Consensus for its implementation was not established within a reasonable period of time. If you want to revive discussion, please do so below or initiate a thread at the village pump.
This is an archive of past discussions. Do not edit the contents of this page.
A single user denying another user adminship? I think that will cause an awful lot of personal grudges. In a perfect world, people would just accept the criticism and move on. That doesn't happen. Removing prod tags from articles can be contentious; denying someone adminship in the same way? I doubt too many people are interested in unilaterally deciding that and dealing with the headaches (and possibly harassment) that will come with it.
On the other hand, requiring multiple users to agree on its removal would just bring us back to a tally. X number of opposition and the RfA fails. So, as far as I can see, this just adds another layer of bureaucracy (albeit a thin layer) without doing away with any of the existing process. KafzielTalk 16:08, 30 April 2007 (UTC)[reply]
I do accept that, but this would run along side the existing RfA process, it wouldn't replace it at all, in fact users who fail proposed adminship will be actively encouraged to take out a full RfA (that could be added to the proposal). It's simply meant to be an easier way for the clear potential admins to get the mop without messing around with pile-onsupport. Ryan Postlethwaite 16:12, 30 April 2007 (UTC)[reply]
There is value to "pile-on" supports. They represent another pair of eyes that has looked through the nomination and not found a major issue that should prevent the nominee from becoming an administrator. That's not the case in this type of system: we don't know if a "successful" nominee's contributions have been researched, thoroughly or otherwise, by three people, ten, or a hundred... or if they haven't been looked at by anyone. Dekimasuよ! 03:53, 1 May 2007 (UTC)[reply]
I dislike it for the same reasons I don't use prod. It's an extra, optional step in an already convoluted process. It's easy to abuse (in this case by grossly unqualified users) and essentially just forces good users to waste time, first by removing it and then by defending themselves from attacks after they do so. I don't see any benefit here, because an RfA with absolutely no opposition is a rare thing indeed. KafzielTalk 16:31, 30 April 2007 (UTC)[reply]
I agree for this case, and I do like prod. I think I made these points in the WT:RFA thread already. Dekimasuよ! 03:53, 1 May 2007 (UTC)[reply]
If you'll read the discussion, you'll see that there is actually hearty support for this proposal! Consensus has certainly been reached that this is not only a good idea, but one worth taking the next step forward with! Cheers gaillimhConas tá tú? 09:17, 1 May 2007 (UTC)[reply]
You're going to have to demonstrate a lot more evidence of such consensus than just a bald statement like that. -- nae'blis 17:56, 1 May 2007 (UTC)[reply]
"2. Made at least 100 edits across all namespaces."
Seems somewhat subjective, and should probably get pared back some. I myself have a fairly high edit count,[1] but don't have 100+ edits in all namespaces (and of course, a non-admin just simply can't have any edits in the MediaWiki space); specifically, I'm lacking in the MediaWiki, Portal, and Category spaces (and their respective talk spaces).
I don't really have a solution to recommend, I'm just pointing out that it may need some rewording; I've been an editor for over a year and an admin for about six months, but according to this, I shouldn't de-prod a prospective admin, which just burns my ass. :) EVula// talk // ☯ // 16:28, 30 April 2007 (UTC)[reply]
EVula, I think you misunderstood the proposal. I read it to mean 100 edits regardless of which namespace the edit was in. That's a really low threshold. It means 100 edits total in some combination of Mainspace, Talk, User and User talk would be enough to qualify an editor to remove the template. --Richard 16:30, 30 April 2007 (UTC)[reply]
Yeah that's what I meant, it most probably should be worded clearer, but we need to attempt to stop disruption only accounts removing the template. Ryan Postlethwaite 16:33, 30 April 2007 (UTC)[reply]
That won't stop disruption-only accounts. I can make 100 fully legitimate edits in 1-2 hours (spelling, minor rephrasings of text, stub sorting, vandalism reverts + vandal warnings). -- Black Falcon(Talk) 16:47, 30 April 2007 (UTC)[reply]
Ah, okay, after re-reading it, that does make sense. EVula// talk // ☯ // 17:22, 30 April 2007 (UTC)[reply]
I misread this part, too. I've made an edit that might make it clearer; please feel free to revert or change it. -GTBacchus(talk) 06:13, 10 May 2007 (UTC)[reply]
My first criticism is that this proposal was already introduced at WT:RFA and failed to gain consensus. Here are some others:
Process
How do we prevent the category from being filled with hundreds of potential candidates? The only way this can result in more promotions is by reducing the transparency of RfA.
PROD was created to lessen the load on AfD. What is the purpose of this? RfA is not under any heavy load or in danger of being overwhelmed. Why create an additional layer of bureaucracy?
Removal
"Made at least 100 edits across all namespaces." I think this may need to be narrowed ... very few people have 100 portal talk edits. ;)
"A good reason to provide to the candidate as to why they are not suitable." How exactly do we define a "good reason" ... the RfA talk archives are filled with megabytes of debate on this. We will never have an objective definition of a subjective concept (unless we just coercively impose an arbitrary one).
Promotion
Is this really like PROD? If an article is deleted via PROD, it can be contested and undeleted at any time. If an admin is promoted, can s/he be recalled at any time by a single user in good standing? This will result in two classes of admins: "community-supported" and "community unopposed".
This proposal assumes that silence equates to consensus. I don't think that assumption is uncontroversial.
Hi Black Falcon! If you'll read the thread, you'll see that there is tons of support to take this to the next step. That is, consensus shows that this is a neat idea and should be taken to the next step. In an attempt to address your concerns, here is the following:
Process
We don't want to prevent the category from filling up with lots of candidates! That is, the more admins the merrier! In addition, perhaps there can be some sort of basic standard to apply this (three months of editing and 1000 edits or something). Also, PROD was not only created to reduce the load at RfA; it was also created to expedite the process, which is what PROA endeavours to do. PROD and PROA are not exactly the same; PROD is simply the inspiration for PROA
Removal
Hehe, use your common sense as to what a "good reason" could be. We don't need to define everything specifically; that is just process wonkery. 100 edits across all namespaces equates to 100 well-rounded edits, not 100 edits in MediaWiki, template, mainspace, etc.
Promotion
No, it's not exactly like PROD, hehe. With regards to these classifications, again, this is just process wonkery. Adminship is no big deal.
No it doesn't. There is a subpage (or there will be) where people can voice their affirmation if they wish. This subpage will then become the standard RfA form if the adminship is contested.
We may want to run multiple experiments. We need several people together, and we need a kind of sandbox. For now we can use subpages of this page as our experimental workspace.
I'm kinda busy, so I hope folks will pick up an experimental design by themselves :-)
--Kim Bruning 17:34, 30 April 2007 (UTC)[reply]
I am not willing to support even an experimental use of this process until several of the above concerns have been addressed. This represents a fundamental change in the way that the sysop bit is given out, and thus differs from 'experiments' in RFA formatting conducted recently. -- nae'blis 17:39, 30 April 2007 (UTC)[reply]
I agree with Nae'blis. Not even a trial run. This has almost no support from anyone at all, and countless arguments have already been listed against it. There's no need to proceed with this. KafzielTalk 17:41, 30 April 2007 (UTC)[reply]
I agree as well. This is not just a simple modification to the format. In the absence of consensus support, this proposal should not be advanced, even if only experimentally. -- Black Falcon(Talk) 17:52, 30 April 2007 (UTC)[reply]
As follows: We do a proposed adminship for someone, exactly as if it were the real thing, but we leave out the final step: we do not actually give the user the flag. We then do an actual RFA, and see if the same concerns are raised, and if the results are similar. The differences would be interesting. We can then also quickly ascertain whether peoples objections are wrong or right.
People participating are somewhat compensated, because a proposed adminship would hopefully help candidates figure out how they will do on RFA.
Basically if people want to experiment, it's none of our business, as long as they don't disrupt the rest of wikipedia. On the other hand, disrupting experiments is frowned upon.
While that sounds like a lovely experimental design, I will stake my reputation on no one passing who would not otherwise sail through RFA (due to the veto feature). Therefore the process only seems to expand the delay before someone can become gain the sysop bit. Unless you're proposing to nominate someone who is somewhat controversial, in which case I can't see what that will prove, except maybe to alleviate the concern that some users will fly under the radar (but the fishbowl effect of any experimental setting renders that data somewhat suspect, in my view). -- nae'blis 18:00, 30 April 2007 (UTC)[reply]
Ok, so in short, no harm done to wikipedia, but you're not participating because you think it's a waste of your time? Fine. Done and done. :-) --Kim Bruning 22:55, 30 April 2007 (UTC)[reply]
Note that your own requirements are what will cause any fishbowl effect. You required it, so you will have to live with it. Consensus cuts both ways. :-P --Kim Bruning 23:29, 30 April 2007 (UTC)[reply]
I need a noun, Kim. What requirements? That we not change things unilaterally? I'm perfectly happy with a fishbowl experiment, I just don't see what you hope to gain here. -- nae'blis 05:01, 1 May 2007 (UTC)[reply]
You don't want the test to be live, right? --Kim Bruning 06:07, 1 May 2007 (UTC) Anytime someone uses the word "Unilateral", I worry if they have read WP:BOLD recently. ^^;; [reply]
Correct, I don't support an 'experiment' in which the sysop bit can be granted through this experimental procedure. Let's not be misleading, WP:BOLD has been abbout updating articles since at least October 2003. Perhaps you were thinking of Bold, Revert, Discuss? In that case, plenty of people have reverted the idea, and here we are discussing. -- nae'blis 17:53, 1 May 2007 (UTC)[reply]
I think experimenting with this process will be extremely difficult, to put it mildly. Suppose we pick a well-qualified user and run them through this. If the user makes it through this process (in either experiment, the "live test" or the fishbowl), we really might not learn anything... and I literally mean nothing here - if no one removes the tag, and there's no area for discussion, we have no record at all of peoples' opinions on the candidate that we can analzye. If the user fails the process, it's very hard to learn anything from that either: was their reasoning good reasoning? Did they really remove the tag because of that reasoning? Were they acting instead to disable an experiment they disagree with or try to avoid a process they don't like? We may actually learn something from the follow-up RFA, but it still won't tell us about the experiment with this process. If the results are in synch, it doesn't say this is a good process... and if an RFA passes where this fails, it's to be expected (many RFAs pass with some oppose comments), whereas if this passes but RFA fails, it's bad but impossible to distinguish from an experiment that didn't get enough attention. Mangojuicetalk 20:02, 1 May 2007 (UTC)[reply]
A trial would tell us quite a lot. Just run this process for a few months and if at the end of the trial the number of administrators successfully promoted is significantly higher and a good proportion of the successful candidates are being promoted through this process, then it's been a success. If it has no effect, or the number of successful candidates actually reduces, then it's been a failure. A single experiment, on the other hand, would tell us little. --Tony Sidaway 14:22, 11 May 2007 (UTC)[reply]
But there's nothing near the level of community consensus to even run a trial of this that actually grants the tools. You run the risk of creating two echelons of admins: those approved with explicit trust through Requests For Adminship, and those granted the tools with tacit trust through Proposed Adminship (whether the trial succeeds or not). I don't think anyone believes that is a good outcome, and there are a lot of concerns below that are being brushed aside (not a good idea with landmines). I'm not the only user to pass RFA unopposed and I didn't find it too onerous or bureaucratic; the only people who should be passing this process should be unopposed at RFA as well, so the problem isn't procedural, it's perceptual. I've had seven people reject my offer to nominate them in the last month...solve that, and we solve the shortage. But not one of them would have passed this process (six had prior RFAs, and one has some problems with communication). -- nae'blis 15:02, 11 May 2007 (UTC)[reply]
I think that the only reason why anyone should be stopped from becoming an administrator is if the high number of administrators causes legal trouble to Wikipedia somehow, because everyone would be able to see deleted content.
