Wikiversity talk:Custodianship/Archive 3

RfC - Amending "Mentorship"
Proposal: Add "Wikiversity participants have one week to question and comment on the candidacy. If after one week, there is no community consensus against allowing the candidate to become a probationary custodian and the candidate can find a mentor, the Mentorship period will begin. Probationary Custodians need to have the community's trust. Mentorship is not a right." to I. Request.

Proposed by Ottava Rima (talk) 19:22, 11 January 2011 (UTC)

Support

 * 1) Ottava Rima (talk)  19:22, 11 January 2011 (UTC)
 * 2) Support with the changes I proposed below. -- dark  lama  00:40, 13 January 2011 (UTC)
 * 3) Support per the changes below. At first I held back any vote but Darklama's modification convinced me not to oppose. Kevin Rutherford (talk) 02:10, 13 January 2011 (UTC)
 * 4)  Good changes by Darklama. Guido den Broeder 13:58, 14 January 2011 (UTC)
 * 5) Support per above. Diego Grez 17:35, 18 January 2011 (UTC)
 * 6) Support per User:Ottava Rima's arguments. --Anonymous Uploader 23:06, 19 January 2011 (UTC)

Oppose

 * 1) Strongly oppose. Mentorship is specifically a method to allay community concerns over possible abuse by an inexperienced or possibly controversial custodian. There is no example shown of a problem from routine immediate action on a candidacy/acceptance. If there is abuse of the process, we can address it. No such abuse has ever been shown. Mentored adminship without a necessity for a community discussion is, in fact, one of the best Wikiversity features, it shouldn't be trashed. The suggestion requires an extra close, to determine "community consensus" on going ahead. If any scholar has been involved in controversy, with anyone holding a grudge, it is guaranteed that there will be "concerns" and, likely, tendentious debate. Having one such discussion per candidacy is enough! --Abd 20:27, 11 January 2011 (UTC)
 * If mentorship is to allay community concerns, then you suggest community concerns are important. Thus, if the community has concerns before hand, it should allow for no access at all. Otherwise, anyone no matter how abusive could effectively get ops. This happened with Salmon of Doubt and happened with you. Both by the same Crat with a history of abuse. Ottava Rima (talk) 23:18, 11 January 2011 (UTC)
 * Agreeing to mentor is a custodian action, only custodians can do it, and it often occurs that a custodian suggests mentorship to a possible candidate. Any custodian who acts contrary to consensus is subject to review and sanction. Thus the proposal here is really a curtailment of custodian rights. If there were an example of abuse, then we should look at it. Below, I ask for examples of problems, none have been suggested there. So the two mentioned are covered here, this can be summarized below if someone still wants to assert them.
 * 1) Oppose, because I think we're a bit premature for the "voting phase". I'm also a bit concerned that the changes suggested would lead to two "RfAs" in the place of one. --SB_Johnny talk 23:47, 13 January 2011 (UTC)
 * Shouldn't you be concerned that the current policy gets around WMF standards of needing consensus before having ops? Furthermore, only concerns would be voiced - if there are no concerns then there is no problem. Why do you feel that admin have the right to ignore community concerns? The Custodian and Bureaucrat policies here suggest that admin should respect community consensus. Are you saying that you no longer feel like you are able to uphold these requirements? Ottava Rima (talk) 00:19, 14 January 2011 (UTC)
 * If the community opposes the request/nomination there would be no Probationary period to begin with and the Full Custodianship discussion phase would never be reached. Since support would continue not to be required, there would also be no need for discussion before the Probationary period starts, unless there was opposition. I think brief discussion already tends to happen before a candidate finds a willing mentor. I think discussions would continue to be brief because anyone opposing the request/nomination would have only one week to argue their case and convince the community to oppose. I think a lack of consensus would also not be able to stop the probationary period from happening either since it wouldn't be a consensus in opposition to the probationary period. I think this means that the burden would be on the ones opposing to argue why the candidate should not be allowed to have a Probationary period and to establish a consensus before moving ahead with the Probationary period can be stopped. I think this would also mean in most cases there would only be a need for the community to decide once, unless the community is divided. -- dark lama  14:46, 14 January 2011 (UTC)
 * I guess I'm concerned about how it would work in practice. Would the 7 days start from the moment the request is made, or perhaps after a mentor is found? If there is opposition, would we list it on sitenotice? It seems like yet another invitation to drama mongers, which goes against the whole point of using the mentorship method in the first place. I'd be happier just going with RfA than having 2 votes like this.
 * OTOH, the natural companion to KISS is ABDF (Ain't Broke, Don't Fix). As far as I know, we've never actually had an issue with a probationary custodian actually going rogue or causing damage. We have had people request permissions who were unable to find a mentor, and a few where the mentor didn't recommend and/or the mentored person disappeared, but that's low-drama as well. --SB_Johnny talk 17:51, 14 January 2011 (UTC)
 * I think those are good questions that should have an answer as part of the addition. I think the 7 days shouldn't officially start until a mentor is found, because without a willing mentor they won't be able to become a Probationary Custodian anyways. I think a sitenotice should generally only be used as a way to deal with a lack of consensus when a mentor has been found, so maybe at least three established users must oppose, and consensus must be mixed enough that it could go either way. -- dark lama  18:13, 14 January 2011 (UTC)
 * Well, that sounds better, though perhaps in that case it should be 7 days after the listing on the sitenotice. I'd also say that the 3 established users should explicitly oppose, rather than be "interpreted as opposing" (again, to reduce the window for drama-mongering).
 * Setting a baseline for who is "established" would actually be a good idea too, since we actually had a problem with that on a confirmation several months ago (a custodian was not promoted due to too many opposing, but not all of the opposers were by any means "established" on WV). --SB_Johnny talk 18:28, 14 January 2011 (UTC)
 * I think a user should be considered established if they have contributed to at least 100 times to one or more learning projects, and have contributed at least 20 times to one or more learning projects in the past 30 days. I think the 7 days is intended to define how long a Bureaucrat must wait before flipping the bit. I think that would be lost if we made it 7 days after listing on the sitenotice, not to mention there may be no need to list it on the sitenotice at all. Maybe just extend discussion by another 7 days if there is a need to list it on the sitenotice? I'm not sure about an explicitly oppose to count. A bureaucrat would be the one flipping the bit, so should read all the discussion and see if there are any unanswered objections and any need to seek feedback from the wider community. -- dark lama  19:01, 14 January 2011 (UTC)
 * Well, as far as the explicit part, and speaking as one of the 'crats who would be doing the flipping, I can say from experience that having things clearly defined would make the job a lot more comfortable. I'd also say that 20 contribs to any namespace would be fine (including logs, welcoming new users, etc.), or perhaps the project namespace instead of main (since "matters of concern" are more likely to stem from things in project). --SB_Johnny talk 19:41, 14 January 2011 (UTC)
 * Wikiversity is for participants to learn and teach, and the community can be seen as those people who engage in learning and teaching. My suggestion of requiring contributions to learning projects is about ensuring the people that are counted actually represent Wikiversity and the community rather than about contributing to specific namespaces. -- dark lama  15:23, 17 January 2011 (UTC)
 * Fair enough, but since discussion and debate are actually part of the scope here, I don't see any reason not to count the others. I also think one could probably run up 100 edits rather easily using scripts and such, which are fine but perhaps wouldn't establish one as part of the community as well as 100 additions to discussions. --SB_Johnny talk 18:30, 18 January 2011 (UTC)
 * 1) Oppose - The current system seems fine to me. Some may not give my view much credence, as I edit here rarely if at all. Which is also fine but I did want to express my view. ++Lar: t/c 18:51, 14 January 2011 (UTC)
 * 2) Oppose - Read the policy and see no problem the change would fix. &rarr;StaniStani  21:16, 14 January 2011 (UTC)

Discussion of Opposition
Alleged examples of abuse of process:


 * Salmon of Doubt. See Candidates_for_Custodianship/Salmon_of_Doubt.

Background. Salmon of Doubt has been accused of being a certain controversial Wikipedia user. Regardless of the truth of this, reviewing the activity of Salmon of Doubt, he was highly disruptive here, being one of the factors leading to the Great Debacle of 2008. His alleged WP account is one whose harassment of me led to the ArbComm case that resulted in the loss of an admin bit by an admin who cooperated with him, but he "retired," a trick which often works at Wikipedia, and was protected, by the admin who was desysopped, through revert warring during the ArbComm case. ArbComm was astonished by some of the behavior during that case. I wasn't. And ArbComm didn't notice much of it. I mention this because one might think that I'd have been upset about this custodianship. I didn't know about it at the time, but if this had happened after I was active here, I'd have been concerned, but not after seeing what SBJ did.


