there and with similar policies, I've found many liked the idea in theory but quickly cooled to the idea once you started developing the actual process. The devil is always in the details. There is still a non-trivial minority that wants it, but I don't see a consensus for it happening any time soon. Dennis Brown | 2¢ | WER 22:37, 26 May 2014 (UTC)

AGK, I suspect linking to a former admins page does not fill readers with a huge amount of confidence. I came up with User:Casliber/Fivecrat in 2008. It is worth throwing it out there. Fact is we can have Open to Recall,,,,and folks can just ignore it. Cas Liber (talk · contribs) 23:07, 26 May 2014 (UTC)

Like Dennis, I don't see a consensus for it, and don't particularly think it's a pressing issue myself, so I'm not going to push for it. Others can, if they wish. AGK [•] 06:43, 27 May 2014 (UTC)

Code of Conduct

I guess this may be the appropriate venue to raise this, as it relates to procedures. I think it would be very helpful for Wikipedia as a whole if the Arbcom could develop a code of conduct that users are expected to adhere to when directly addressing the ArbCom, and in other dispute-resolution venues. This would, I feel, go a long way to reducing the tainting effect of conflicts brought to these venues that exacerbate in the venue, and the nastiness trickles down from these venues throughout WP. --LT910001 (talk) 10:36, 27 May 2014 (UTC)

Maybe WP:NPA would cover most of it? All the best: Rich Farmbrough, 19:20, 27 May 2014 (UTC).

Thanks, but I am specifically requesting a code of conduct for conduct within the dispute resolution venues, which at the moment seem to be regarded as protected places for conflict. --LT910001 (talk) 21:51, 27 May 2014 (UTC)

Raising issues

Specifically useful would be a defined on-wiki way to raise matters that are currently only have mailing an email list as the procedure. It may require technical glue, for those functionaries that don't log on to Wikipedia very often. All the best: Rich Farmbrough, 19:20, 27 May 2014 (UTC).

So we can better consider this, could you give some examples (subject to WP:BEANS and so on)? AGK [•] 21:47, 27 May 2014 (UTC)

• " For other matters, the arbcom-l mailing list is the primary venue for internal Arbitration Committee discussion on all matters under the Committee's remit. " (No other venue is given.)
• "The arbcom-audit-en mailing list is used by the subcommittee for internal discussion, communication with functionaries who are the subject of investigation, and to be a venue for the receipt and resolution of complaints." (Again no other venue is given.)
• "To appeal your block or ban, send an e-mail to the subcommittee through Special:EmailUser/Ban Appeals Subcommittee." (Here, of course, the appellant is unable to post anywhere but their talk page, if that. They should however, be able to have their appeal on record if they wish.)

That's the three I am aware of, there may be more. All the best: Rich Farmbrough, 13:58, 30 May 2014 (UTC).

The arbcom mailing list is for arbcom's internal discussions. If we have those discussions in public, they are no longer internal discussions, they are public discussions where arbs would be unable to mention anything confidential or indeed to speak their minds freely about certain issues. The arbcom mailing list has been made out to be this sinister thing over the years, but I can tell you from experience that most of it is just internal coordination, trying to come to an agreement on how to proceed with our public actions. I have endeavored to help introduce more transparency in committee proceedings, and I think the majority of current arbs are pretty willing to explain themselves and to have public conversations about matters that ought to be public, but the list remains a very important tool for getting the committee's business done.

AUSC is by definition something that must operate off-wiki. Confidential information is always involved.

BASC is the avenue of last resort for users who have exhausted the patience of the community. It is not and should not be a venue for blocked users to grandstand. Indeed, the lack of an audience seems to help some users calm down and really see what it is they were blocked for or conversly for really nasty users to reveal just how depraved they really are. In only six months on the subcommittee I have seen both more than once. As almost everyone who appeals to BASC has already had three or four public appeals it seems to me to make perfect sense that the final avenue be something different. Keeping BASC appeals private can actually protect the user as well as preventing further disruption f the appellant is a vandal/troll/stalker/etc. Beeblebrox (talk) 03:29, 24 June 2014 (UTC)

Permissions for former arbitrators

"Arbitrators who are checkusers or oversighters ex officio will lose the permissions on leaving the committee." What is the reason for this? If there are too few functionaries already then don't we need these people as well? Thanks, --Pine^✉ 08:34, 28 May 2014 (UTC)

Historical artifact, I think. At the time the rule was established, there was a lot of controversy over ex-arbs with holdover permissions from their time on the committee. Many of them ended up being among the small number of functionaries whose rights have been forcibly removed. If I remember right, this issue was a major topic when the concept of functionaries (and access to various private mailing lists) was first discussed. Nathan ^T 01:17, 29 May 2014 (UTC)

OK. Could the draft include some statement that arbitrators who want to retain their permissions after leaving Arbcom can and must go through this process if they didn't have these permissions prior to joining Arbcom? --Pine^✉ 07:08, 29 May 2014 (UTC)

• Part of the reason that it's an historical artifact is that almost all successful Arbcom candidates for the last few elections are either checkusers or oversighters or arbcom clerks or former arbitrators before they are elected. Indeed, many feel that holding one or both of those hats is almost a prerequisite to a successful arbcom candidacy. They're probably right; other than a few individuals whose admin work is widely known and respected, CU and OS are pretty much default choices since they've already jumped through the "access to private information" hoops. The other historical artifact involved was that there was this veneer of mystery associated with checkuser in particular, implying that it required extensive technical knowledge in order to for the candidate to be useful; in fact, it's a pretty simple tool to operate, and the bigger challenge is identifying the patterns in socking behaviour. Risker (talk) 02:21, 9 June 2014 (UTC)

"Neutral votes and attaching comments to votes is prohibited."

As someone who participates in the FPC process I find neutral votes and comments to be valuable. Is there a reason for prohibiting them in CUOS votes? I am especially opposed to prohibiting the attachment of comments to votes because I think users will want to know why other users voted as they did. --Pine^✉ 08:38, 28 May 2014 (UTC)

It makes CU/OS elections real elections and not a psudo-discussion --Guerillero | My Talk 02:41, 29 May 2014 (UTC)

How will editors be able to discuss evidence regarding the suitability of candidates? Sending everyone to the talk page would be inconvenient. --Pine^✉ 07:07, 29 May 2014 (UTC)

How do people discuss evidence regarding the suitability of candidates for arbcom? --In actu (Guerillero) | My Talk 14:22, 29 May 2014 (UTC)

That is a good question. There are Questions to Candidates, of course. I have missed the last two elections, so I am not up on the procedure. All the best: Rich Farmbrough, 14:00, 30 May 2014 (UTC).

Actually, there's very, very little discussion of the suitability of candidates for arbcom, and there hasn't been much of it ever since we went to electing Arbcom by SecurePoll. Before that, the talk pages of the voting pages of Arbcom candidates was chock full of discussion about the candidate's suitability. The questions are pretty meaningless; very few people read all of them (mostly those who create "voting guides" although even amongst them, many openly base their votes on only a few questions); they're mostly a measurement of candidate stamina. But actual discussion about the suitability of candidates, their qualifications, or the makeup of the committee as a whole has gone the way of the dodo; it was last seen around here about 2008. Risker (talk) 01:17, 1 June 2014 (UTC)

Why do we want real elections and not a discussion? Hawkeye7 (talk) 07:24, 1 June 2014 (UTC)