Another terrible thing is the ability of administrators to unblock themselves: if there were no legal problems and administrators could not unblock themselves, then every user should have the administrator tools, and if the user abused them, the user would just be blocked and would not unblock themself. A.Z. 06:22, 1 May 2007 (UTC)[reply]
I think there are other issues as well. Look what damage User:Robdurbar did in just 17 minutes. Even if there were no legal problems and admins could not unblock themselves, vandals would stop inserting gibberish into articles and delete the article instead, which would not only be another problem for RC patrollers but would stop readers from seeing the article. Hut 8.5 17:53, 1 May 2007 (UTC)[reply]
A.Z., this is not the place to discuss if everybody should be an admin or not, this page is for discussing the proposal for PROA. If you must, bring this up at WT:RFA. Greeves(talk • contribs • reviews) 21:43, 3 May 2007 (UTC)[reply]
The reason PROD works is that, at the time it was initiated, a large majority of AFD debates turned out to be unanimous. The reason this doesn't work, and indeed isn't needed, is that only a small minority of RFA nominations are unanimous. Radiant! 12:41, 1 May 2007 (UTC)[reply]
I hope that, given your history with the prod system, editors will listen to you when you say this. I tried to say it a few times, and it had little effect. Dekimasuよ! 13:14, 1 May 2007 (UTC)[reply]
There is a minor niggeling problem where we need to promote more admins.... :-) --Kim Bruning 15:59, 1 May 2007 (UTC)[reply]
Too bad you don't nominate people anymore, then.... :-) Dekimasuよ! 16:02, 1 May 2007 (UTC)[reply]
If I nominate them, will you support? --Kim Bruning 16:17, 1 May 2007 (UTC)[reply]
Probably, but someone will oppose. Dekimasuよ! 16:20, 1 May 2007 (UTC)[reply]
I was thinking of writing a bot for the purpose. However, it seems that our bot policy is utterly insane. --Kim Bruning 16:47, 1 May 2007 (UTC)[reply]
I understand your concern; what I don't understand is why more RfAs aren't unanimous. --Iamunknown 19:27, 1 May 2007 (UTC)[reply]
Because almost everyone sucks in one way or another. KafzielTalk 19:30, 1 May 2007 (UTC)[reply]
Exactly. I wish we could get beside ourselves, realize that we are not all perfect and try to identify editors who would act well as administrators including, but not limited, not deliberately abusing power and being willing to realize when they make a mistake. Incidentally, I believe most editors at WP:RFA right now could be characterised as such and would be excellent administrators. Most opposes I have seen haven't been concerned with this, though, but instead with edit counts, project space edit counts, article edit counts, WP:AIV edit counts, etc. --Iamunknown 19:34, 1 May 2007 (UTC) I don't mean to make a point about edit counts, those are really just the opposes I could identify off of the top of my head that I considered rather silly. --Iamunknown 19:35, 1 May 2007 (UTC)[reply]
Well, I think people do that already. No one's insisting that admins be saints ... just that they be experienced, civil, and psychologically stable (not the best term, but it's all I can come up with in this tired state). As for opposes based on edit counts, those are really just a shorthand way of stating "this user lacks experience in discussing policies and participating in Wikipedia process". That certainly is of some importance to adminship. The fact of the matter is that most RfAs (when we exclude nominations that are SNOWballed) are successful. So, what if they're not unanimous? Is 50/0/1 result all that different from a 50/1/1? -- Black Falcon(Talk) 19:46, 1 May 2007 (UTC)[reply]
I understand that most successful RfAs are unanimous. That does not, however, say anything about how well the RfA process performs. What of the editors who are not currently administrators but who would not abuse the tools and would listen to justified complaints, but who will likely not become administrators because they have gotten into a few tiffies? We can't judge that the RfA process works by how many nearly unanimous RfAs pass; we must consider that many RfAs never happen because people are intimidated by it, think that they must get at least one FA under their belt, develop one JS tool, etc. --Iamunknown 19:53, 1 May 2007 (UTC)[reply]
Maybe they should do that stuff. Wouldn't hurt. But I don't want to get into the territory of codifying which opinions are valid and which aren't. As far as I'm concerned, there aren't any invalid opinions. Just because some guy wrote a humorous essay about editcountitis doesn't mean it suddenly becomes law. In fact, editcountitis is quite explicitly not law; it's not even a guideline. It's nothing. And the one on meta isn't, either. Opposition based on edit counts, project participation, featured article writing, or even personal grudges, racism, etc. all have to be valid if we're going to even try to pretend we're seeking consensus on RfA. We need more good admins, not just more admins. Crap admins just end up wasting other people's time. KafzielTalk 20:07, 1 May 2007 (UTC)[reply]
I did not suggest that we promote bad administrators just for the sake of promoting more administrators; I suggested that some administrators who are not promoted would nonetheless be good administrators. Do you think, even if our stated goal is to get consensus to promote good administrators, that no opinion should be discounted? --Iamunknown 20:12, 1 May 2007 (UTC)[reply]
Yes, that's what I think. I'm all for moving to a straight vote system, with no justifications necessary at all. Cut through the BS and get a quick, visible decision based on pure percentage. Voting is indisputably the best way to make quick decisions in large communities. But if consensus is really what everyone wants, then we have to take every opinion into account, even the ones we don't like. Even the ones we think are unfair or ridiculous. KafzielTalk 20:19, 1 May 2007 (UTC)[reply]
Then go ahead and get rid of Wikipedia:Consensus. In the meantime we'll keep trying to design or encourage a system of promoting administrators that is consensual. You're encouraged to contribute! --Iamunknown 20:21, 1 May 2007 (UTC)[reply]
Hey, come on - don't patronize me. I am contributing. I'm saying that your suggestion of throwing out opinions goes against the very meaning of consensus. Regardless of my own feelings about consensus vs. voting, consensus means that everyone has a say. And this little project has nothing to do with consensus; it suggests that the best way to find admins is to choose them without any discussion at all. KafzielTalk 20:28, 1 May 2007 (UTC)[reply]
I'm sorry that I was patronizing. I should have re-read my comment before posting it. I concede that sometimes I feel a vote would be easier than all of this madness; but it wouldn't be consensual and, to be honest, I'm a stickler for consensus. I really like it. I guess we shouldn't just concern ourselves with consensus; we should concern ourselves with both consensus and "adminship is not big deal" and then try to reconcile the differences between those two. A lot of RfA patterns seem to be going away from the two towards...I dunno what. That is unfortunate. --Iamunknown 20:32, 1 May 2007 (UTC)[reply]
That's probably where we differ. "Adminship is not a big deal" was some offhand comment Jimbo made years ago. Wikipedia was a very different place when he said that. Today we're more than ten times the size of Britannica, with a new registered user every 2 seconds or so. People keep repeating the same old rhetoric, but recent events have shown that adminship is a big deal. Not every mistake is a little one, and not every mistake is fixed in a second or two. Some mistakes cause emergency de-sysopping and weeks or even months of debate and discussion. When we screw up, the Wikimedia Foundation cares. The press cares. The editors care. The public cares. Quality admins matter. We're big enough now that we need to be focusing on quality, not quantity, and that goes double for admins. A system that lets people be admins by virtue of nobody noticing them is definitely not the solution. KafzielTalk 20:44, 1 May 2007 (UTC)[reply]
"But if consensus is really what everyone wants, then we have to take every opinion into account, even the ones we don't like. Even the ones we think are unfair or ridiculous."
Whoa, whoa, whoa! At least one method of determining consensus in the real world excludes invalid opinions. I could probably sort out a way to implement it here, but what it boils down to is that if the blocking opinion doesn't persuade others in the group, then it is not a 'valid block' because it doesn't represent an affirmative decision by the group; it's just one person's axe to grind. An analogous situation on RFA is when one person opposes because the person is Black/Hindi/11/whatever, and 17 other people come by and comment that they are not persuaded by that opinion. So long as the opinion is not widely held, the RFA still passes (which is why we get to call it rough consensus). If someone brings up an edit war or a racial epithet or whatever, and 17 people oppose based on it, then it's a different matter. Make sense? It's not binary, voting/consensus... -- nae'blis 23:11, 1 May 2007 (UTC)[reply]
But your example proves what I'm saying. The invalid opinions need to be identified and discussed in each case. If enough people do agree that race (for example) is a factor for whatever reason, it's still consensus. The opinion itself is not automatically invalid; it's only made invalid by a louder majority disagreeing with it. We don't have precise rules about what is and is not immediately tossed out. It's impossible to flatly say that any particular objection is universally unacceptable. Each case needs to be discussed on its own merits, every time.
Now, in this proposal, it only takes one objection to sink the nomination. Is the objection valid? We can't be certain, because there's no debate involved. You just remove the tag, post your reason, the end. And if there is a big debate about whether it's valid or not—which I guarantee there will be, almost every time—then we're back to the standard RfA anyway, except that now it's spread out all over the talk pages of different users.
This proposal ranges from no consensus at all (unilateral rejection) to even lengthier and more disorganized discussions than we already have (numerous debates on countless talk pages). Either way, it doesn't work. KafzielTalk 23:37, 1 May 2007 (UTC)[reply]
I think we're agreeing very loudly, Kaf. :) -- nae'blis 23:49, 1 May 2007 (UTC)[reply]
This system does not establish that the community has trust in the user, which is a requirement to become an admin. HighInBC(Need help? Ask me) 17:54, 3 May 2007 (UTC)[reply]
There are some objections to this, mainly along the lines that "it will never work" presumably because every proposed adminship will be opposed by someone. While that's a valid objection, it's something we could test quite easily by a trial run of, say, three months. If at the end of that time hardly anybody has been promoted using this method, then it has failed. But looking at the quite large number of no-oppose pile-ons we get, I can't help feeling that there might be a place for this method. --Tony Sidaway 11:11, 2 May 2007 (UTC)[reply]
I agree with Tony here, to all the people who say it's not going to work, it seams like it certainly merits a trial, be it temporary if you wish, but we won't know until we put it into practice. Of course there may be a few teething problems at first, but they can easily be ironed out. If it doesn't work, it doesn't work simple as that and we can go back to the drawing board. Ryan Postlethwaite 11:51, 2 May 2007 (UTC)[reply]
Well I reckon we can remove as many as we need to as long as we're not doing so in order to disrupt the process. That is to say, we should have a valid objection. Using a highly tendentious objection, such as "you don't have 5,000 Wikipedia-space edits", might be considered disruptive if it's done just once. Using an obscure objection on every single candidate, or most of them, might also be considered disruptive by the community. But most objections are likely to be reasonable so we needn't concern ourselves about this. --Tony Sidaway 12:24, 2 May 2007 (UTC)[reply]
But then we're back to the problem Nae and I were discussing above. Objecting because of edit count is valid. There's no rule against it. I happen to think a few thousand edits is a reasonable standard, and just because somebody once wrote a humorous essay to the contrary doesn't mean I'm wrong.