 * When I became aware of SoD's brief term of adminship, and saw what SB_Johnny had done, I was in awe. It was an example of how a skillful 'crat could sidestep disruption, and this made the loss of SBJ during the Debacle of 2010 all the more visible as a loss, hence my encouragement of SBJ to return.
 * Concerns were expressed about SoD. However, SBJ offered to mentor. Any sysop may do this, and it doesn't require that one be a 'crat. I thought that SBJ had himself pushed the button, but, I was incorrect in that. Mu301 was the acting 'crat. So, if we look closely, if there was abuse here, it would be Mu301, for "ignoring concerns." However, the procedure was clear, and was followed. The normal time for an evaluation of concerns is in the full custodianship !voting. Before that, concerns may elicit promises from the candidate, specific agreements for fast desysop, etc. In reality, these are concerns that are for the benefit of the mentor, so that the mentor can be specially watchful. If there is concern that a mentor will not be able to adequately supervise the candidate, then that should be raised for the attention of a possible implementing 'crat. A truly inactive custodian should not mentor.
 * SBJ, offering to mentor, stated conditions: I will mentor on two conditions: (1.) you will act in exactly the manner you described above, and (2.) you will leave an edit on my talk giving me clear authority to ask a steward to revoke your tools without discussion, reason, or explanation. SBJ made sure that SoD made aan edit that could be pointed to when going to meta, and SoD complied. The mentorship offer was at 20:04, 12 September 2008 (UTC), and Mu301 set the bit at 02:23, 13 September 2008. There were, prior to the decision by Mu301, "concerns expressed," and no support for SoD. The only specific advice given to SBJ about SoD, AFAIK, was by Moulton, who was involved to his eyeballs and was on a short track to a global ban at that point, and JWS was about to be desysopped, another train wreck. Nevertheless, Mu301 approved.
 * The proposed rule would have made this impossible, there would have been a disruptive and highly controversial candidacy. In the end, SoD lost interest quickly. It's boring being a WV sysop if your every move is watched by someone who isn't going to tolerate disruption, it's not at all like Wikipedia, where there is so much going on that it's impossible to watch activity unless it is very important to you. No harm was done. Period.
 * There were no admin actions by SoD after September 14. Then SBJ suggested that he "hold the keys" while SoD was inactive, see . This was immediately done at meta, September 18. In other words, this "abusive person getting ops" amounted to ops for five days, with a handful of actions, and I've seen none of them that were alleged to be harmful. Highly efficient. The only wasted time there was in the discussion that Ottava now wants to require and expand to five days minimum.


 * Abd See Candidates_for_Custodianship/Abd_2


 * Ottava is claiming that because my previous custodianship was unsuccessful, and because he was my mentor the first time, and disapproves, I'm barred from applying again. This is blatantly contrary to policy, and the change proposed here would not prevent it, and making any individual have absolute veto power is obviously a bad idea.
 * In this case, my candidacy was opened 04:13, 26 December, 2010. I already had a mentor who had agreed to serve, Jtneill, but, as it turned out, he was on wikibreak. The candidacy attracted comment from:
 * JWSchmidt, whose comments were totally unsurprising to any who know what he's been doing for over two years now. I responded in detail.
 * Ottava Rima, who, when he's not blaming someone else for his loss of the admin bit, blames me.
 * User:Caprice, who did not express support or opposition.
 * Enric Naval, an editor with no Wikiversity history until this, but major conflict history on Wikipedia, and interest in Cold fusion, my major academic project here. Enric Naval, however, was mostly concerned about pages at netknowledge.org, an alternative wiki where some study of Wikipedia editorial practice is under way. This was, in fact, irrelevant to custodianship here. I made sure to emphasize that Enric Naval was welcome to contribute here, and that any prior conflict was irrelevant (but because of that conflict, I'd not use tools with him absent emergency, and he is not a disruptive Wikipedia editor, has never been blocked there, AFAIK. There are other problems, he's been active in banning editors, but that's Wikipedia, not here.
 * TeleComNasSprVen asked a normal question.and implied another.
 * SB Johnny expressed concerns, boiling down to an opinion that I write too much and that I might misrepresent Wikiversity to newcomers.
 * To avoid further moot discussion, since Jtneill was absent, I closed the candidacy 17:44, 1 January 2011. Thus discussion was open for 6 days.
 * Jtneill returned to accept, 01:08, 10 January 2011], but reverted himself when he realized that the candidacy had been withdrawn.
 * Discussion ensued at User talk:Jtneill. Jtneill expressed an intention to only be involved as a mentor, not as a 'crat, and SBJ offered to do the 'crat work if Jtneill confirmed the mentorship. Ottava objected, and noted that, "A proposal has been put up to make it blatant that you aren't allowed to act this way. If you proceed anyway, then a proposal to terminate the mentorship all together will be necessary.
 * Jtneill re-opened the candidacy, 19:51, 11 January 2011.
 * Ottava objected. Note that Ottava complains about Jtneill as if Jtneill were acting as a 'crat. Deciding to go ahead with mentorship was a custodial decision, not a bureaucrat's decision. Contrary to Ottava's claim, Jtneill analyzed and considered the community's concerns, and addressed them.
 * Ottava then added a desperate attempt to wikilawyer this away: As per policy, candidates that aren't recommended by their mentor for full custodianship after 4 weeks do not get a second chance,] which is a blatant misrepresentation of the policy, which does not limit the number of candidacies and even allows immediate new mentorship.
 * SB Johnny granted probationary custodianship, noting it at 20:47, 11 January 2011.
 * The candidacy was open for discussion for six days before action was taken. The only serious opposition was from Ottava and JWS, both of which were totally predictable.
 * There was no community consensus in opposition to the candidacy. Indeed, looking at the comments, I only two clear oppositions, those of Ottava and JWS. JWS expressed his "dismay," not just at me, but with the entire custodial corps. Ottava wasn't so extreme, but see Community Review/Jtneill, Community Review/SB_Johnny, Community Review/Mikeu, that's our three most active bureaucrats. At meta, Ottava has attempted to gain intervention here, currently, trying to get SB Johnny desysopped for simply following standard procedure, then filing an RfC on SB Johnny there, suggesting stewards intervene to force a policy change here.
 * Clearly, the request here wasn't about fixing a broken process. It was worded, however, so that it would remove 'crat discretion, radically. It's been amended to tone it down a little, but still adds some Bad Ideas. Which I will discuss separately. The point here is that there was no abuse underlying this change. Nothing to fix.
 * Analyzing the comments to find clear votes, there were three against, including SBJ. There were three for, Jtneill, myself, and SBJ, who apparently changed his mind given Jtneill's response, or, alternatively, didn't believe that violating the policy was necessary. There was no consensus against the probationary custodianship.
 * Further, 'crats in closes have the discretion to consider arguments, and decision-making by the numbers is deprecated. Only when numerical consensus is clear should a 'crat refrain from deciding in contradiction to a supermajority. --Abd 21:32, 13 January 2011 (UTC)

Conclusion: community concerns were considered and addressed. Only two WV editors were actually opposed to the candidacy; and all expressed concerns received response. There was no community consensus against probationary custodianship, if all involved are considered. Mentorship was affirmed by Jtneill, a highly trusted user, one of our 'crats, and the determination of process propriety was by SB_Johnny. No abuse of process. Comment was actually open for 6 days. Both editors strongly opposed are opposed to the entire active bureaucrat corps, and one also is dismayed by the entire admin corps. --Abd 21:40, 13 January 2011 (UTC)
 * What are you talking about? Just because you "responded" to a concern doesn't mean it is addressed. There was no one who was voicing saying that there weren't concerns about you, and Jtneill even said that the community had concerns. You had all opposition and no support. Ottava Rima (talk) 01:38, 14 January 2011 (UTC)
 * Perhaps, Ottava, you are hoping people won't read the description of the process. I listed the people. There were two people opposed, that was you and JWS, whose opposition is Just Plain Weird. At least yours makes sense, you hate my guts, and it's obvious. Then there was SBJ, whom you are attacking for making the promotion decision. He was "opposed," when he commented. In support, contrary to your misinformation, would be myself and Jtneill, of course, but then SBJ actually decided that the concerns, including his own, were not enough to violate the policy, which seems to be what you'd have wanted him to do. Of course consensus can override policy. But that was not a consensus against probationary custodianship. This probationary custodianship would still go ahead, with those comments, over a 7-day period (one more day would have made no difference), under the policy you have just proposed. --Abd 04:15, 14 January 2011 (UTC)
 * "There were two people opposed, that was you and JWS, whose opposition is Just Plain Weird. " Belittling someone is not a substitute for an argument. Quite the contrary, it verifies that you lack an argument. I thank you for conceding the point. Ottava Rima (talk) 04:34, 14 January 2011 (UTC)
 * By the way, you assume that the other concerns were not opposition. Erkan's, for instance, raised serious issues. The proposal above is about concerns, not just blatant "oppose". It is to make sure admin think twice before ignoring concerns and giving justification to remove careless administrators who give ops to users not trusted by the community. Ottava Rima (talk) 04:38, 14 January 2011 (UTC)
 * Raising of issues is not "opposition." Users "not trusted by the community" is a huge category, including billions of people (as possible users); there is no offense involved in what amounts to "sharing" ops with people not yet trusted, and the legitimate concern here would be reasonable fear of abuse, and mentorship is designed to reduce any possible harm. Again, there is no example of actual harm having been done by a mentored custodianship, no example of anything that this change would have prevented, except maybe the very brief custodianship of Salmon of Doubt, and that case, in fact, shows how mentorship turned what could have been very contentious and disruptive into no problem at all, and the community gained a few intelligent custodial actions. I'll stand with the "Just Plain Weird" description of JWS's opposition, which is not belittling JWS, only pointing out the obvious about his arguments over the last two years. Uh, Erkan's what? --Abd 17:10, 14 January 2011 (UTC)
 * I meant Enric and not Erkan. And raising serious issues is considered opposition. You don't have to say "oppose" to have concerns. Crats are supposed to read all statements and see if there are any major concerns about something, and Jtneill admitted that there were. He even said he was concerned about the concerns. Ottava Rima (talk) 18:50, 14 January 2011 (UTC)
 * raising serious issues is considered opposition. Yes, to those who think in terms of battles and who see raising a concern as firing a salvo at the "other side."
 * Jtneill admitted that there were [concerns]. Like, he admitted the obvious. And he addressed those concerns, considering them not worthwhile withdrawing his mentorship offer, that he considered the probationary adminship safe. And SBJ, in spite of being personally opposed, in his initial opposition, appears to have considered Jtneill's response adequate. He was not obligated to approve, even if the policy appeared to require approval. That's part of how abusive policies are transcended: nobody enforces them.
 * Enric. Raising serious issues is not opposition, it is a request to consider the issues. Enric was, in any case, not at all an established user here and his showing up to comment could be considered an extension of Wikipedia conflict. I welcomed him anyway, and answered his questions. And he did not end up stating an actual opposition. Enric and I were sometimes able to cooperate on Wikipedia, I'm hoping we can do so here. People are complicated, but Ottava, it seems, tends to reduce everything to For Ottava's Position and Against Ottava's Position. That's part of the problem here. --Abd 19:42, 14 January 2011 (UTC)