I don't usually participate in RfAs unless it's clear that consensus is going to be hard to determine. But this proposal has nothing to do with consensus, so I'm not going to let it get pushed through just because I'm afraid people might not like my reasons for opposing. This can't work if everyone who opposes for their own reasons is going to be accused of disrupting the experiment; it's not our fault that a single-point of failure system is a bad idea. If somebody prefers a system where one person's "unreasonable" standards can be attacked, debated, overruled, and disregarded, we already have that at RfA. KafzielTalk 12:59, 2 May 2007 (UTC)[reply]
The point is, that it's not a bit deal if the template is removed, if it is, the user can just run through the normal RfA procedure with no questions asked. People probably will remove the template for there own stupid reasons, but if they are a valid concern, and not something stupid like Hasn't got 3000 wikipedia space edits. If someone says a user with 2,500 edits is too low, then so be it. If concensus on the user in questions talk page that it shouldn't have been removed, then it could be reinstated. Ryan Postlethwaite 13:05, 2 May 2007 (UTC)[reply]
Although I personally wouldn't agree with an objection based solely on edit count, there is consensus that this can be a reasonable reason (and we can all agree that experience of editing is an important factor). So I don't think someone who removed the tag from candidates who had less than X,000 edits for some low value of X would be acting disruptively. Someone who removed the tag from a few candidates solely because the candidates all had less than 10,000 edits, however, might have a hard time arguing that he's not being deliberately disruptive. So I don't anticipate many serious challenges to removal of the tag, and those that do emerge and have consequences are likely, in my opinion, to be cases of egregiously bad behavior. --Tony Sidaway 13:09, 2 May 2007 (UTC)[reply]
And what would you propose to do if someone removed the tag for failure to have contributed to 1FA, use edit summaries > 95% of the time or receive Wikiproject endorsement? I would suggest that there be an area for comments where editors are encouraged to record negative information but not "blackball" the candidate on relatively minor grounds. --Richard 15:34, 2 May 2007 (UTC)[reply]
And just what exactly is a "relatively minor ground"? Just a few days ago, someone argued that issuing a death threat to another editor, if it was in the past and has since been apologised for, can be overlooked. I disagree with that. Who's to say who's right? Also, although every can agree that requiring 100,000 edits to support is ludicrous, we certainly can't agree on the cutoff point (if any) ... are 10 good edits enough? How about 500? 2000? 3432? How are we going to judge what is a disruptive reason to oppose and what is not? I strongly oppose this kind of trial, wherein candidates may actually be promoted. Any promotion process should go through consensus (even if it's rough consensus) ... this proposal does not have consensus support. -- Black Falcon(Talk) 18:29, 2 May 2007 (UTC)[reply]
Issuing a death threat to another editor? I presume you mean that I was arguing against it, well yes I was because it wasn't even a death threat. Anyway..... If people have a reason they believe they should oppose a candidate, they are more than welcome to remove the tag, there's no minimum standards, its up to personal judgement from the person removing the template. This proposal doesn't have consensus oppose either. Ryan Postlethwaite 18:34, 2 May 2007 (UTC)[reply]
What? I'm not sure what you're referring to, but I was referencing a discussion I had with someone else about the validity of reasons to oppose (about a week ago at WT:RFA). He stated that such threats could be overlooked if they were old and unique incidents and I disagreed. The reason I brought that up is to illustrate the inherent impossibility to judge what is or is not a valid reason to trust or not trust someone. Anyway, to get back on topic ... I agree that this proposal doesn't have consensus oppose either, but since it constitutes a change in how admin promotions are handled, lack of consensus should translate into a lack of change. Per WP:PG, "a rejected page is any proposal for which consensus support is not present" (emphasis added). It's the presence of consensus support that mandates a change; "no consensus" translates into "no change". -- Black Falcon(Talk) 19:02, 2 May 2007 (UTC)[reply]
I think the sense of this is that, unless someone gives a really egregious reason to remove the tag, the community will accept that removal, and the user still has option of applying for adminship in the usual way. --Tony Sidaway 19:29, 2 May 2007 (UTC)[reply]
What if the tag is removed with the reason "This form of becoming an admin does not provide the community discussion needed to determine community trust. I do not know this person well enough to skip that step."? HighInBC(Need help? Ask me) 18:03, 3 May 2007 (UTC)[reply]
That seems a bit pointed (and I'm an opponent of the process as outlined). However, "I do not have confidence that this person is trusted by the community, as evidenced by <x>" would seem reasonable and actually related to the candidate. -- nae'blis 18:29, 3 May 2007 (UTC)[reply]
Unless I can see a demonstration that the community has accepted this as a way of becoming an admin(by a very broad consensus composed of people from all over wikipedia), I don't see how that would be disrupting Wikipedia to prove a point. You say this system is based on lack of objection. HighInBC(Need help? Ask me) 18:30, 3 May 2007 (UTC)[reply]
Another possible reason for opposing could be: "I do not trust a candidate who seeks to become an admin through a process that lacks community support." That's directly relevant to the candidate and his behaviour on matters related to community consensus. -- Black Falcon(Talk) 18:34, 3 May 2007 (UTC)[reply]
No trial can start unless and until there is community consensus, so someone removing the tag for that reason would be doing so against consensus. He might get a few fingers wagged in his direction for acting disruptively. --Tony Sidaway 09:18, 4 May 2007 (UTC)[reply]
Well, yes, but my statement assumed that the trial begins now, at a time when it doesn't have consensus support. Under these circumstances, removing the tag for that reason would not be against consensus, even though it may comes across as more than a little WP:POINTy. -- Black Falcon(Talk) 21:54, 5 May 2007 (UTC)[reply]
Well good, I think we are on the same page. I'm not interested in a trial that we can't all back 100%. If we can agree to a trial, good, but it isn't worth it unless and until we do. --Tony Sidaway 22:41, 5 May 2007 (UTC)[reply]
Some of this have been posted already, but I may as well say them again. Just for clarity, as much as anything.
The Process
I like this. It's simple, it's contestable, therefore it's balanced, and it seems effective. The current "RfA patrol" will probably spend their time patrolling the appropriate category, so it's not like people will get through unseen.
Addition of Template
Seems simple enough. Clarification on HOW it must be substituted would be appreciated.
Removal of the template
I don't like #1, because I believe that admins are everyone's responsibility, and no amount of experience (or lack of, for that matter) should influence what say someone has in this regard. After all, admins can effect accounts that are less then 30 days old, so the reverse should also apply.
I'm very passionate about the comments concerning edit count on WP:ANOT, and thus firmly oppose #2.
#3 seems like common sense. I'd have the whole thing run with ONLY #3.
I also like the 30 day cool down period before adding the template again. Thumbs up on that.
Directions for closing bureaucrat Before closing any expired proposed adminship, bureaucrats should check through the contributions of the proposed administrator and decide themselves if they would make a suitable candidate.
To me, this undermines the entire process. If this is removed, as well as points 1 and 2 mentioned above, I will support the program. At the moment, I am neutral. ~ G1ggy!blah, blah, blah 11:21, 2 May 2007 (UTC)[reply]
We need to to have some way to stop the trolls and vandals from simply removing the template, that's why we need the restrictions on who can remove the template, otherwise, an IP page blanking a userpage of a candidate would in effect mean that they fail, it's not about editcountitis, it's about protection. Also, the final 'crat check is required, to make sure that somebody slips through that shouldn't do, it doesn't undermine the whole process, it's just a safety net. Ryan Postlethwaite 11:49, 2 May 2007 (UTC)[reply]
I don't see why a blanking (that removed the template) couldn't be reverted, with a note left on the talk page, or in the template, saying that an rvv took place. As for the b'cat, if a safety net must be imposed, I'd like it to be VERY loose, so that 95% of candidates at least get through it. It should ONLY be there to remove those whose template probably wasn't even viewed. It shouldn't be another review process.~ G1ggy!Reply 23:48, 2 May 2007 (UTC)[reply]
Along those same lines, of the problems that occasionally comes up with PROD is that it gets applied to articles it shouldn't (because they've already been prodded, or because they've been through AFD, etc). I'm assuming no one supports people trying Proposed Adminship more than once, but is there a mechanism/way to track who has been removed, and (ideally) the reason given)? That would also help spot the spurious removals, or at least track trends on why the template was removed, if nothing else. -- nae'blis 17:54, 2 May 2007 (UTC)[reply]
Forbidding self-noms would help, but you could easily get a sockpuppet/meatpuppet to nominate you. I'm not sure how we propose to coach b'crats to 'vett' candidates, since it's not something any of them agreed to when they ran for RFB. -- nae'blis 16:28, 3 May 2007 (UTC)[reply]
While I personally still think this process is doomed to failure, I have made a number of changes to the proposal itself in the interest of tightening language and reflecting concepts I see on this talk page. In summary:
I removed the concept of self-nominations. We do not allow the same administrator to delete a prodded article as nominated it, and the sysop bit is considerably less overturnable in case of malfeasance.
Strengthened the requirements for declaring the reason for removing the template.
Removed the possibility of re-nomination through this process, except when the candidate takes down the template themselves. We don't re-PROD articles, and any valid reason given once is probably reason to go to RFA for the full process. I also clarified that while it's not technically forbidden to run through this process after a failed RFA (I'm thinking mainly of the 60th-edit RFAs that are doomed for experience, not actual malfeasance), it is likely to draw an objector or two.
Reworded the b'crat prerogative to 'review the candidate' to simply state that they may remove the template themselves as a normal editor, instead of promoting.
Make sense? Check my edits for sanity, clarity, bias, etc, please. -- nae'blis 18:48, 3 May 2007 (UTC)[reply]
Does anyone else think it would be a good idea to get a straw poll going for this? It's such a big change to how adminship can be gained that a full community input should be gained into it. We're not going to get consensus on this page, and even WT:RFA isn't going to get a wide enough view. I'd like to propose we plan the questions for a week, mention the poll on the village pump, WP:AN, then let voting prevail for a further week. Sound OK? Ryan Postlethwaite 19:21, 3 May 2007 (UTC)[reply]
My suggestion: What if after someone supports or opposes the candidate, we have someone else check to see if the reason given was valid and then the crat checks again when closing it so we get a few pairs of eyes in there. Greeves(talk • contribs • reviews) 22:50, 3 May 2007 (UTC)[reply]
Wikipedia is supposed to be a community by and for people. By simply allowing a user to be promoted to admin, it means throwing out all experience and community we have strived for for years. The only way we can continue a reliable and reasonable way of promoting admins is to keep the voting process where it is. Jmlk17 23:45, 3 May 2007 (UTC)[reply]
This takes the concensus that Wikipedia is built on right out of everything. By doing this, a user can simply be promoted without the voice of the community. If I were to put the tag on my good-faithed, but not yet ready friend, Bmader, and he goes inactive for 7 days, and no one visits his userpage, he's given access to the tools. Terrible idea. Cool Blue 22:29, 5 May 2007 (UTC)[reply]
Not if RFA isn't edited at all in that week (OK, this is unlikely), or enough people enable the "Don't show page content below diffs" user preference. Watchlists and categories just don't mix. —Cryptic 00:56, 6 May 2007 (UTC)[reply]
Lack of evidence of community distrust is not the same thing as evidence of community trust. This system gives us the former, we need the latter. HighInBC 22:38, 5 May 2007 (UTC)[reply]
Whilst I disagree, I can fully understand that. Ryan Postlethwaite 22:41, 5 May 2007 (UTC)[reply]
I think this is making adminship more than it really is. It's just access to some tools that can if misused do a little more damage, and temporarily at that, to the project. Recently an administrator whom nobody had any reason to distrust, and who was sysopped after a full Request for adminship discussion, deleted the main page, blocked Jimbo Wales and several other prominent editors, and did some other damaging acts. This happened at perhaps the quietest time for Wikipedia, around 1000 GMT. The account was completely disabled in about fifteen minutes, and no permanent damage was done. So let's not exaggerate the power of the administrator to cause damage, and let's not pretend that our administrators are all angels. They're not. --Tony Sidaway 22:52, 5 May 2007 (UTC)[reply]
The only reason permanent damage was not done is because the administrator in question either was supremely unimaginative, or wanted to create as much noise as possible without causing permanent damage. Obviously I'm not going to spell methods out, but as an intelligent fellow and a former admin yourself, I'm certain you can think of them. —Cryptic 00:56, 6 May 2007 (UTC)[reply]
I'm not sure what your argument is here. Are you saying that, although the current system does not ensure that administrators selected aren't going to cause damage, it magically ensures that those who do go postal only cause limited damage? Do you think that the proposed system here would select more imaginative, more malicious administrators? If so, what is your reasoning for this? --Tony Sidaway 08:27, 6 May 2007 (UTC)[reply]
The current system does do a fairly good job in selecting against the time bombs. (Not to say that I like it; I don't. Heck, who does?) My reasoning is that when proposed deletion - the process on which this is based - screws up and deletes an article that should've stuck around, it tends to do so in much more obvious cases than when afd does. I have nightmare visions of a Category:Proposed administrators that's as full as, say, Category:Wikipedia administrator hopefuls, and inappropriate administrators slipping through due to too little review. —Cryptic 11:03, 6 May 2007 (UTC)[reply]
We have bureaucrats to review the candidates and reject the obviously bad ones. In any case we cannot know whether the category will fill up with no-hopers unless we give it a trial. --Tony Sidaway 11:16, 6 May 2007 (UTC)[reply]
It's not the obviously bad ones that concern me. I'm thinking more along the lines of RFAs that are initially successful, but turn around after someone points out egregiously bad but not-immediately-obvious behavior in the past. In at least one case I can recall - though I don't want to name names and imply the candidate would've turned into another Robdurbar - said egregiously bad behavior was on a page that had been deleted before the RFA, and it wouldn't be at all reasonable to expect that the closing bureacrat would find it on his own. —Cryptic 12:03, 6 May 2007 (UTC)[reply]
In this proposed method, just one person with a valid reason (not necessarily one we would all support, just one that makes him feel uneasy) could remove the proposed adminship tag. If there is any doubt about whether a given editor could be trusted, the tag should be removed, and the candidate can apply using the normal method. So there really shouldn't be any problem here in my opinion.
But obviously we don't want to run a trial unless we're all agreed that this method is worth a go and isn't so potentially dangerous that even a trial might lead to serious problems. So while I'd love to see a trial, I see my task at the moment as trying to answer reasonable objections like this. I hope you'll continue to apply some thought to this proposal, because I think that the more you consider it, the more you'll see that it's a pretty safe one. --Tony Sidaway 12:36, 6 May 2007 (UTC)[reply]
Admins can cause damage to the project in one of three forms (this is damage they can cause in addition to the damage any regular use could cause):
Rampaging around like a bull elephant in heat, deleting articles and block users left and right. These actions generally target high-profile pages, attract a lot of attention, and are caught rather quickly and reverted. They cause no long-lasting damage.
Admin abuse of tools. These are cases where admins deliberately break policy to achieve their desired results. This includes actions such as blocking editors in order to push a particular POV, misrepresenting consensus to keep/delete articles that they like/dislike, speedy deleting articles that do not meet the speedy deletion criteria, and protecting pages to shut out other editors. These types of actions attract significantly less notice (especially inappropriate speedy deletions) and can cause significant damage. The damage comes not only in the form of deletion of valid content, retention of unencyclopedic content, and so forth, but also (and perhaps more importantly) in the alienation of present and potential editors.
Admin misuse of tools. These are cases where admins violate policy, but do not do so deliberately. This includes actions such as being too quick to block editors, misinterpreting consensus, and misapplying (or failing to apply) policies. This type of misuse usually results from inexperience, but it too can cause damage to the project, as above.
I agree that "adminship is not a big deal", but I disagree that the phrase implies that admin tools are harmless. In case you're interested, my understanding of the phrase "adminship is no big deal" is summed up in this edit. -- Black Falcon(Talk) 23:39, 5 May 2007 (UTC)[reply]
I would rather keep the de-sysopings to a minimum. HighInBC(Need help? Ask me) 22:54, 5 May 2007 (UTC)[reply]
You both appear to have misunderstood my points. RFA doesn't tell us who can be trusted. The damage done by sysops is very limited. In all the history of arbitration there have been a small number of cases of administrators abusing their tools to push a point of view. Even good administrators will sometimes block editors to quickly, misinterpret consensus, and misapply policy. We've survived quite well with imperfect administrators. I don't see any reason to suppose that things would change for the worse under the proposed system. --Tony Sidaway 23:45, 5 May 2007 (UTC)[reply]
The issue isn't that RfA guarantees perfect admins ... it doesn't. And, as you noted, even good admins make mistakes. It's just that the only way this proposed trial can increase the number of admins is by decreasing the extent to which editors review candidates. Such a process is bound to result in the promotion of admins who are more likely to make mistakes. -- Black Falcon(Talk) 23:55, 5 May 2007 (UTC)[reply]
On the contrary, this proposal doesn't put any barriers in the way of review of candidates. Anyone who wants to can do everything on the user's talk page that he would otherwise do on RFA, and unlike in RFA it takes just one opposer to halt this process completely. No, I think that if this increases the number of administrators it is likely to do so by attracting good candidates who for whatever reason are not motivated to use the current process. --Tony Sidaway 08:22, 6 May 2007 (UTC)[reply]
I for one, strongly object to the proposed template and process. I do not feeel that it is likely to reasult in excluding candidates who are unworthy of trust. DES(talk) 21:28, 9 May 2007 (UTC)[reply]
Rather than templates putting some user in some category, I would prefer a central page where those users are listed, so that I can watchlist it. I can't feasibly watchlist a cat for new entries. Radiant! 08:42, 7 May 2007 (UTC)[reply]
I can't see any drawbacks with this method, so long as we still require a tag on the user's page itself (mostly to facilitate discovery of repeat tagging, but it also allows watchlist discovery from that end). It also gives us a sort on time for free, instead of needing to maintain clunky "Proposed administrators as of 7 May 2007" categories. —Cryptic 10:19, 7 May 2007 (UTC)[reply]
Yes, a list page would be very helpful. Any comments added to the page could be removed to the candidate's talk page if appropriate, or else the talk page of the list page. --Tony Sidaway 10:47, 7 May 2007 (UTC)[reply]
How would a bot do that (since that's the only near-foolproof way to avoid the 'human error meant this person wasn't properly vetted' failing of this process? Walk through the automatic category and assign/remove any new entries since its last pass to the list page? I think less often than the current RFA-bots would be fine, maybe twice a day, though running it at the same time/on the same bots would be good for other reasons. The edit summary should definitely name all additions, and probably all removals. -- nae'blis 11:59, 7 May 2007 (UTC)[reply]
The edit summary cannot be guaranteed to be big enough to include all usernames, but yes such a bot could do the edit in several passes. But I was really just thinking of a user adding a template to the list. If a link to the user's talk page is included by the template, that's all that needs to be done. No need for any categories. An objector would remove the link from the list and provide a reason on the user's talk page. --Tony Sidaway 12:05, 7 May 2007 (UTC)[reply]
Interesting how the more we fiddle with this proposal, the closer it gets to RFA and the more acceptable it seems to get to me. I would place such a list at the top of the RFA page itself, similar to the Uncontroversial Moves section of Requested Moves or the History-only/Contested Prod sections of Deletion Review. But then, if you're willing to do that and think you'll pass unanimously, we go back to the old question of why not just run through RFA? Hell, even I passed unanimously and I'm neither the best nor the brightest (although you might have been on break then, Tony). ;)
But I am mystified at some people's reluctance to go through RFA even when they will likely win. I'm 0 for 8 on nominating people so far; one was rejected at RFA, and seven have declined to run/run again. -- nae'blis 12:51, 8 May 2007 (UTC)[reply]
Yeah, this is a good example of circular reasoning. People point to RFA's alleged harshness as a reason for why they don't want to run, and then claim that RFA is harsh because all these people aren't running. I'm at a loss as to how to break this negative spiral, though. Radiant! 13:19, 8 May 2007 (UTC)[reply]
I think this proposal could, no promises, be a way of breaking the spiral, by providing a low-cost, lightweight way of testing the water. --Tony Sidaway 08:55, 10 May 2007 (UTC)[reply]
I was interested in becoming an admin for quite a while before I finally went through with RfA, and "harshness" was a big part of why. RfA noms get picked apart for all sort of reasons, and someone's very minor flaws can be blown way out of proportion. I don't like this proposal, but it probably would help get some qualified but timid candidates approved. —dgiestc 23:31, 8 May 2007 (UTC)[reply]
You know, just today I was thinking, "Boy, I wish we had more timid admins." :P
The way I see it, the crap you get at RfA is nothing compared to the crap you get the first time you block an experienced user for 3RR, or even the first time you speedy some garage band's article. The only problem I see with the harshness of RfA is that people expect you to sit there and grovel, rather than standing up for yourself like an admin should. I certainly don't think we need more admins who can't even handle a little unconstructive criticism. KafzielTalk 23:46, 8 May 2007 (UTC)[reply]
Yeah, I didn't mean to say "let's encourage people who can't handle criticism to become admins". What I was trying to get at is that criticism in RfA isn't really a fair fight because it works much more like a vote than a consensus discussion, and people who rebut their accusers tend to be accused of harassing them or not accepting of others' criticism. —dgiestc 03:19, 9 May 2007 (UTC)[reply]
The moment we start listing info on a central page, we've lost the entire point. There's no way a single-central-page concept is going to scale. (Unless I'm terribly wrong) --Kim Bruning 14:15, 9 May 2007 (UTC)[reply]
My problem with a central page is not that it "loses the point" (it doesn't, and someone will make the tally page regardless of if it's official or not), but that there's no way to tell if someone has been vetted or not under this system. So we may actually be taking up more time with PA than with RFA, as people will often stop voting at 20-30 supporters when it looks like a clear winner (my RFA was clearly an outlier). With PA, everyone might try to look at every application (in case no one else has), therefore taking even more time away from the encyclopedia. I think {{prod}} has some sort of {{prod-2}}, but that's getting even further afield from the original intention, which is lightweight (and in my opinion, too lightweight to fly here). -- nae'blis 19:25, 9 May 2007 (UTC)[reply]
I don't see a problem with having plenty of scrutiny over the candidates. Ideally lots of people should independently look at RFA candidates, but usually what we get is sheep-voting and pile-ons. This method doesn't suffer from that drawback. --Tony Sidaway 08:57, 10 May 2007 (UTC)[reply]
I think the term "sheep-voting" drastically understates the capabilities or the responsibility of RfA participants. While probably true that people are more likely to "go with the crowd", I think most RfA regulars look at a candidate's history and make their own decisions. However, if people are overwhelmed with dozens of candidates (as is possible under this method), they may change to the strategy of not looking at a candidate's edit history with the hope that someone else will do it for them. -- Black Falcon(Talk) 17:49, 10 May 2007 (UTC)[reply]
Yes, that's a reasonable worry. We could only really see if we gave it a trial. Remember that promotion by this method is at the bureaucrat's discretion. --Tony Sidaway 17:52, 10 May 2007 (UTC)[reply]
RFA is really not so harsh as people think. >Radiant< 15:10, 9 May 2007 (UTC)[reply]
I'm sorry, but but please provide a reliable source for your claim; tertiary sources, such as Wikipedia, should be eschewed in favor of primary or secondary sources. Really, Radiant, I'm disappointed in you. :P EVula// talk // ☯ // 15:20, 9 May 2007 (UTC)[reply]
It's sad seeing this pic isn't it, just think, if PA was active the hundreds of comments that have been made in those 100% RfA's could have been used on something constructive! :-) Ryan Postlethwaite 19:17, 9 May 2007 (UTC)[reply]
If PA was active, a third of those users (anyone not at 100%) would have been rejected. KafzielTalk 19:26, 9 May 2007 (UTC)[reply]
And free to run through the normal RfA procedure. Ryan Postlethwaite 19:27, 9 May 2007 (UTC)[reply]
Which would put us right back to having those hundreds of comments, just slightly delayed. KafzielTalk 19:33, 9 May 2007 (UTC)[reply]
Actually, Ryan, I imagine most of the time is not spent on the comments (although some of them are quite funny and probably do require some thought), but rather on searching through the candidates' edit histories. That should occur regardless of which system we use. -- Black Falcon(Talk) 20:26, 9 May 2007 (UTC)[reply]
Success ratio of RfA's in 2006 in terms of week opened.Selection bias, you can see the variability in this chart. Also, the high ratio of currently successful candidates may simply mean failed ones are getting SNOWed earlier. —dgiestc 15:27, 9 May 2007 (UTC)[reply]
The glass is half full, people :) >Radiant< 15:39, 9 May 2007 (UTC)[reply]
I'm more of a "the glass is at 50% capacity" person myself. Besides, it's much more fun to just be an ass. EVula// talk // ☯ // 16:11, 9 May 2007 (UTC)[reply]