Discussion
This would change the mentorship process from the current one which can essentially be a mentor-mentee arrangement between two people to one which would require community consensus in addition to mentor-mentee pairing. Note that "no community concerns" could mean that one small complaint could prevent probationary custodianship which I think is too harsh. -- Jtneill - Talk - c 20:06, 11 January 2011 (UTC)
 * The current process is a Crat giving someone free adminship against the wishes of the community, which is not acceptable practice at Wikiversity or the WMF. It is the definition of abuse and irresponsible. Ottava Rima (talk) 20:21, 11 January 2011 (UTC)
 * P.S., you had no problem desysopping a highly active admin over something that wasn't even a violation of policy, so your claims about "one small complaint" are disingenuous. Your actions suggest what can only be described as using ops to wage war upon others. Ottava Rima (talk) 20:23, 11 January 2011 (UTC)
 * This transparently a manifestation of what led to that "desysopping," which was based on a community decision, not just the opinion of one 'crat. Ottava is attempting to rewrite policy to suit a particular case where he has a huge personal axe to grind. No example of a problem from granting mentored adminship under current policy has been shown, at all. It's been claimed that there might have been a problem, but the mentor made sure there wasn't one. --Abd 20:31, 11 January 2011 (UTC)
 * Abd, the community doesn't want you here. You only exist here because of abuse by a Bureaucrat. Stop with the nonsense. Ottava Rima (talk) 22:46, 11 January 2011 (UTC)
 * Eh? I only exist because of abuse? What abuse? The day I was born? Ottava, as Lar said on meta, get a grip. --Abd 03:09, 12 January 2011 (UTC)


 * Comment. Setting up an RfC without prior discussion is disruptive. It wastes the community's time by soliciting support/oppose positions on issues before they are made clear through discussion. Wikiversity has no RfC process, as such; RfCs on Wikipedia are centrally listed and categorized, to solicit broader participation. Please don't list this RfC, as it is, in the site announcement! Anything listed in the site announcement should be ready for broad community discussion, and should clearly need that. Probationary candidacy, as designed, doesn't need that broad discussion, but permanent candidacy does, which is why such process gets listed. Our process is efficient and respectful, and hasn't broken yet. This would make it worse, not better.
 * Comment. I can imagine possible abuses of the mentorship system, and we could discuss how to tighten it up to head them off, but that's a separate issue. The community discussion proposed is a truly Bad Idea, redundant to the final discussion, and not addressing possible real problems at all. (If a mentor wants to allow someone to be a probationary custodian, who would be abusive, then there is a problem with the mentor! And that should be addressed, not something that has worked very well.) --Abd 20:27, 11 January 2011 (UTC)
 * More making up things about policies which just aren't true. Ottava Rima (talk) 22:46, 11 January 2011 (UTC)

I think "Wikiversity participants have one week to question and comment on the candidacy. If after one week, there is no community consensus against allowing the candidate to become a probationary custodian and the candidate can find a mentor, the mentorship period will begin" would be better. That would allow the current default of granting the tools if a mentor can be found to continue without interference, while giving the community a way to block if there are actual concerns. -- dark lama  21:00, 11 January 2011 (UTC)
 * Changed, while keeping the last sentence that is already part of the policy (at the bottom, though). Ottava Rima (talk) 22:49, 11 January 2011 (UTC)
 * The last sentence should read "Probationary Custodians need to have the community's trust. Mentorship is not a right." -- dark lama  23:06, 11 January 2011 (UTC)
 * Done. Ottava Rima (talk) 23:17, 11 January 2011 (UTC)

Closing this: policy changes need to be "advertised" to the community for causal input and collaborative chat before it gets to the "support/oppose" phase. --SB_Johnny talk 21:17, 11 January 2011 (UTC)
 * This post was moved. 1. SB Johnny has no authority to close polls before a set period of time. 2. There is no rule saying any amount of discussion is needed before a poll on changing a policy. 3. SB Johnny is clearly involved in the matter and has no right to make such actions. This is a 2 year long problem within SB Johnny's actions. Ottava Rima (talk) 22:50, 11 January 2011 (UTC)
 * Any user may do what SBJ did, and if it's wrong, any other user, not already involved, may reverse it. I'm going to restore SBJ's closure, which does not close this discussion, and when there is a mature discussion here, if there is still a need for an RfC, it can be opened. Consider this: the proposed change has already been edited. So if someone has voted, which version did they approve or oppose? Changing the question during a poll is very much a bad idea; that's why questions should be negotiated before a poll is taken. I'll note that present policy is clear on mentorship and procedure, and Ottava has been misrepresenting it, blatantly, and he just tried to get SBJ desysopped at meta simply for following present (long-standing) policy. That's disruptive, and it should stop. --Abd 03:16, 12 January 2011 (UTC)
 * No, they can't. You can't just archive discussions like that. Abd, you are officially and abusively edit warring. Anyone can open up a poll at any time. Consensus is the rule and does not allow you to randomly close polls. This has no consensus for it and has no standing. Ottava Rima (talk) 03:33, 12 January 2011 (UTC)
 * I can't? I'm officially edit warring and you weren't? How do you figure that? Generally, one revert does not revert warring make. I acted from my understanding of fair and efficient process, likeliest to find true community consensus, based on many decades of experience with consensus process. Don't like it, I have no authority over you, but, beware. The community does. I am, by no means, the only person with edit buttons here. --Abd 03:42, 12 January 2011 (UTC)
 * Warring is an action to push something in direct contradiction to closing. By reclosing something without having any process right to, not only are you directly going against Wikiversity:Consensus, you have no grounds to even rationalize it. In short, there is no way to claim what you are doing is acceptable. Anyone has the right to open and hold a discussion. Re-opening it is acceptable practice. Closing is not. Ottava Rima (talk) 03:58, 12 January 2011 (UTC)
 * It should also be noted that the RfC's above were not closed in such an inappropriate manner and had immediate voting along with the discussion. Ottava Rima (talk) 04:00, 12 January 2011 (UTC)
 * A single user acting restoring a former condition that the user created or previously restored, without discussion, insisting on the content, without compromise, is revert warring. Sometimes a single such action is allowed, it's called 1RR on Wikipedia, but it's iffy. Multiple users each making one restoration of content, that they did not themselves create, is not revert warring, though it is sometimes discouraged: there should be discussion and compromise if possible. However, when a position is only being asserted by one user, the existence of other users willing to act contrary to the action of that single user establishes a "rough consensus" -- two to one, in this case -- for the action. That's reversible at any time, and this is routine. Given that the core process here is discussion, particularly at this point, I see no reason why the Support/Oppose section must remain open for the time being. It's quite likely that if we decide to change policy, the language won't be the same as was present when those sections were opened up by Ottava. It's already changed once. I don't know if Ottava has ever considered this: if the proposal is changed, what is to be done with prior !votes?
 * This is why it's highly recommended to simply discuss changes first, before jumping to an RfC. That way, if an RfC is necessary, the question is more mature and more likely to be stable.
 * As to the RfCs above, there is really only one, presented by Ottava also. And it should have been shut down immediately, as too !voting, the proposal wasn't ripe, and, indeed, dispute over it is part of what led to Ottava's desysopping. Ironic, eh? So, because we failed to act quickly before, we should now not act in a better way?
 * Ottava has considered all this an emergency, demanding on meta that SBJ's custodian and 'crat bits be yanked. Denied, of course. Obvious. Then he created a meta RfC on SBJ, with a laundry list of offenses, all moot for meta, but requesting steward interference in Wikiversity self-governance. As part of that, he claimed that this discussion was "prevented." As anyone can see, that's false. If someone disagrees with what SBJ and I did, this is a wiki. I don't own it, nor does SBJ. Even Ottava doesn't own it. --Abd 22:39, 12 January 2011 (UTC)