Lets get the ball rolling here! Ryan, you've done a bang-up job, mate! How's about tagging this as a policy so people can actually start using this as a means by which to secure the few extra buttons that the sysop flag affords? gaillimhConas tá tú? 20:49, 9 May 2007 (UTC)[reply]
Please note that people should feel free not to use PROA - that is, we're not forcing this on the community, but rather making available a much easier process gaillimhConas tá tú? 20:51, 9 May 2007 (UTC)[reply]
You've still not dealt with numerous concerns above, though. Has this been posted to Centralized Discussion or the Village Pump? -- nae'blis 21:02, 9 May 2007 (UTC)[reply]
I think it's premature to start "tagging this as a policy", considering that at least half of the people who've commented on it are opposed to the idea or certain parts of its implemention, and that many of the issues raised have not been addressed. -- Black Falcon(Talk) 21:22, 9 May 2007 (UTC)[reply]
I still don't like it. A prodded article can be restored if a single user objects after the fact. This is to account for the fact that someone might not notice a prod until it is too late. The slower discussion-oriented processes are not overrulable by a single after-the-fact objection because by virtue of being a public discussion, they represent consensus. If someone gets sysopped simply because nobody noticed until it's too late, then it stands to reason they should be desysoppable if anyone objects later. It's only fair, right? But then that completely nerfs them as an admin because if they do anything controversial they get desysopped. —dgiestc 21:05, 9 May 2007 (UTC)[reply]
To nae'blis, haha I may have jumped the gun a bit! I'm just enthusiastic to finally see a de-emphasis placed on a few extra buttons. I'll get it over to the Pump as soon as I'm done with this post. To dgies, last I heard, it takes about four minutes to de-sysop someone if he/she is acting out of sorts (though I would like to put forward a simple way to desysop someone as soon as we get this up and running). Also, I think you're getting a bit stuck on the whole PROD thing. While this is inspired by the PROD system, it is its own unique entity. Also, to say nobody will notice is a bit silly, as I'm sure the RfA regulars will watchlist the category or page where a bot lists the PROA people. gaillimhConas tá tú? 21:14, 9 May 2007 (UTC)[reply]
Prod with deletion is okay, because it is easy to reverse a prod, if someone complains after the fact the prod is reversed and a proper AfD may ensue, we cannot be deadmining prodded admins after the fact then doing an RfA. We don't need two classes of admins, ones who have demonstrated they have community trust, and those who have simply demonstrate a lack of community distrust. Not the same thing. HighInBC(Need help? Ask me) 21:18, 9 May 2007 (UTC)[reply]
I don't think that you're relating the "trust" and "lack of distrust" appropriately. As mentioned, this will be highly visible and there will be (or rather, I'd like to see) a subpage where people can voice their support in case the PROA is rejected or opposed and goes to RfA (this subpage will then be the RfA page) gaillimhConas tá tú? 21:22, 9 May 2007 (UTC)[reply]
If the category is filled with lots of usernames, individual PROAs will not be very visible. -- Black Falcon(Talk) 21:43, 9 May 2007 (UTC)[reply]
That does not resolve the fact that the lack of evidence of distrust is not the same as evidence of trust. This test creates a demonstrate of a lack of distrust, whereas our criteria is a demonstration of trust. HighInBC(Need help? Ask me) 21:40, 9 May 2007 (UTC)[reply]
I understand where you're coming from, but as mentioned, people can voice their support on the subpage or reject the PROA if the candidate is deemed unsuitable by the 'rejector'. As mentioned several times, this will be highly visible where all the RfA regulars can see it. The PROA policy does not create a need for a lack of distrust, but a collective approval of everyone who sees it, which should be quite a few people. gaillimhConas tá tú? 21:45, 9 May 2007 (UTC)[reply]
Why couldn't we just transclude, or otherwise maintain, a category or list of admins who are currently proposed for adminship, and have it appear in full on the regular WP:RFA page? That way, anyone who would normally be voting in an RfA will still have the opportunity, in the very same place, to look through all those candidates, and anyone who would normally oppose them can simply go object to the proposal. Would this remove concerns about "not having the trust of the community" or about this being a "stealth" process? SeraphimbladeTalk to me 21:26, 9 May 2007 (UTC)[reply]
I think that's a great idea! The suggestion of "sneaking people through", so to speak, has been brought up by fellows who, in good faith, have apparently missed what this is all about. It's about making things easier. It (the list of PROA people) should be highly visible through a category that is updated by a bot and a page which can be watchlisted to show everyone in the category. WP:RFA is a great place to display these names, or at least a place where a prominent link to a subpage could be displayed (so as not to clutter the main RfA page). gaillimhConas tá tú? 21:39, 9 May 2007 (UTC)[reply]
But you left an oppose vote for a candidate today stating that "I'm rather underwhelmed by the responses to the questions, which demonstrate that he might not actually know how to use the buttons, which is a bit of a concern". Doesn't this indicate that it's important to have potential candidates give a statement and communicate with the voters as a display of that kind of knowledge? Isn't that safeguard being eliminated here? Dekimasuよ! 00:20, 10 May 2007 (UTC)[reply]
People have talk pages. If you want to hear what the candidate has to say on something, you visit their talk, you say "Hey, related to your proposed adminship, I'd like to know what you think on...". If they refuse to answer, or you don't find the answer satisfactory, it'd still be your option to dispute the proposal based on that. SeraphimbladeTalk to me 03:38, 10 May 2007 (UTC)[reply]
Brilliant... in fact I suggested something similar below, but having the current RfA able to use the proposed option after the process is started. LessHeard vanU 22:45, 9 May 2007 (UTC)[reply]
This worthy proposal falls foul of one constant; someone, somewhere at sometime will oppose. At RfA there can be a vote of 50-0-1 (with a well founded and argued opinion against) and a rightly succesful application. With this method it doesn't matter how many eyes/ayes see it, that one nay is all that is required to end it. Even with a 51-0-0 result there has been the case of an oppose (or neutral tending toward oppose) that has been persuaded to change their mind (and strike out the original vote) by discussion or by the proposee answering their concerns.
On that last point, a neutral at RfA may be inclined to oppose here as that is the only available option to staying silent. Rather than easing any burden at RfA this proposed system will likely create very few successful candidates, and (as previously argued) add another layer to the process.
Um, whilst writing this the thought occurred to me that this could be an option at RfA if an application is sailing through and a consensus develops that there will be no oppose vote, and with a drastically reduced time scale (48hours?), with a note at the RfA. If the PROA is opposed then it reverts back to the RfA where the reason for opposing is turned into a vote. The full process continues. If the proponents of this suggestion want to tweak this then please do.
Probability of an RfA succeeding in terms of age and support percentage Have you seen this chart? It essentially confirms what you proposed (though for about 54, not 48 hours). The only caveat I'd add is that since that date was analyzed, there was a recent RfA wich went unopposed until about 6½ days when some misconduct surfaced, and the bureaucrat kept it open longer while the opposers piled on. —dgiestc 23:09, 9 May 2007 (UTC)[reply]
Thanks. Yes, interesting. It doesn't seem that my suggestion has been taken up, though. As for the recent example of a late change of heart I would comment that i) no system will ever be foolproof (because there will always be a bigger fool?), and ii) you have to ask what it was that took so long to uncover, how recent it was, and if it really had relevance to the application before them. Using the above system the proposed Admin might have got the mop&bucket; any later evidence of wrongdoing would have meant that it would have had to be reviewed. If it was serious then the appropriate actions would need taking (including desysoping), if not so then perhaps the appointment would have been unchallenged. I can really see a benefit to using this system, as a process within RfA (but not separately and certainly not instead of). LessHeard vanU 20:03, 11 May 2007 (UTC)[reply]
I'm a non admin willing to see how many hours (minutes hopefully not) it would take for a no, I'll report on the real why for whatever person it is jbolden1517Talk 02:44, 10 May 2007 (UTC)[reply]
Post an app at WP:RFA in the conventional way, but replace all the questions with a note explaining it is a trial of this process, then strike the support and neutral sections. Dgies 03:01, 10 May 2007 (UTC)[reply]
i don't want a failed RFA that's permanent and searchable. Remember the fact that good people won't do it is the reason this all started. jbolden1517Talk 03:27, 10 May 2007 (UTC)[reply]
Do it without a subpage. We can't have a trial without it being posted somewhere pretty public. Dgies 03:33, 10 May 2007 (UTC)[reply]
Without a lot more discussion, any kind of trial would only prompt a lot of opposition. Let's talk this through more. --Tony Sidaway 09:07, 10 May 2007 (UTC)[reply]
OK results are in:
30 minutes
some guy I don't know who raised the point that if you don't like public criticism (from the RFA process) you won't like being an admin (As an aside I tend to agree)
Tone was much less hostile then what's usually found in an RFA oppose. I suppose since the issue is now dead.
One thing bothers me about this process more than any toher, because I fear it could be a potentially fatal flaw...
...Some vandals are clever. Some add information to articles that is incorrect, or even irrelevant, but isn't always easily spotted, for example. On the other hand some people are vindictive and/or spiteful (it does happen!) What guards are there against a user removing the tag with a spurious reason, or one that may seem reasonable on the surface but is in fact anything but reasonable? Any editor that's been here for a while will have at some point encountered at least one or two (probably numerous) situations where they have rightly disagreed with somebody only to have that person hold bad feelings against them (you get the rough idea). Is any old reason to count as valid enough for the removal of the PA tag, no-matter how petty or illegitimate?
My apologis if this has been mentioned before - I've just scan-read the talk page and didn't see it if it has been. Crimsone 03:22, 10 May 2007 (UTC)[reply]
What's it matter, really? If one person holding a grudge removes the proposal, they can still run through a regular RfA. If that's the only person who has any reason to oppose them, they'll still fly right through. Seraphimblade 03:33, 10 May 2007 (UTC)[reply]
The way proposed deletion works, almost any reason for removal is accepted. Like proposed deletion, this is intended to be a painless method that reduces the effort spent by the community on uncontroversial decisions. A proposed administrator candidate who attracts controversy is likely to have his tag removed, and probably would be better off going straight to requested adminship. This proposed policy isn't for every candidate. --Tony Sidaway 09:11, 10 May 2007 (UTC)[reply]
How many RFA's pass completely unchallenged without a single oppose? Of course, it's worth mentioning here that I'm not talking about candidates that attract controversy - I'm talking about candidates/nominators finding the PA tag taken down by vandals or by those who hold grudges for no good reason (personally I've been involved in a fairly intense informal mediation where at various points, both mediated parties were upset at me for not taking their side and admonishing the other absolutely for example. Or maybe it could be someone upset at the fact that you nominated a page they created for deletion).
It seems to me that there aren't too many completely uncontested candidates, and of those that exist, a number (undefined!) may be brought out of this process either by vandalism or by petty revenge (or such things) where it isn't actually warranted. It is of course no reason not to trial the process, but I would be interested to see how such issues might be dealt with if they become a problem. I'm also concerned that a failed Proposed Adminship may end up counting against the candidate at RfA as yet another one of those criterion that has nothing to do with use of the tools themselves (though I must admit, having been away for a while, I don't know what the current RfA climate is with any certainty - I'm going on what I saw of it the last time I looked).Crimsone 13:31, 10 May 2007 (UTC)[reply]
The reason for a failed proposed adminship would obviously count. If it's a poor reason then it's not a good reason not to promote in a subsequent normal RFA. I agree that few RFA's go through without a single oppose, but rather suspect that it would not apply with this method. The whole torches-and-pitchforks atmosphere simply wouldn't be there. -Tony Sidaway 13:40, 10 May 2007 (UTC)[reply]
I would agree as well, and just like a prod, in the case of, say, pure vandalism (a vandal-only account blanks the page that happens to have the proposal on it, or someone you warned for vandalism comes and removes it with the reason "YOU SUCK!"), I don't think anyone in his or her right mind would object to re-adding it and considering that objection invalid. In the case of a more realistic but rather silly objection, well, I suppose people are allowed to oppose for silly and stupid reasons, so if someone wants to object on the basis that they don't like a nomination for deletion you made, I suppose they can. SeraphimbladeTalk to me 14:01, 10 May 2007 (UTC)[reply]
While I agree that the given reason behind a failed proposed adminship should be considered at RfA, I'm concerned that there may be a potential for the very fact of a proposed adminship failure itself may eventually be taken as a cause for concern for those considering RfA candidates. Crimsone 15:13, 10 May 2007 (UTC)[reply]
If this is to be like PROD, we can't give a meaningful definition for what is a valid reason or not. Except for obvious vandalism, any objection must be taken for valid, and the other process (RFA/AFD) should take over. >Radiant< 14:11, 10 May 2007 (UTC)[reply]
Perhaps the process could be like PROD, but with an exception list (for example, an exception may be that the reasoned nomination of a single page for XfD is not a valid reason for refusing adminship (because let's face it, it's not! lol - a string of relatively recent innapropriate nominations however would be a reason to be quite cautious). Such an exception list would contain a number of such silly PA-tag removal reasons (including vandalism), each of which would have to be discussed. Such a list would be specific (ie, not a rough guide to what may not be an acceptable reason), and would contain no comment on potentially contentious issues (edit count/summary for example) - it would merely contain specific exceptions to the "PROD rules", and I doubt they would be great in number. The exact content of such a list would need extensive discussion though. An alternative may be to allow beaurocrats to review given reasons on request with a view to potentially restoring the PA tag for a continuation or re-start of the process - for which some simple guidelines would be needed (they wouldn't be beaurocrats unless they had common sense, and the aim would be that of protecting the process from silly or entirely unfounded objections so that the process can do what it suggests, and promote uncontraversial candidates without expending the communities time on doing so. These are of course just a fe ideas that may or may not be agreeable at this time, or perhaps instead for the future, should the process stumble heavily on such problems (or indeed, perhaps never! lol) There's no harm in ideas :) Crimsone 15:13, 10 May 2007 (UTC)[reply]
I don't think a list is important or useful. If we get consensus for this method, then someone makes a nuisance of himself by removing the tag from lots and lots of proposed adminships for frivolous reasons, then that's a matter for dispute resolution, up to arbitration if necessary. --Tony Sidaway 15:40, 10 May 2007 (UTC)[reply]
This is perhaps a perennial suggestion, but compiling a list of "valid" objections only encourages people that have an objection not on that list to lie. It doesn't actually solve anything, and has undesirable side effects, ergo WP:CREEP. >Radiant< 15:44, 10 May 2007 (UTC)[reply]
I see that point Radiant, and on the subject of the list idea, I would be tempted to agree. On the subject of beaurocrat review though (in the sense that I suggested it), I've just realised that there wouldn't need to be guidelines at all. The beaurocrats would of course understand thier role in that situation - protecting the process from silly or entirely unfounded objections so that the process can do what it suggests, and promote uncontraversial candidates without expending the communities time on doing so. Not everybody has common sense of course, and those that do have different ideas about what it is, but I'm fairly confident that beaurocrats could quickly exercise it in the interests of the community in these terms. I'm actually less concerned about a single user being disruptive, and more concerned about the process not being as effective as it could/should be (or even failing) for lack of adequate protection. Essentially, if the reason is openly silly (a single reasoned XfD nomination that somebody didn't like, for example - it's an easy example! lol) then it's easy to see that - If it's the sort of reason that would take much time in investigation then it's potentially a contraversial or contentious one. I doubt such a role would take a beaurocrat long at all. The difficult part of the suggestion would be in making it clear as to what sorts of reasons for a PA-tag removal might prompt a request for review, which could be the undoing of the idea if nothing else - in an ideal world, only reasons which are silly and uncontraversial (using basic common sense) should ever need to be reviewed, but in practice there's a question of how to ensure only such reviews are requested.
It's a shame that basic common sense is more often than not impossible to codify. lol I do like the idea for this PA process, though I personally do feel that it's potentially a little flawed in terms of it's stated purpose. Perhaps it could run as it is, though only a trial run would show whether or not it would need protection. I'd be happy to be wrong, because apart from this I'm out of ideas. lol Crimsone 16:18, 10 May 2007 (UTC)[reply]
If the purpose of this proposed process is to reduce the amount of bureaucracy involved in admin promotions, then creating a list of silly and 'good' reasons to oppose and establishing a review process for that (and possibly an appeals process for reviews) probably defeats the purpose. Removal of the tag by someone on a spree can be reverted as vandalism, but a list of 'good' and 'bad' reasons is not feasible when it comes to the basic issue of trust. Although I'm sure everyone will agree that "a single reasoned XfD nomination that somebody didn't like" is not a valid reason, people will have different standards of what constitutes a 'reasoned' nomination. I think protection from vandalism would primarily come from the relative obscurity of the process except to RfA regulars. -- Black Falcon(Talk) 18:02, 10 May 2007 (UTC)[reply]
I've already agreed that the list idea probably wasn't a very good one, both in the opening sentence of the comment to which you replied, and in a comment further down the page (that's the trouble with replying out of sequence). The "appeals process for for reviews" as you have called it is actually intended as less of an open appeals process as such a name would suggest, and more a means by which obviously invalid reasons could be prevented from derailing what should otherwise be an efficient process designed to fastrack uncontraversial nominations. In my view, as things currently stand, while everyone would indeed agree that a "single reasoned XfD nomination that somebody didn't like" isn't a valid reason, it would nevertheless be reason enough to derail the process - which would defeat the object. A simple excercise of common sense by a 'crat would solve such a problem. Common sense in this respect would also mean that if the one single reason given has the potential be contentious/debateable with regards to sysopping, it should not be overturned and should go to RFA if desired. If the reason is obviously invalid, it should be overturned. It really would be that simple, allows this process to be as efficient as it should be in removing said beaurocracy, and should waste far less time and effort than forcing said unlucky and uncontraversial candidates through RFA (or such is the idea) I'm going back to the bottom of the page now, where I'd probably rather have stayed. I hate going out of sequence - it's confusing for people to follow. lol Crimsone 18:45, 10 May 2007 (UTC)[reply]
But, the problem is that "common sense" isn't ... common. People have different understandings of what constitutes a common sense or a silly reason to oppose. A "single reasoned XfD nomination that somebody didn't like" can derail a proposed adminship because there is no fixed definition of what constitutes a 'reasoned nomination'. The only reason that is "obviously invalid" is blatant vandalism. Having 'crats apply "common sense" beyond that amounts to having them apply their own standards for admin promotions. I'll post this here for consistency; reply at the bottom of the page if you wish. Cheers, Black Falcon(Talk) 18:52, 10 May 2007 (UTC)[reply]
I'm with Radiant! on this one; either people will lie to get a reason on the list (and we can't even codify what we think are valid reasons at RFA), or any and all non-vandalism reasons must be accepted. We're also talking about a significant expansion in b'crat duties, since they are now responsible for being gatekeepers, and anyone who closes a Proposed Adminship request must satisfy themselves that the candidate is worthy as well before flipping the bit. Promoting on 50/1/2 is much simpler for them; has anyone asked them if they're willing to take on that task? -- nae'blis 16:40, 10 May 2007 (UTC)[reply]
I'd be against any proposed "list of exceptions", aside from blatant and obvious vandalism. Just like a prod, this should only reflect unanimous consensus. If someone has concerns, no matter how silly or stupid anyone may think they are, those concerns should be addressed in a full RfA. If they really are silly, no one else in the RfA will take them seriously, and the candidate will pass right on through anyway. SeraphimbladeTalk to me 16:56, 10 May 2007 (UTC)[reply]
I think you misunderstand the list idea, Nae'blis - people would lie to get a reason that isn't on the list (but it's still a bad idea for the same reasons. lol) It wasn't a list of reasons (or types of reason) for which a PA nom would fail, but a list of reasons (or types of reason) which should not cause a PA nom to fail (such as a single reasoned nomination of a page for XfD).
As for the 'crat review idea, it should be a quite small "duty" if done right, and is an entirely seperate idea from the list idea which relying instead on the common sense of the 'crats within the constraints of merely avoiding silliness and ensuring that uncontraversial candidates aren't halted in this process by reasoning that no reasonable person would consider a valid objection (back to the single reasoned XfD nomination thing again!). As for 'crats being satisfied that a candidate is worthy before flipping the bit, if nobody has objected within 7 days, it seems fairly likely that they are - one thing wiki is famous for is objections! It shouldn't take a 'crat too long in most cases to decide whether he/she has any objections or not. What might help however, is not only having criteria for who can object, but also basic criteria for who can be nominated (1 year on wiki, 1,500 edits in any namespace, etc) - on the other hand, that again comes mainly down to common sense, and if a person is likely to fall short of such criteria, you could almost guarentee that somebody would object. If the point of this idea is to prevent the waste of the communities time on RFA'ing uncontraversial candidates, it may not be very efficient if silly reasons for PA tag removal are going to force uncontraversial candidates through RFA anyway. To me, that would defeat the object. Again, mainly thoughts, for what they may be worth. Crimsone 17:04, 10 May 2007 (UTC)[reply]
reply to Black Falcon - Regardless of what constitutes a "reasoned nomination" all it needs is a reason considered in good faith. A single nomination to XfD in itself is no valid reason to refuse adminship, and if it's been done with a good faith reason it's even less valid (hey, even current admins make minor mistakes, and that's what XfD is there for - concensus building on the subject of page deletion). This is what I mean by common sense - it's fairly obvious that in such a case, this reason would not be a valid reason to oppose an admin candidate, and I'm absolutely sure that a consensus would agree (as you alluded to in your earlier post). However, if somebody wished to say "candidate demonstrates a pattern of poor reasoning behind XfD nominations" (or similar) and wished to cite one as an example, that may well be a valid issue for consideration, and so an RfA should be required. "Common sense" would require 'crats to consider whether reasoned concensus (remembering that it's the reasoning that should matter, not the vote) would exist to support overturning a proposed adminship, or whether doing so would be contentious. They've enough RFA closing experience to be well placed to do that, and if they feel that a reason for a failed PA is quite clearly invalid in itself, they could continue or restart the process (a very simple matter of replacing the tag and leaving a note). If there truly is a reason that a candidate is contraversial, someone else will object, else the original user may return and offer a reason worth serious consideration on its own merit. Again, I state that my personal definition of an invalid objection reason is one where a concensus would agree that the reason, on its own merit, would not be a cause for concern/caution, and would not be a contraversial issue in terms of its effect on an RfA. There is undoubtedly a better definition out there somewhere.
Perhaps a better alternative yet is to require objectors to be specific and detailed in their objection (at least a few coherent lines), possibly even offering diffs. On its own this should discourage invalid/silly objections, and it would make most such objections stand out a mile. Crimsone 19:41, 10 May 2007 (UTC)[reply]
I think requiring a relatively detailed objection (a few lines) in order to remove a PROA tag is a good idea, but I still don't think that judging whether an objection is valid (except in cases of obvious vandalism) is a viable strategy. If someone holds a grudge against an admin candidate, he or she won't come out and state "Oppose. Nominated for deletion an article that I like.", but rather "Oppose. Made a frivolous/poorly researched/bad faith nomination." The latter cannot simply be dismissed as an invalid objection as whether an XfD nomination is "frivolous", "poorly researched", or "bad faith" is largely a personal/subjective judgment.
Also, the stated purpose of PROA is to cover only completely uncontroversial candidacies: overturning a proposed adminship should not require a "reasoned consensus", as you state above; a single objection suffices to end the process. If PROA is indeed to be for uncontroversial cases only, then we shouldn't exclude any objections. Either a candidacy is uncontroversial or it's not. If PROA is to be implemented (which I oppose for various reasons stated above by myself and HighInBC), it should be implemented without extensive restrictions. -- Black Falcon(Talk) 20:38, 10 May 2007 (UTC)[reply]
I agree with both of you. To avoid any further complicated processes while maintaining fidelity to the PROA intentions, a "good reason" is all that is needed to remove the template. We should assume good faith that everyone knows what a good reason is, as well as respect that people have varying perceptions of what this entails. To this end, I completely agree that one should note why he or she removed the template on the fellow's talk page or on the RfA subpage, much like one does when rejecting a good article candidate (I agree that few lines with a bit of rationale and constructive criticism should suffice). gaillimhConas tá tú? 22:54, 10 May 2007 (UTC)[reply]
It has been repeatedly suggested that this proposal "de-emphasizes adminship". I still don't really understand how that's the case. Can someone explain it to me in concrete terms? Here's what I see:
1) It does not reduce load on RfA: If working properly, as many people should look at these nominations as any other nomination. (If not working properly, the candidate is not being adequately reviewed, which is a very bad thing.) In fact, it increases load: users will have to review the PROA, and then vote again on the RfA of any that fail. This fact is recognized by both supporters and opposers of the proposal.