'Comment' The proposal makes sense and I don't see why it wasn't originally adopted."Anonymous Uploader 23:35, 12 January 2011 (UTC)
 * See below, AU, it wasn't originally adopted because Wikiversity did not want to reproduce the train wreck at Wikipedia over administrative privileges, and it's stayed the way it is because it works and there is no example of failure, and it allows relatively unknown users to show the community what they can do. Please see the question below. Are you a newcomer to Wikiversity? I hope you will look around and realize that we do things very, very differently from the other WMF wikis, we are a kind of refuge, in a way. Users who have been very controversial elsewhere can become custodians here, Ottava was in that class. Indeed, so am I, though I'm not blocked anywhere (Ottava is), but I'm not a permanent custodian here. In order to become one, I'll need consensus in a full custodianship discussion. So why have two discussions? Probationary custodianship depends on the reliability of what is ordinarily two trusted users: a custodian mentor agreeing to supervise and an implementing 'crat. There is some possible issue if a 'crat agrees to mentor and pushes the button, but that has not resulted in actual problems so far, AFAIK. And it wasn't the case in the present flap. I.e., one 'crat offered to mentor, before I opened the candidacy page, and another 'crat acted on it. Rigorously proper. --Abd 00:06, 13 January 2011 (UTC)
 * It wasn't the community's original belief that Mentorship would be deemed a right nor given against community concerns. Ottava Rima (talk) 00:09, 13 January 2011 (UTC)


 * Nobody has claimed that mentorship is a right. However, wiki communities almost always have "concerns" about specific administrators or candidates. A custodian who agreed to mentor an abusive user, placing the community at risk, shouldn't be trusted himself or herself! Likewise a 'crat who ignores cogent concerns, and who awards a bit purely on the basis of "procedure." All of our administrative and bureaucratic procedures require discretion; it's obvious that in the cases Ottava has in mind, the 'crats considered there to be no substantial risk to the community. Our existing process allows the expression of concerns, but it also understands that initial, rapid response, to a candidacy, may well be warped toward the negative, with everyone who has ever been involved in a dispute with the candidate coming out of the woodwork. Concerns are addressed as part of the process, and candidate will make promises, mentors will set conditions, etc. It is not broken, and no example of damage that would be avoided by the change has been asserted.


 * I was a probationary custodian before, and there is a review of my activity linked from the current candidacy page. That review would makes completely obvious why Ottava has been so strongly opposed. As well, one can see from it that there is no basis for claims that I'd use the tools abusively.


 * Probationary custodianship is supervised. That's being missed. Some commenters may not be familiar with how Wikiversity works. One who fears that I will abuse the tools may not realize that I'm closely watched. This is not Wikipedia. --Abd 21:51, 13 January 2011 (UTC)
 * "wiki communities almost always have "concerns"" Looking at past discussions I see very few candidates who had community concerns voiced against them. Ottava Rima (talk) 22:11, 13 January 2011 (UTC)

Possible canvassing
New users at Wikiversity are welcome to express opinions about proposals, but with what amounts to a drastic policy change, creating more opportunity for divisive discussion, which this proposal would implement, we should be careful about reading community consensus from a relatively obscure discussion with possibly biased participation. For best process, When there is apparent consensus here that a proposal is mature, then a full RfC should be announced in the site message. Meanwhile, I note several things:
 * Back-and-forth reversion of a temporary closure of the Support/Oppose section of this poll, with attempts at meta to bypass Wikiversity process and force this change as supposedly required by Foundation policy. This is a direct attack on an important part of the Wikiversity tradition, how serving as a custodian here is more widely open, even to users banned at other sites, yet the community is still protected; problems with custodianship at Wikiversity have not come from probationary custodians, but from full ones. The temporary closure of !voting as premature was not as represented at meta, as an attempt to prevent discussion. And, obviously, no admin tools were used, merely editors expressed their positions through editing. I don't see that consensus was sought, however.
 * The appearance, then, of two new users !voting here, showing no awareness of Wikiversity tradition and how mentorship currently works, who have no significant history on Wikiversity. These would be Kevin Rutherford and Guido den Broeder. Note that Kevin's only edits to Wikiversity have been this and to oppose the desysopping of Ottava.
 * Indications are that Ottava considers this a battle, and expects that Wikiversity users will line up on his side, see . Ottava has been waging a campaign against the entire administrative core of Wikiversity, see Community Review/Jtneill, Censure_SB_Johnny_for_Community_Review, attempt to remove SB_Johnny's privileges, and Community Review/Mikeu. Ottava's meta RfC on "SB Johnny" is an attempt to bypass local discussion of this policy change. Ottava has also attempted to bring Jimbo into our issues: and . It is hard to imagine a stronger attempt to disrupt Wikiversity. --Abd 15:35, 14 January 2011 (UTC)


 * I find this interesting. Abd has canvassed using WR before. Many people who voted in the above RfCs were admitted to be canvassed by SB Johnny. It is a little odd, especially when you started a thread that led up to this matter on Wikipedia Review, then you are worried that people came here after that. Also, entire administrator core? I find it odd how you can say SB Johnny, Mikeu and Jtneill represent all administrators. Darklama, Geoff Plourde, and many others would find that a little troubling. Did you instead mean to say the only three people willing to back you up? The only one disrupting Wikiversity is you and SB Johnny. We were fine until you arrived and SB Johnny decided to abuse his ops in every possible way. Ottava Rima (talk) 18:54, 14 January 2011 (UTC)
 * And I thank you for linking to Wikipedia Review so everyone can see how you are not acting in good faith here. You make incivil and nasty attacks, using adminship as taunting, demanding it as your rights, and all sorts of other inappropriate things. Wikiversity is not a game and your WR posts show a mentality that is inappropriate for any WMF project. Ottava Rima (talk) 18:56, 14 January 2011 (UTC)
 * Is WR a neutral place? It could be claimed to be so. I have no problem with my activities being visible, I like them that way. I raised the canvassing issue for consideration. It's not an attack. The point is that this proposal was obviously made as part of undisclosed axe-grinding, and therefore the possible effects should be very carefully considered. Darklama has already modified the proposal once, reducing the potential for drama from Ottava's original. I'd simply say we are not there yet, as to real consensus. Ottava's posts on WR and on meta betray the agenda. This is not about fixing what is broken, it's about abusing our process, pursuing "payback." The evidence I cited is conclusive, but, of course, as always, it's up to the community as to what it wants to do. Parts of the proposal may be worth considering. Or if we identify the real problem, then we can design better solutions. --Abd 19:29, 14 January 2011 (UTC)
 * At one place, Ottava opined that the entire administrative corps would need to be changed, I can't recall where that was. I just mentioned the three active bureaucrats as the most notable sample. It's enough to review those RfCs, the arguments given, etc., not only the fact that the community rejected Ottava's positions. --Abd 19:33, 14 January 2011 (UTC)
 * "This is not about fixing what is broken, it's about abusing our process, " Actually, that is only applicable to Abd 2 and why you were given ops. You and SB Johnny consistently abused our processes. My only agenda is to protect the community from your rampages. Moulton even states over and over that your statements are all nonsense. That isn't me saying it. How do you not get that? And those three "active" Bureaucrats are only one truly active member, Jtneill, and even he isn't that active. Ottava Rima (talk) 19:43, 14 January 2011 (UTC)
 * As for this: "that the entire administrative corps would need to be changed". Yes, when 50% of our admin are, on average, inactive for 2 years and only 12 have been "active" within the past 3 months, that is a serious problem. When one decides to use his power to intimidate others who want to fix that, that is a serious problem. Ottava Rima (talk) 19:45, 14 January 2011 (UTC)
 * If anything, Abd, your consistent assumptions of bad faith make it clear to me at least that you are not custodian material. You are not only attacking Ottava Rima with your insinuations of canvassing, but Kevin and me as well, and, by implication, also Lar. You would do well to offer your apologies. Guido den Broeder 19:50, 14 January 2011 (UTC)
 * Agreed on the need for a much better tone here. OTOH, WV:CANVASS might look a little different from what you'd expect ;-). --SB_Johnny talk 20:42, 14 January 2011 (UTC)
 * The Wikipedia page on Canvassing looked something like that as well, originally. It took just one user to change it overnight and give the concept the entirely negative meaning that it has on WP today. :) Guido den Broeder 20:54, 14 January 2011 (UTC)

What's broken? Anything?
The RfC above seems to assume that something is broken about Wikiversity probationary custodianship process. While I do see a possible "loophole," that could be abused, perhaps, I know of no example of actual problems resulting from it. We could, indeed, profitably consider this and plug the loophole, and the present flap is over a candidacy where the mentor specifically avoided the problem. Jtneill said that he preferred to not mentor and make a 'crat decision with the same candidate. So he didn't! Because he did not, it took two trusted users to start up a probationary period, as is usual: the mentor and a 'crat. Crats are not rubber stamps, and if a crat saw a problem with a candidacy, sufficient to not allow a probationary period, I'd expect the 'crat to prevent it, or, at least, to discuss it. In any case, the first order of business is to determine what is actually broken. Allowing a single 'crat to mentor and begin the probationary period is a theoretical loophole, but is there any example where this actually caused harm? If so, please describe the situation so we can review it.