2) It does not remove pile-on support: In response to objections by HighInBC and others that this process precludes evidence of trust in the user and evidence of active support for the candidacy, it was suggested that users note their support for the template. Meanwhile, the qualms of users who noted that PRODs may be undeleted on request after the fact, but that admins may not be desysopped in the same way, have gone unanswered.
3) It does not remove the questions for the candidate: In response to objections that it removes the candidate's statement and answers to the standard questions from the process, which are often seen as important guides to the candidate's understanding and interpretation of policy, it was suggested that the questions be asked on the candidate's talk page.
4) It does not make requesting adminship more friendly: If the candidate's discussion at RfA was going to be unanimous, there was never going to be anything but friendliness. If the template is removed by a single user, it can cause aninimity between users. Alternatively, a user might withhold a valid objection in order to avoid such problems. This was mentioned in the first thread on this page, and never really addressed.
5) It does not increase the probability of a successful adminship request: Any users that could become admins through this process would also pass RfA.
In light of these things, I fail to see how this is a simpler, easier, or kinder process than RfA. What specific problem is this designed to address? Dekimasuよ! 06:46, 11 May 2007 (UTC)[reply]
Answers
1) This is during the test run only. We check the PROA result using RFA. This is exactly how you should run a test. If/When folks feel confident, RFA can be dropped, and people could actually nominate candidates in this way.
2) Supports are not generally required.
3) I thought we'd finally be rid of those darned standard questions.
4) I'm not sure that's a big issue. We'll find out if it is if we give things a whirl. (It's taking quite a while to give things a whirl though, eh? Much talk, little do!)
5) Um, Yes, exactly :-) . That's the whole idea. The only big difference should be that this method is supposed to scale rather better. :-)
What we're actually trying to do is solve scaling issues. As the most recent example in the case of RFA, there was a peak of RFA activity that couldn't quite be handled by the software! So... if you can think up other ways to deal with RFA scaling, that would be as good or better. :-)
1) That is not what Dekimasu said; he was talking about any that receive an objection. It is impossible to count how many users will have reviewed that PROA up to that point, who will then have to go participate in RFA anyway to register their trust, rather than leaving it tacit as in this system.
5) That's a little disingenuous. What actually happened was a bloated template hit a performance limitation, which could have happened on any list page anywhere in the software. Trimming the template once doubled or tripled the capacity of RFA, and further reductions in overhead are very possible. Some of us feel that a decentralized system would actually scale worse than RFA. -- nae'blis 13:19, 11 May 2007 (UTC)[reply]
Kim, this is about RfA reform, not scaling issues. And, as far as I know, we're not talking about doing away with RfA completely; the whole reason this has even marginal support is that when the tag is removed we still have the real system in place. This would have next to no support at all if it was being proposed as the end-all of adminship requests. KafzielTalk 13:26, 11 May 2007 (UTC)[reply]
I think Kim is broadly right when he talks of scaling issue. Simply put, we're not really getting adminship proposals into the existing process at a decent rate, because it's unnecessarily bureaucratic. This proposal is much more lightweight because in the reducing case it's just a matter of sticking a tag on a page and waiting for seven days. If nobody objects nothing further happens until a bureaucrat reviews the application and decides whether to promote.
By contrast, our existing process has all kinds of encrusted nonsense: a ceremonial gauntlet to run, questions to answer, and quite a lot of energy expended on something that isn't supposed to be a big deal.
I hope we'll get more candidates this way. The poor candidates will fail at the first hurdle when someone removes the tag. The good candidates will become administrators without all the fuss we have now. --Tony Sidaway 13:35, 11 May 2007 (UTC)[reply]
What about the issues in point #2, a prodded article can be "oops undeleted", and admin cannot be "oops desysoped". And the fact that the demonstration of the lack of distrust does not equal a demonstration of trust has also not been addressed. It seems these points are being avoiding like land mines. I think this will result in a lack of confidence in the community trust in admins promoted this way, as the admin never demonstrated community trust, never answered any of the questions, does not have a page of users showing support for them. Lack of objection is not enough, I want admins to have support as well.
I still don't think this is going to solve any of the problems with RfA which I am still not even sure exist. HighInBC 13:53, 11 May 2007 (UTC)[reply]
As I think someone has already observed, most if not all of our problem administrators who have had to be desysopped had had very successful requests for adminship. The current process is not a good measure of trustworthiness.
Trust here has two components: trust in an editor's good faith, and trust in his competence. Lack of community distrust in the good faith of an editor with a reasonable amount of experience implies community trust, because we operate a policy of assume good faith leavened by commonsense based on observation. If you doubt a candidate's good faith, say so and remove the tag and no questions asked.
The other component, competence, is easily demonstrated: if you're not confident that a candidate is experienced enough to do a good job as administrator, or you feel that his judgement is not to be trusted, just say so and remove his tag and no questions asked. Thus a candidate who makes it through this process is so trusted by the community that not one person has serious doubts about his good faith nor his competence. --Tony Sidaway 14:15, 11 May 2007 (UTC)[reply]
Which will result in far fewer admins. Even those who pass RfA with 99% support would fail PROA. I'm pretty sure the people calling for RfA reform are looking for more admins, not less. KafzielTalk 14:27, 11 May 2007 (UTC)[reply]
You could be right, but I don't think it will work out that way. I propose that we give it a trial. If you're right we will have learned something at little cost.
As I remarked earlier, I think the problem is that while we have thousands of people who would make very good administrators, very few of them actually apply. My diagnosis of this is that Requests for adminship may well appear excessively bureaucratic to good admin candidates. Popping a tag on your page and waiting for a week, perhaps answering a few questions on your talk page, is much less bureaucratic, while retaining very high standards as you have suggested, so as to attract only the best candidates. I think it's worth a try, and it has the advantage that it would run alongside our existing process. --Tony Sidaway 14:41, 11 May 2007 (UTC)[reply]
What do you mean by "trial"? The results will not be actionable unless there is a very large community consensus to change how we promote admins. If by trial you mean a simulation, then I cannot see the harm, but I don't think the 'crats should be promoting anyone off any trial until there is a site wide consensus. This change effects the entirety of the Wikipedia community. HighInBC(Need help? Ask me) 14:47, 11 May 2007 (UTC)[reply]
If filling out an RfA page and answering some questions is too much bureaucracy for someone to handle, I don't think I want him to be an administrator. KafzielTalk 14:43, 11 May 2007 (UTC)[reply]
I agree, it is not a gauntlet, it is a simple procedure. I did it, twice, I was not ready the first time and I failed, I was ready the second and I passed. It worked great, and neither times did I find it bitey or difficult in any way. HighInBC(Need help? Ask me) 14:48, 11 May 2007 (UTC)[reply]
The results of the trial are quite actionable. The result now is that a "bureaucrat" posts on someones talk page: "Conratulations-ish: If this had been a real process, you would now be promoted to admin. Now try RFA, and compare to see how you do there!" --Kim Bruning 15:25, 11 May 2007 (UTC)[reply]
Ok, now outlined the experimental procedure on the project page. --Kim Bruning 15:36, 11 May 2007 (UTC)[reply]
Do 10 names under the simulation, then do the same 10 people with an actual RfA. I bet you that people who rightfully fail their RfA will have passed through the proposed admin net. The reason for this is that the added step required in look through another list of candidates will reduce the examination a user normally undergoes. HighInBC(Need help? Ask me) 16:03, 11 May 2007 (UTC)[reply]
Note that the reason for RFA reform (and reform of several other parts of the project) is ultimately because people see something wrong with RFA. We think that the problem is mostly to do with scaling issues. Ok, there's just one way to find out ;-) (remove scaling and see if that's sufficient to fix it then). I'm not sure how a decentralized system could ever scale worse than a centralized system. In a consensus based system, measurements show that decentralization is a prerequisite for scalability.
If people can come up with a better decentralized or otherwise hoped-to-be-scalable design, I'd really love to hear about it :-)
--Kim Bruning 15:36, 11 May 2007 (UTC)[reply]
I'd take issue with the idea that the problem of RFA is scalability (we don't actually know what the problem is yet, so throwing solutions at it is unlikely to work). In addition, one thing that RFA does well is show trendlines and early returns, which actually helps scalability: if 40 people have vetted a candidate independently and marked down their support, I trust the community enough to not go do my own analysis, in most cases. Under this decentralized system, there is no way for me to know that that explicit trust exists, so I must in good conscience evaluate everyone. That's why the conflation of 'scalability' with 'tacit/explicit trust' in this discussion is so perilous. I'm curious if you can come up with a decentralized system that preserves explicit trust well, which seems to be the dealbreaker for many people here. -- nae'blis 15:52, 11 May 2007 (UTC)[reply]
Excellent point. I was thinking the same thing myself. This is what prompted the reply that supports could still be recorded, which led to point 2 in my original comment in this thread. Dekimasuよ! 05:09, 12 May 2007 (UTC)[reply]
I agree, I have a lot of trust in community judgment. I can look at support and opposes, read their reasoning, and this saves me a lot of time in my own research. If I want to look through the RfA candidates under this system I must look at RfA, then go through a list in the proa cat. If I want to know the answers to the questions I have to go to every user talk page, and if not already asked I must ask. The only advantage I see is that the potential admin does not need to fill out a form.