Present process does not provide for a period of community comment until later, with permanent custodianship. It seems that the process was designed specifically to avoid that extra red tape. Many of our users, and even some custodians, have come here from other WMF wikis where there were problems. Our process here is vulnerable to users with whom they had conflict, on other wikis, dropping in to voice their "concerns," and we probably lost Diego Grez as a custodian because of canvassed !votes like that. The policy provides for renewed mentorship if a candidacy fails, so that problem could have been bypassed if Diego had asked for a new mentorship; I suggested it, in fact, as I recall, but he was discouraged, I think.

Is there any example where a failure to have a period of community comment before beginning probationary custodianship caused an actual problem? What are the real situations we might be looking at? --Abd 22:51, 12 January 2011 (UTC)
 * My understanding of Wikiversity is that here we have a unique international forum for the exposition of new ideas [brand new] that will progress to a solid understanding of matters that were hitherto considered to be " mysteries " whereas in fact they were always and are now just problems of logic. This I believed and continue to believe is the rationale for the existence of Wikiversity. Incidentally some one has written: " A place of bad memories. " This comment requires an explanation. User: SHAWWPG19410425 13th March 2011.SHAWWPG19410425 11:18, 13 March 2011 (UTC)
 * While it's true that political problems can be modeled as errors in logic, those errors are frequently subtle enough that it takes a genius of the order of Socrates, Hillel, Buddha, Jesus, Becket, Augustine, Newton, Jefferson, Dostoevsky, Poincare, Gandhi, Lorenz, King, or Girard to analyze them.  And even so, most people won't have the mathematical depth to appreciate subtle logic errors.  —Barsoom Tork 11:33, 13 March 2011 (UTC)

Convert "policy" to "proposed policy"?
SB_Johnny changed the description of the attached page from "policy" to "proposed policy." This has stood as policy since it was converted from "proposed policy" to "policy" by JWSchmidt, 12 February 2007. It was actual practice for that time and longer. I have reverted the change. Policy should not be changed without discussion and consensus. --Abd 01:37, 8 June 2011 (UTC)


 * My understanding, from discussing things with Darklama a while back, was that at some point JWS declared many things policy without seeking community consensus. Several of those thing have in recent years have been changed back to proposed policies so they could be discussed and voted on by the community.  I am personally in favor of changing to "proposed policy" if there was never a formal adoption procedure for it. Thenub314 02:58, 8 June 2011 (UTC)


 * It's been followed for years, this wasn't merely a declaration by JWS! At the time, JWS was extremely popular, he was a shoo-in for 'crat, but turned down the nomination. There is no requirement for any formal adoption procedure, where something was so broadly accepted, obviously. Now, if there is evidence that significant elements in the community actually opposed what JWS did, that could be another story. I'm not seeing that at all, in this case.
 * It's not clear what the purpose of the change is. We do have a procedure in place and it's been followed for years. The procedure was followed, I think, before the page said "policy," but it said "proposed policy." Is this an attempt to shut down the procedure, but indirectly? Why? --Abd 03:05, 8 June 2011 (UTC)
 * I do not think there is any attempt to shut down the procedure. But (I am guessing) SBJ hoped to highlight that there has never been a community wide discussion whether or not the pattern we follow is policy.  This makes sense as he is trying to review the way we have been doing things.  To be honest people have a tendency to follow things marked as policy even when they don't make sense.  It is good to be clear about when/if pages have been carefully vetted by the community at large, and not just the small subset that does custodial work. Was the page brought up in a site notice, or in the Colloquium before it was declared to be policy?  I think these are rather necessary steps to take.


 * I like the mentorship approach, but I really think it needs to be clarified. This wouldn't be an issue if Abd hadn't suggested that he would simply "go to the stewards" rather than rely on local process. My position (and I believe Mike's as well) is that I (as a 'crat) really need clarity from the community on how people want the buttons used. Without that clarity I'm hesitant to use the buttons, particularly in a controversial case, which this case certainly is.
 * Due to the explicit threat to try to use this as-yet-unapproved policy to over-ride local decision making, I think it would be wise to keep it marked as proposed until the CR discussion produces some result. --SB_Johnny talk 12:07, 8 June 2011 (UTC)
 * It's been standard practice here, as it is elsewhere, when a situation requiring custodian or bureaucrat action does not receive response in a reasonable period of time, to request action by a steward. I did it before, nobody objected at all. There is no "threat," there is only a suggestion of going to stewards as a possibility, if no 'crat is available to act and chooses to act. There is reliance on local process *first.* How long should we wait? That's all. --Abd 17:33, 8 June 2011 (UTC)

For me, policy or proposed policy is a bit moot. Let's improve the policy via the talk page here and in Community Review/Custodianship process. -- Jtneill - Talk - c 23:35, 16 June 2011 (UTC)


 * The reduction of the page from policy to proposal did not stand. The page is presently policy, and should not be changed without consensus. I second Jtneill's suggestion. However, I don't think the policy is broken, and I've suggested a change, self-reverted with, that could address some of the difficulties experienced last year and earlier this year. I will discuss this in a section below. --Abd

Split the tools and allow users to be mentored towards each tool or apply directly
Up until now, the "custodial" tools are:
 * 1) Rollback
 * 2) Protect
 * 3) Import
 * 4) Delete
 * 5) Block
 * 6) Editing MediaWiki namespace

and the current "bureaucrat" tools are:
 * 1) User rights (add not remove)

-- Jtneill - Talk - c 23:35, 16 June 2011 (UTC)
 * Proposed changes to admin rights on WV
 * 1) Instead of rolling (too) many rights into the custodian role, this proposal is to split the rights so that users can gain access to each right separately.
 * 2) Access to rights is available via mentorship and probationary use and followed by community approval, or directly via community approval.
 * 3) Rights removal is via emergency followed by community consensus or community consensus.
 * 4) Rights are added or removed by those with user right access.

Clarifying possible removal process for probationary custodians
Self-reverted proposed change to policy.

Because certain unwritten possibilities were not clear, my previous probationary custodianship periods were disruptively terminated. I will describe these events from my point of view, and this could make plain how to improve the situation.


 * Period with Ottava Rima as mentor. This period was largely uneventful until I came across Ottava insisting upon incivility, as was confirmed by later review. I warned him and when he rejected the warning, I short-blocked him. His response confirmed what some had suspected, that he was not suited for custodianship; however, it took some time to work that out. As to my custodianship, it was terminated when Ottava went to meta and gave an argument that it had "expired," in which he ignored actual policy on withdrawal of mentorship. Stewards will not investigate deeply, and they will assume that a request from an administrator is made in good faith, because, in the event they err, any local bureaucrat can quickly undo the damage. Hence I did not challenge this at meta, nor did I request restoration of the bit. However, a permanent custodianship poll was set up, I did not support that, preferring to wait until the smoke cleared. However, I also felt that I should not prevent the community from discussing the issue, so I did not withdraw, either. As was not surprising, the poll failed to approve, with "blocked mentor" being a major argument. Point: no disruptive process was required to terminate custodianship. Mentor could have properly terminated custodianship simply by withdrawal, and if no new mentor were found within the specified time, could then have had the bit lifted. Emergency desysop is always available at the discretion of any 'crat, as well. These requests are generally honored by stewards, again, because of the easy reversal of an error.
 * Period with Jtneill as a mentor. SB_Johnny had implemented, in spite of clearly having personal reservations. Conflict arose with Ottava Rima, who started a series of community discussions, against SB_Johnny, against me, against Jtneill, against Mu301, i.e., three active bureaucrats. I closed the Mu301 review and protected it as closed, giving my reason on the Talk page. The protection stood. I also protected the Jtneill CR, protection was removed by Jtneill. Removal of protection by any custodian would have been fine with me, but Jtneill especially. I did not protect the SBJ review, which had been started in a different place. But it was a mess. It was eventually protected by Mu301. I had closed it as No Consensus, which is clear.
 * However, SBJ closed Community_Review/Abd as showing consensus for desysop. That review had been created without any evidence the policy condition allowing such reviews: "Custodians can lose their status for egregious violations of policies." Rather, it was based on this version of a Colloquium proposal for a topic ban. Ottava copied "most" comments from that discussion to the Community Review. It appears that he had canvassed for comments, probably on IRC, see the full Colloquium discussion. I criticize SBJ's closure in his Community Review, see . Be that as it may, if there had indeed been an emergency as claimed, SBJ had the authority, explicitly confirmed by me, to request desysop at any time. I still do not know why he did not choose to exercise that authority. I have one speculation: he's opposed to my being a sysop, forever. By establishing this as a "community consensus," he has cover. This speculation, however, is in no way a suggestion that disruptive process should begin to remove privileges. Quite simply, SBJ does not do enough damage, there is no "egregious violation of policy," there are merely some events that have caused me to lose faith in his neutrality and wisdom.

Wikiversity has procedures that make it difficult for an individual administrator to effectively "ban" a user from anything. Probationary custodianship is one such. It allows a kind of "amplification" of custodian effectiveness. Mentored custodianship is safe as long as the mentor is reasonably responsive to communication. It must also be said that if a mentor, after opportunity to review the action of a mentored custodian, approves of the action, and permits continued similar action, the mentor must be considered responsible for it, as if the mentor had done it himself or herself.

In my proposed clarification, which I will restore if there is no objection within a reasonable period of time, the implementing bureaucrat also has the right to undo what he or she has done. That is normally a basic wiki principle, any editor may undo their own action. Normally, if an administrator has blocked, they may unblock even if they are later accused of recusal failure, it is the same with the reverse, if an administrator has unblocked, they may restore the block (increasing the block, though, would be offensive, if recusal failure is reasonably claimed). There are reasons why bureacrats cannot unilaterally desysop, it is a safeguard. However, that safeguard prevents a 'crat from undoing their own action. So that it is clear at meta, I've added the language to the policy that makes it explicit that the 'crat who implemented may request removal, without fuss.

It is already established practice that bureaucrats may, in an emergency, request desysop at meta, even without policy. The stewards will, generally, do this, because any bureaucrat may, locally, undo the effect, so if it isn't undone, there is obviously an effective consensus of all active bureaucrats.

I also suggest, in the proposed change, that nondisruptive process be followed first, before the more disruptive process of a Community Review. Community Reviews on individuals will attract every user who has ever had a conflict with the subject. We commonly see, in such reviews and similar processes, people show up who have little or no regular connection with Wikiversity. For a Community Review to attract broad participation, it should probably be in the site notice, and it should remain open for a substantial time. The 7-day process established in policy is really for emergency desysop, because of the requirement for "egregious violations of policy."

I also refer to the possibility of individual custodian agreements regarding conditions for desysop. Had those been used, Community Review/Abd might not have been needed, if there was actual abuse, or it could have been quickly closed. In my current candidacy, I refer to a standard stop agreement. That provides a process whereby any custodian may effectively (and flexibly) prohibit further use of tools by a probationary custodian who has agreed to it. The probationary custodianship of anyone who has agreed to this should be quite safe. Rather than discouraging probationary custodianship, which is a position being advocated, we should be encouraging it. Wikiversity needs more custodians. I commonly see situations where a custodial action is required. Frequently, if it's important enough, I place a speedy deletion tag, or request an action on the page for that; however, response is often quite slow. Regardless, I've been acting in quasi-custodial mode, for a long time. I do things that are often only done by custodians, such as close RfDs. At first there was some opposition to this, but my actual closes have been universally sustained, and the opposition has disappeared. (I was surprised to see this, in fact, I thought I might at least make occasional mistakes, but I am careful and consider it my duty, in actions like that, to simply respect and reflect community consensus. I'm cleaning up request pages because they then work better, not to pursue any personal agenda.)

In the past, mentors have sometimes required special agreements. That is totally within the discretion of the mentor and mentee, but such agreements can and will be enforced, the example of SBJ and Salmon of Doubt stands out as a way that SBJ protected the community against the possibility of harm by this user, whom some suspected of being here for disruption. SBJ managed to protect the community while possibly gaining skilled sysop work. No damage was done. SBJ was able to "take back the keys" solely as his own decision, and immediately.

This is not an argument for my custodianship, that will be worked out elsewhere. However, my point is that the disruption that occurred around it was completely avoidable. The real source of that disruption was not me, and I was attacked (as were others, including three 'crats) for serving the community, in ways that the ultimate flow of events confirmed. It is well-known on Wikipedia that any administrator who actually serves the community will make enemies, that is one argument that is given for making adminship there permanent, requiring a formal Arbitration committee majority vote to desysop. I do not, however, consider any user to be my enemy. That's up to them, and the past is the past.

We need to keep it simple here. We need mentored custodianship because it allows that simplicity. There are multiple safeguards in place, made clear by my proposed change. --Abd 18:14, 31 July 2011 (UTC)
 * Undid self-reversion, implementing change, no objection appeared here. --Abd 16:28, 12 September 2011 (UTC)
 * Lack of objection does not mean support. You need support before you can institutionalize change, not hide something in obscurity, wait a few months, and declare it true via fiat. Ottava Rima (talk) 16:50, 12 September 2011 (UTC)


 * Darklama edited my change, accepting the basic idea. Generally, that's acceptable to me, except for one thing. A mentor must allow a probationary custodian 48 hours to find a new mentor, if withdrawing support, before requesting removal, that's in the policy. So I edited this to cover that complication. What 'crats can do is more complicated. 'Crats may request desysop on an emergency basis, and have done so, sometimes. Aside from that, the probationary custodian's candidacy may include agreements that provide for contingencies about fast removal, see WV:Candidates for Custodianship/Abd 3 for an example. In general, my opinion is that what a 'crat can do, the 'crat can undo, so if the 'crat implemented the custodianship, the 'crat may undo it at meta, without argument. But if the custodianship was implemented by another 'crat, not necessarily. --Abd 20:51, 12 September 2011 (UTC)

Comment period for full custodianship
I changed the evaluation period to seven full days from five, and self-reverted. The templated evaluation page also contains a reference to five days, and will need change if this is accepted.

The reason is that a full week allows users who might be checking in once a week to notice the process. Because, normally, a candidate has the tools, there is no rush on completing the discussion, and nothing happens when it is complete unless it results in denial. Little-noticed: the probationary period can be restarted, all it takes is mentor willingness, by policy. If a mentor has expressed willingness to continue supervision, a 'crat should not routinely go to meta to request desysop. However, if the community is clearly opposed to continuation, mentor and 'crat should consider that. --Abd 16:38, 12 September 2011 (UTC)
 * There was no response to this proposal. I have an open voting period at the moment, and I think it might be useful to extend that, but I didn't make this proposal for myself. If no objection appears in the next week, I'll undo the effect of my last revert, or anyone else may do that, as anyone could have done since the proposal in September. --Abd 21:44, 21 November 2011 (UTC)
 * Thanks, Darklama. This implements the period as one week, and makes it clear that the period may exceed a week, a week is minimum. --Abd 21:25, 22 November 2011 (UTC)

I agree that at least a week instead of 5 days seems reasonable. -- Jtneill - Talk - c 23:40, 24 November 2011 (UTC)

email for custodians?
The policy currently reads: ''* Custodians should set their "user preferences" so as to provide for email contacts from other Wikiversity participants. If you do not use email, then you must make yourself easily available by some other means such as IRC chat.''

What the first sentence gives, the second takes away. We do have an active custodian who does not have email set. I've gone onto IRC quite a number of times and did not see that custodian as active. IRC is also not a place to make a private request, and while, if the custodian is on-line, one requests private chat, that's public. Email is private, almost fully (email is logged, but is apparently heavily anonymized. Certainly ordinary users and custodians have no access to the logs.) Further, IRC communications are not necessarily validated identity. Email must be validated for the user, and it's easy to confirm that an email actually came from a particular user. Given the availability of free, private email, easily anonymized, I see no justification for a custodian to not follow the basic policy, and IRC rules out many users who won't set up IRC just to contact a custodian! "Some other means," if allowed, must be *more* convenient than email, not less. Further, if the custodian does not have email enabled, the custodian cannot then use email to contact a user, which limits possibilities for developing cooperation with users. The second sentence should just be eliminated. --Abd 17:19, 24 November 2011 (UTC)


 * I think public discussion about requests that may need custodian action should be recommended and private discussions discouraged. The availability of free, private, and easily anonymized email also means emails from a requester are not necessarily indication of a validated identity. Wiki discussions are more convenient than email because other people can participate in a discussion and read what lead to choices made that private discussions like email don't allow. I think custodians should, like any user, be able to decide to where and by what means they are willing to discuss for discussions that do not require custodian action. -- dark  lama  19:11, 24 November 2011 (UTC)
 * Blocked users may request review by email, that's standard. Some discussions, by their nature, should begin, at least, privately. What is correct about Darklama's position is that no decision should be made as to custodial action based solely on private email. The problem Darklama is addressing also applies to IRC, which is a highly warped set of participants, consensus is not to be formed there, it's almost as bad as email (for that purpose). Email is for gaining attention and gaining and giving advice, not for wiki decision-making.
 * Darklama is incorrect, though, about validated identity, i.e., the possibility of validating that a mail is from a specific Wikiversity user. In order to send a WV email, the user must have email enabled, and that email has been validated as belonging to the user by the interface. This is not to be confused with validating real life identity, that's not really possible. If a custodian receives a direct email, it would be because the custodian has revealed that email address somewhere, and this email would not show as coming from Wikiversity. I'd recommend not answering such mails, unless you independently know the sender and their address. You'd be revealing your email address (and IP, by the way). Rather, respond to a participant through the WV interface, until you know for sure that an email address is that of a specific participant. --Abd 23:14, 24 November 2011 (UTC)


 * I think private discussion is absolutely necessary, especially with certain people claiming any questioning of others as "incivil". It basically makes it impossible to have any honest discussion over disruptive behavior. Admin need to be contacted, and "public" discussion is completely inappropriate in most cases, especially when there is personal information involved, gray areas, etc. It absolutely needs to stay if this is ever made an official policy. Ottava Rima (talk) 23:34, 24 November 2011 (UTC)


 * The page is policy, it's long been so labelled. However, the policy required email until this was changed by JWS in December 2007 with no discussion that I've found. JWS had originally written that section. Oddly, JWS had commented on email in Candidates for Custodianship/Darklama, recommending email to Darklama. The matter did not come up in the Permanent Custodianship voting, and Darklama became permanent at the beginning of October, 2007. Let me guess: it was discussed by IRC. Otherwise why would JWS, out of the blue, change an obviously sensible policy that he'd supported? (And he'd also brought this up with Custodian Remi, earlier in 2007. Remi apparently complied?) Maybe I missed a discussion somewhere. --Abd 00:31, 25 November 2011 (UTC)


 * I recall IRC is where initially I was introduced to what Wikiversity is about, and where a few people persuaded me to reluctantly accept custodianship. I may have mentioned the email requirement as one of many reason for me to not accept custodianship. Even if email from users is enabled, people can create throw away accounts they never check or read, which makes them as unreachable by email as any other means if they don't use it. I think privacy concerns should probably be addressed by checkusers or oversighters since they are trusted with access to sensitive information already. I think the fewer the barriers there are to custodianship the more people there are likely to be who are willing to volunteer to address the community's needs. I think custodians tend to focus on some small part of Wikiversity, just like most participants do.
 * Some people may think there are appropriate and inappropriate ways to honestly discuss behavior concerns. I think behavior concerns should be addressed publicly because concerns privately addressed can lead to new inappropriate behaviors that seem to appear out of the blue for no obvious reason. I think problems that arise out of the blue because of private discussion can be near impossible to resolve. I think Wikiversity has already unfortunately seen some problems arise from discussions having been held in private. -- dark lama  01:28, 25 November 2011 (UTC)


 * While I tend to use email and IRC, I would suggest that it shouldn't be a requirement to Custodianship. As with any user, custodians should be able to set their email preferences freely. Also, it is a misconception to think that email is more secure or somehow more verified then IRC.  I personally have faked mail headers to make emails appear as if they are coming from someone else, it is extraordinarily easy to do if you have the right bit of technical knowledge, and unless the custodian on the other end checks the headers carefully, it would easily be missed. Thenub314 04:22, 25 November 2011 (UTC)
 * First of all, faking email headers to fool a skilled user is about impossible. Spammers would love to do this, but a skilled user can see the insertion of the fake into the chain of senders. However, the point has been missed. If I receive an email from the WV system, I can see that it comes from that system, and if I have any reason at all to doubt this, I can check the history of mail transmission in the headers. However, that's not what I had in mind. If I respond to such an email through the WV interface, this then validates that I'm sending to the user in question. If the user says "Huh? I didn't send you the mail you've copied to me," I then know to check the original headers carefully. If I get an offensive mail appearing to be from a user, I'll check carefully anyway, if it is at all out of character. If however, this is a blocked user requesting unblock, which is a common situation where email requests would be important, I don't actually need to be so careful. I'd simply look at the situation on-wiki. I have, however, never personally encountered a faked email, appearing to be through the user interface, in years of WikiMedia experience. Doesn't mean it couldn't happen, and I'm still careful. (This mail would be illegal, at least in some jurisdictions, by the way.)


 * The question here should be our policy, and there was obviously long sentiment for requiring email for custodians. If we do have a policy, that doesn't mean that we couldn't make an exception, we'd just need to document that. But I'm not seeing any sensible ground for exception here, and discussion of this, for the custodian in question, seems to have been off-wiki. If it had been by email, then, at least, it would be simple to open that up, if the custodian cared to, and the involved users consent. But it was apparently by IRC. No record, unless someone happens to have logs from so long ago! IRC is very insecure compared to email.


 * Wikipedia seems to assume that admins will have email enabled. See [w:Wikipedia:List of administrators] and [w:Wikipedia:Requests_for_administrator_attention]. (On the latter page, emailing a blocking admin to request unblock or before unblocking is recommended.) One more point. If Darklama were to ever lose his password, he would be unable to recover, password recovery depends on email. How would we know that a new account, claiming to be him, was really him? --Abd 16:50, 25 November 2011 (UTC)


 * I think your comments about email security are simply not correct. But your correct that the issue here is about our policy, and so we should just let that aspect of the conversation drop.  It is my opinion that not having your email enabled is a choice that really has no bearing on the duties of being a custodian.  Your hypothetical about DL losing his password are silly.  He could just has easily have lost his email password and we would be back in the same situation.  The answer is simply that if someone comes by claiming to be some inactive custodian we have no verifiable method for establishing identity.   That situation is easy to handle, we simply have a new RfC with the new account.  It is more or less irrelevant if it is the same user and they lost their login information, a new mentorship period doesn't seem so extreme.  Overall I am in favor or removing this note entirely, or at worst leaving it as it is.  Custodians should be as free to choose their settings as other users.  If not where do we draw the line, does WV start requiring more strict requirements of passwords?  Language user boxes so people know which language communication can be emailed?  Time zone disclosures so users know which custodians snappy to respond to immediate issues?  It all sounds much too much to me.


 * The fact that we have custodians who are uncomfortable with disclosing their email address is exactly why there should be no requirement. It clearly sets a bar that makes otherwise excellent custodians uncomfortable with the position. Thenub314 20:18, 25 November 2011 (UTC)


 * Something is puzzling here.
 * Password loss. If a custodian doesn't have email enabled, losing one password demolishes the account, unless the user can establish identity some other way. If the custodian loses his email password, it has no effect here, he simply recovers that password independently, or doesn't, and if he doesn't, he may designate a new email account. What Thenub is doing is equating one hazard with a risk of 1/N, say, with the simultaneous occurrence of two risks, each perhaps 1/N. That's 1/(N^2), i.e., probably completely negligible. As a community, we have far more invested in a custodian, normally, than in an ordinary user.
 * Simply have an RfC with the new account. This is tantamount to a referendum on the custodian. It can be far from simple, it could possibly be quite disruptive.
 * Custodians should be as free as other users. Actually, no, not and retain custodial privileges, necessarily. Custodians are often judged based on issues not about their specific performance with tools. In addition, custodians have responsibilities, and not having email enabled can damage performance on those responsibilities. I've found it frustrating, myself, that Darklama doesn't have email enabled. IRC isn't practical for me; I did use it for a time, but it is way too demanding, it requires continuous and immediate attention, practically speaking. If you really want to stick with your position, Thenub, we'll take this to the community for consideration, unless you can show why using email would be an onerous burden on a custodian. As I wrote, we could have the policy and still make exceptions for some necessity. The question here is not really about Darklama, he's just an example, though the only one I know of. Are there any others? You used the plural:
 * Custodians who are uncomfortable with disclosing their email address. That's a red herring. Having email enabled does not disclose your email address, it would be difficult even for a developer to hack it out. Responding to a mail will disclose your address to your correspondent, obviously, but your address could be an anonymous account. The point is to have a means of contact, not to expose the RL identity of the custodian. Some admins on Wikipedia have accounts at gmail, or the like, that are only used for admin purposes. If the account is compromised, i.e., someone reveals it publicly, and they start to get too much harassing mail, they just drop it and start up and enable a new one. No damage done, a few minutes for the one person to fix. There is no problem here, Thenub, as far as I can see. Darklama did not publicly answer the request before, in his original custodianship candidacy.
 * Darklama's position above is based on a view that discussions should be public, but I know that Darklama uses private chat in IRC, and I think almost all of us agree that these can be useful under some circumstances. So ... what's the problem here? --Abd 20:45, 25 November 2011 (UTC)
 * Erm, about your first bullet, why are you assuming independence? I don't personally believe your mathematical argument makes any sense.


 * There is not ever any need to look at the record the previous custodian. Loosing your access to your account is just a CLEANSTART.  Yes it might be annoying for this to be the case, but it is the custodians/users choice.  Also, the user may have an email address without enabling emails from other users.  It is not all about losing passwords.


 * About your third bullet. You left out 'to set their preferences.  It is not up to us to dictate how individual users guard their personal security.  Lots of people anonymize all their internet traffic, so the sites they visit have no identifiable information.  There are companies offering this service, firefox extensions to make it easy, etc, etc.  This is a reasonable way to conduct oneself.  Revealing your email address to WMF might not be acceptable to all users.  Having an email that Google/Yahoo/etc can monitor and collect data from and advertise to may also be unacceptable to some users.  That is their choice, I would choose to allow as much freedom as possible for any user.  Your assuming that no one should object to multiple gmail accounts as necessary because they are free in the monitary sense.  But lots of people object to the invasion of privacy from projects like google start projects like this.  I do not think we should tell people they must get a google account because we see "no harm" in it.  It is up to each individual user to guard their privacy from whomever they see necessary.  I do not see this as any reason one shouldn't be a custodian.    Besides, I know a little about how mediawiki works, and if I am not terribly mistaken it is very easy for anyone with commandline access to dump an email address.  You only need a command like: "SELECT user_email FROM USER WHERE  user_name = 'MyUserName'"


 * About the forth bullet, there is a big difference between a user choosing to do something, such as user IRC chat or email, and being required to do something. This is a volunteer effort, even at the custodial level.  Theyonly requirement in my eyes is "don't abuse the tools".  DL is not really the only case.  I occasionally go a few months between checking the email associated with this account.  And the account is about to be deleted.  In all honesty, as I am not required to read or react to emails (and I often don't) what is the utility in making a it fictitiously appear that our admins are available via email? Thenub314 21:57, 25 November 2011 (UTC)
 * You've made some valid points above, but ... it also seems you are strongly disagreeing with the intention of the policy. Are you proposing to make a change? If so, then seek consensus on your change. I will be proposing removing the optional status that was added, apparently to cover one custodian. And we'll see what the community thinks. --Abd 22:08, 25 November 2011 (UTC)
 * I don't consider it apparent what the reason/intention for adding that clause was. I would have strongly objected and possibly have reversed the edit myself, if anyone had clearly claimed their intentions and decision to make such a change was based or derived solely out of consideration for me. I consider myself, possibly with a strong negative bias towards myself, replaceable and of no significant value. In fact, I often hope some custodian can/will replace me someday soon. -- dark lama  23:15, 25 November 2011 (UTC)
 * You do know that the issue was raised in your candidacy by JWSchmidt, right? (As he'd raised it before for another.) Did you ever discuss it? However, he didn't vote in your permanent status discussion. We could ask him, but I'd rather not disturb him. It doesn't really matter that much. The issue is the policy, and, as I wrote, exceptions to policies can be made, generally for cause, or just because the community feels like it. --Abd 23:38, 25 November 2011 (UTC)
 * I remember worrying I would have less time to devote to learning if the nomination was successful, and not looking forward to reading my candidacy page. I might have avoided reading it at the time. I don't remember there being any issues raised for my candidacy except coming from me, trying to talk other people out of wanting me as a custodian. I don't remember email being discussed specifically, just acknowledge that I might have brought it up, could have been why JWSchmidt mentioned it in my nomination, and could have been why JWSchmidt changed policy, but I hope policy wasn't changed because of me. -- dark lama  00:11, 26 November 2011 (UTC)
 * Thanks, Darklama. I might ask him. It might simplify things. --Abd 02:27, 26 November 2011 (UTC)

Change of policy status to "proposed"

 * "The page is policy, it's long been so labelled" This page, even though it was labelled such by a banned user, was never made policy and never officially consented to by the community. The RfCs above were about that and still didn't pass. We have a process, and things don't happen because you someone claims they are true. Abd was banned on Meta for such magical rethinking of policy and concepts of what is true or not, and he lost his bid for adminship on such behavior. The community even agreed before to topic ban him from discussions such as this even though the admin refused to enforce it. Ottava Rima (talk) 19:17, 26 November 2011 (UTC)


 * (edit conflict with below) I hardly know where to begin. So, the first sentence. It was marked policy by JWSchmidt, who is blocked but not banned. Any custodian could unblock him, but all of us, including myself, have wanted to see some assurances that he's not yet prepared to give. He was one of the major founders of Wikiversity, and he, marking the page as policy, represented consensus.
 * This was the original labelling as policy in 2007. Nobody removed it. In fact, the page was considered policy before JWschmidt so tagged it, see Mu201's addition of the page to Policies. On the contrary, many worked on the page, and many, many custodianships followed the policy.
 * There was an inquiry into the policy status of the page in 2010, see It went nowhere. The major issues raised seemed to be that probationary periods were sometimes longer than four weeks, and the 48-hour period to find a new mentor was mentioned; this boils down to "sometimes policies are not followed." However, SB_Johnny did remove, in June, 2001, the policy designation, but I reverted that, and the reversion stood (see a later edit by SB_Johnny, the same day, reverting back to my revision.) That was discussed here. The policy must be considered approved by defacto consensus, and should not be changed without consensus (or, at a minimum, majority support among active editors and custodians. Discussion is not necessarily required to show consensus!
 * The attempt by Ottava to change mentorship procedure this year failed, which is the best thing that can be said about it.
 * We will not examine my meta block here (I'm not banned there; a steward who promised to review the situation has gone on a long wikibreak.) We will examine, in due course, how I lost my adminship here, but not now, and there is a current RfC covering the policy that I was enforcing, with consensus unclear. Ottava's proposed "topic ban" for me was ludicrous, it only had support from a claque of non-Wikiversity users canvassed by Ottava, it was obvious. --Abd 00:02, 27 November 2011 (UTC)


 * It does seem looking back that JWS unilaterally declared this policy as "it seems to work". He did this with many policies over the years, and in most other cases they were later changed back to proposed policies, so I am doing the same here. Thenub314


 * Please read the above. Thenub, this has been discussed before. Please don't change a policy without consensus, especially not based on misinformation. I'll revert. --Abd 00:02, 27 November 2011 (UTC)
 * I reverted Abd's clear vandalism. It is against the rules to declare things as policy that were never determined such by community consensus. Abd knows this and even argued this previously here without any community support for his disruption then, making his current acts of disruption something that clearly warrants a long term block. Ottava Rima (talk) 00:05, 27 November 2011 (UTC)

Returning to the Policy/Proposed Policy comments. I had not been aware of the previous discussions (though it seems I was part of one of them and forgot) but it really seems to me that with the exception of Abd, the comments have either been in favor of calling this a proposed policy or have been generally neutral. Labeling it as a Proposed Policy seems to make sense to me. Thenub314 05:29, 27 November 2011 (UTC)
 * No, that's an incomplete analysis. You are looking at a very small number of comments, relatively recently. The point was made, not by me, that a change to policy would require site notice. If you look at the RfCs Ottava mentions, you'll see that they assumed that the page was policy, they were trying to amend it. Here is my point: it's disruptive to come along and change a page that has been labelled policy for four years, that was obviously accepted as policy by many users who edited the page and who followed it, to deprecate it to "proposed policy" based on some alleged defect in how it was created, on some claim now that there was no consensus then, when all signs point to an actual consensus, a lack of contention and acceptance, until someone, quite some time later, noticed that there was no vote.
 * What makes some think a vote is required? Whether or not a vote was taken, consensus may still exist. This is particularly true in the first stages of formation of a community. Things are done informally. What I see, reviewing the record, is almost complete unanimity, lack of objection. A bit later, there comes in some opinion, "Hey, this never was voted on, and I don't like X, so, let's call it proposed." When JWS added the policy label, replacing the previous proposed policy label, he was very popular. He was representing consensus, and the proof is the response of the community. It wasn't that nobody noticed, this wasn't an obscure page.
 * Policies are still subject to change through the ordinary editing process. If someone makes a change, it's possible that it wouldn't be noticed for a while, but next time the policy comes up as an issue, people will notice that change and can object. But that kind of change isn't what happened with JWS's label as policy. This stood for four years, until Thenub and Ottava changed it, out of the blue, without any extended discussion and without the site notice that would be required for a significant policy change, and deprecating a policy to "proposed," is a huge change. I've seen the damage from Blocking policy being labelled "proposed." It's treated as if we have no policy. If there is no policy, then custodians can do whatever they please, there is no a priori restraint, it's all the Wild West. --Abd 05:52, 27 November 2011 (UTC)
 * As the proposer of the RfC, I did not assume one way or another if this was an official policy approved of by the community. To the contrary, I was processing a change to a page through consensus, which is something I seem to be the only one willing to do. We have discussions, get community consensus for these alterations, and then get the community to approve it. That goes for actually ratifying the policy. I pushed for a couple policies to be ratified earlier this year. It ended abruptly when you decided to start trying to get in the way of every proposal. Ottava Rima (talk) 14:08, 27 November 2011 (UTC)
 * Wow! I did that all by myself? I didn't realize I was that powerful. --Abd 19:09, 27 November 2011 (UTC)