It saves a little work for the applicant, and puts all that work, plus overhead, onto those reviewing the applicants. HighInBC(Need help? Ask me) 16:06, 11 May 2007 (UTC)[reply]
And yet the wiki encourages systems involving thorough review and lack of opposition. It would be interesting to see how many people converge on a single candidate, to see if the scaling issues you fear show up. I do agree that we seem to lack a form of mark-as-patrolled. (And not just for PROA) :-). Durin, Greg Maxwell and Kelly Martin have done some number crunching for me, and we're all pretty darn sure that scalability is the main issue. --Kim Bruning 16:47, 11 May 2007 (UTC) some of these folks might not be too community minded, but they sure can crunch numbers, Ayup![reply]
Absolutely; that's what I meant when I said RFA already scales better than this will, in some ways. We don't expect every editor to review every diff on every article on their watchlist; we assume that they have reviewed the article during their last edit session, and go forward from there. RFA and other promotion processes are different because while edits to an article generally only affect that article, the sysop bit potentially affects every part of the wiki from there on forward, hence the centralized vetting process. If I see 7-8 users whose opinions I trust and generally agree with opine support on an RFA, I consider it 'patrolled' and move on. I have a theory for how we might be able to combine explicit support with decentralization, but a) it's not fully formed in my head yet, and b) it's not really related to this proposal anymore. -- nae'blis 17:39, 11 May 2007 (UTC)[reply]
I believe it's pretty much been established on this talk page that sufficient editors object to this process that any "proposed admin" candidates will be denied by at least one of such editors, thus forcing them to use the regular RFA process once more. This is a far cry from the initial reactions to PROD. Radiant! 15:04, 11 May 2007 (UTC)[reply]
Yes, I think we'd need a lot more support to go ahead with this. I'd like to see those who are skeptical about it agree to a trial. If that isn't feasible then it's unlikely ever to work. --Tony Sidaway 15:41, 11 May 2007 (UTC)[reply]
Most of us are not just skeptical; we are against it. It has nothing to do with consensus and does nothing but add a bureaucratic step to what we already have, without reducing any of the red tape at all. There's no need to try it because even if it works, we don't like it. KafzielTalk 15:43, 11 May 2007 (UTC)[reply]
I've removed the RFA back end, which sort of kills a lot of red tape. I have no idea who added it, but that sort of defeats the object, eh? --Kim Bruning 16:02, 11 May 2007 (UTC)[reply]
Falling back to established process when the simplified process fails is common practice on Wikipedia, and any system which can get caught in a recursive loop is bad process. -- nae'blis 16:05, 11 May 2007 (UTC)[reply]
I'm not quite as far out there as Radiant in declaring this finished; I think it's unworkable and will not achieve the goals originally intended, but I do not intend to remove the tag as a protest statement from otherwise good candidates. -- nae'blis 15:44, 11 May 2007 (UTC)[reply]
I see the use of this tag as an affirmation that the candidate either fears criticism or dislikes the consensus-building process, both of which are an every-day part of being an admin. In light of Kim's recent changes to the proposal I think I should also say I'm not going to discuss tag removals with any candidate. I will leave my reason, and it is not open to endless debate. Call me pointy if you like, but I say forcing this proposal into a test phase despite a sea of objections is pointy in and of itself. KafzielTalk 16:03, 11 May 2007 (UTC)[reply]
The conclusion that "use of this tag as an affirmation that the candidate either fears criticism or dislikes the consensus-building process" doesn't seem compatible with our Assume good faith policy to me. --Tony Sidaway 16:54, 11 May 2007 (UTC)[reply]
I would cetainly plan to inspect any candidated tagged for this process, adn i would probably either nominate the canddiate at RfA if I thought s/he would mage a good admin, or else remove the tag. DES(talk) 16:58, 11 May 2007 (UTC)[reply]
I don't have to "assume" anything, Tony: if the user placed the tag on his page, he has demonstrated his reluctance to go through even a tiny taste of the criticism and red tape he will face as an admin. I don't have to debate or defend my reasons, because there's no "list of valid oppose reasons". I've been saying that since this proposal got started and, if you recall, I knew people would start to get defensive and start throwing AGF and POINT around when the rejections actually started. You assured me that wouldn't happen; what changed? Does my opinion on this subject suddenly make my behavior "egregiously bad"? KafzielTalk 17:19, 11 May 2007 (UTC)[reply]
Lets hope it does not get backlogged. HighInBC(Need help? Ask me) 17:05, 11 May 2007 (UTC)[reply]
This system SHOULDRFC 2119, be self maintaining, and if it does, it MUST NOT RFC 2119 fall back to RFA in case of doubt. One way to do that is to simply apply Bold revert discuss. I'm open to other ideas, of course. :-) --Kim Bruning 15:59, 11 May 2007 (UTC)[reply]
It doesn't now... (borrowing one of Kim's trademark innocent looks) >Radiant< 16:02, 11 May 2007 (UTC)[reply]
Infinite court of appeals is a dealbreaker for me. -- nae'blis 16:04, 11 May 2007 (UTC)[reply]
Fine by me. Replace it with something more useful? --Kim Bruning 16:17, 11 May 2007 (UTC)[reply]
I agree with the others - as I said above, I will not debate my decisions. If a candidate wants a chance to debate, he can use RfA. And, again, nobody (except Kim) is suggesting this actually replace RfA, any more than PROD replaced AFD. KafzielTalk 16:07, 11 May 2007 (UTC)[reply]
You SHOULDRFC 2119 debate your decisions, this is a requirement, because this is a wiki, and wikis operate on the basis of Consensus. The fact that PROD does not (yet) replace AFD is extremely problematic. AFD does not scale --Kim Bruning 16:17, 11 May 2007 (UTC)[reply]
Okay, I don't really follow most of what you just said, but... no. I will not debate. KafzielTalk 16:19, 11 May 2007 (UTC)[reply]
Er, Kim, the precise reason we need AFD aside from PROD is that there's a need to, as you just said, debate (some of) our decisions, which isn't possible in PROD. >Radiant< 16:23, 11 May 2007 (UTC)[reply]
Yes, this is an issue with PROD. It does not properly allow WP:BRD on the PROD tag ;-) --Kim Bruning 16:25, 11 May 2007 (UTC)[reply]
And that is because many novice users are unable to distinguish between BRD and a regular edit war :P >Radiant< 16:28, 11 May 2007 (UTC)[reply]
Hmm... well, BRD can safely become quite heated, while there are still no 3RRs or page protections (or even (other) mediators) in sight? ;-) --Kim Bruning 16:35, 11 May 2007 (UTC)[reply]
Kazfziel, you're the one who started violating WP:WOTTA first ;-) Basically, on a consensus based wiki, you are required to (be able to) explain your actions (edits) to anyone who requests so, at any time, provided such a request is reasonable and made in good faith. This is a requirement that flows from the nature of the wiki, it would be very hard to maintain the community without it. If you are unwilling to discuss issues, there are likely several projects -even within wikimedia- that are less tied to consensus (although in my opinion, such projects under-utilize their wikis) --Kim Bruning 16:32, 11 May 2007 (UTC)[reply]
I am willing to discuss (to some extent) my decisions on RfA. But since RFAPROD is silence-driven instead of consensus-driven, I will not debate them here. People don't have to debate their reasons for not removing the tag (as nobody will even know they viewed the candidate's request) so people don't have to debate their reasons for removing it, either. KafzielTalk 16:37, 11 May 2007 (UTC)[reply]
Actually, the whole theory is based on starting a debate when you remove the tag. This allows you to find candidates who still need instruction on how to be an admin, and teach them how to improve. If you don't debate, then the excersize is kind of pointless. <looks apologetic> --Kim Bruning 16:54, 11 May 2007 (UTC)[reply]
No, actually the whole process is based on sending the person to RfA when you remove the tag. I'm not interested in finding candidates in need of admin coaching; there's already a process for that. But I'll agree that the exercise is kind of pointless, yes. KafzielTalk 16:59, 11 May 2007 (UTC)[reply]
Yes, that was insane, I just threw that out today. (note also the heading of this section). Also, admin coaching often coaches the wrong people. --Kim Bruning 18:06, 11 May 2007 (UTC)[reply]
So because you (and you alone) say we must not use RfA, that's the end of it? It's suddenly "insane"? Every argument you have leads back to that statement, as if it's some kind of obvious fact, but it isn't. You took out the RfA bit. It can be put back in. KafzielTalk 18:27, 11 May 2007 (UTC)[reply]
This is significantly different from the current functioning of RFA. In first instance, you just want a scalable RFA, not a request for banning-adminship :-P --Kim Bruning 18:08, 11 May 2007 (UTC)[reply]
First objection to retries: There's no history of prior attempts. Unless you wish to go through all of a user's userpage history, and also ask someone who can view deleted edits to make sure the page wasn't deleted upon request, then it's very difficult to obtain context on prior requests. With RFA, when Wikipedia:Requests for adminship/William Rotelle 37 shows up, I can at least know to look a little closer. My second objection is that multiple people have removed the retry clause by now, so it seems to be off the table. Semantics of whether it fails-over to RFA or some other process are not really relevant. PROD works the same way. -- nae'blis 18:14, 11 May 2007 (UTC)[reply]
Makes it impossible to BRD. Other Options? (Note: PROD should not fail-over either, or if it does, it should fail-over to a different scalable process.). --Kim Bruning 18:29, 11 May 2007 (UTC)[reply]
You know BRD isn't policy, right? You keep using that to invalidate all this stuff, but it's just one method of consensus building. KafzielTalk 18:31, 11 May 2007 (UTC)[reply]
Kim, I've already mentioned that BRD works on articles, not policies (proposals are neither fish nor fowl). You're not even arguing logically, so I think I'm done here. -- nae'blis 18:34, 11 May 2007 (UTC)[reply]
Actually, I've mostly tested BRD in policy disputes, where it works rather well (if rather slowly) --Kim Bruning 23:24, 11 May 2007 (UTC)[reply]
Hmm, 99% of the coolness factor of this is if people treat the template as any other wikicontent on a page. "You may not restore the template" sort of throws a 0RR spanner in the works. That and it typically gets re-added by people who consequently go "And it will not possibly work". Nope, it won't, big surprise ;-) --Kim Bruning 13:43, 12 May 2007 (UTC)[reply]
I must question the purpose of the experiment if it is to be limited to a small sample of users and to result in "If this had been a real process, you would now be an admin." What does that prove? That people may potentially be promoted through this process? I don't think anyone's disputing that ...
I just can't see a point to a trial. However, as it seems essentially harmless, I'm not opposed to it per se, as long as it's made clear that:
(A) For some reason, I started the list with "B" ... Since I really can't think of anything for "A", I'll just leave this comment. -- Black Falcon(Talk) 23:29, 11 May 2007 (UTC)[reply]
(B) The trial will not result in any promotions;
(C) The proposed adminship process does not have consensus support;
(D) The experimental trial does not have consensus support; and
(E) Lack of objection to any user who participates in the trial does not constitute community approval of that user.
Finally, in response to a comment made about 3 sections up ... I find the standard questions useful in evaluating a candidate and will expect all proposed administrator to answer them (either immediately before or after adding the tag or upon request). -- Black Falcon(Talk) 18:55, 11 May 2007 (UTC)[reply]
First actual experiment (as far as I know, since template transclusions are not trackable): added 11:55 Eastern US time, removed almost exactly 30 minutes later (by Kafziel). Now, this was a self-nom, so it doesn't meet all of the current parameters, but it is data of a sort. -- nae'blis 21:21, 11 May 2007 (UTC)[reply]
I was wondering if I was the first. And I was starting to wonder if anyone was even going to notice. Can't track it by the template since I removed it, can't track it by the category since I took him out of it... kind of highlights the need for some kind of noticeboard. Or not. KafzielTalk 21:28, 11 May 2007 (UTC)[reply]
In fairness, I'd never have noticed if jbolden hadn't posted the results up higher on the page. -- nae'blis 21:36, 11 May 2007 (UTC)[reply]
I agree with all points except possibly (E)... ;-) --Kim Bruning 23:26, 11 May 2007 (UTC)[reply]
Okay, if people are actually going be using this template it should not say "The user will be sysopped in 7 days from the date of this notice unless another user has a serious concern." This is just not so. It should say "This is a test, if this process is ever accepted by the community..." HighInBC(Need help? Ask me) 14:32, 12 May 2007 (UTC)[reply]
So if I propose somebody for adminship, and a single person who doesn't want any more admins, who has a pet peeve about adminship, or about the particular person, removes the nomination, that's it, it's done, and this person can't be nominated again? Corvus cornix 22:40, 13 May 2007 (UTC)[reply]
The person can be nominated at RfA just like they would now. KafzielTalk 22:45, 13 May 2007 (UTC)[reply]
I just don't understand what this accomplishes. One person removing the tag gets an absolute veto. One person voting oppose on an RfA doesn't get an absolute veto. How is this an improvement? Corvus cornix 22:59, 13 May 2007 (UTC)[reply]
In theory, it's an improvement because it's less "paperwork" (so to speak), the same way PROD was said to improve the AfD process. An extra layer of bureaucracy that might, in the long run, lessen the load on the more involved processes as long as it doesn't become too convoluted itself. That's the idea, anyway. KafzielTalk 23:10, 13 May 2007 (UTC)[reply]
Ok, so the prototype really doesn't work (as predicted). ;-)
Can we try a run allowing normal page editing rules (and thus also BRD for the fans ;-) ) ? In this variant, anyone is allowed to put the template back provided there is rough consensus to do so. Normal page editing concepts apply to the template (including WP:3RR, WP:HEC, etc...).
If the template stays on unchallenged for 7 consecutive days, the candidate is promoted.
If after ummm (picking numbers out of thin air) 14-21 days, there's no consensus to replace the template, a candidate must by default wait 3 months before they can be nommed again. (Why 3 months? That's traditionally the same amount of time it would take to get back to a point where you get nominated, if you were to start with a new account/identity.)
Would that work out? I'll list it as variant B... we'll likely have many kinds of variants after a while. --Kim Bruning 23:38, 13 May 2007 (UTC)[reply]
When these kinds of adaptations are necessary to make it possible that the process will work, I think we're at the point where this has ceased to be a simpler operation than RfA. Can a process as complex as RfA, and dealing with the same issues as RfA, be more scalable than RfA? Also, I'm not convinced that your proposal is is in the spirit of this proposal (to promote people when there are no objections). We have to argue over whether there is rough consensus to replace the tag, and then the survival of the tag is used to determine consensus? Under that variant, it appears that consensus for promotion has to be determined prior to the nomination. Dekimasuよ! 02:17, 14 May 2007 (UTC)[reply]
Actually afaict there's quite less text there than for RFA, and I've thrown out the entire RFA requirement. I don't see how it could possibly then still be more complex than RFA, but I'm sure you can do a side by side comparison to prove it. :-)
If variant B changes the original intent, so be it and this is a different intent?... Because the test we just did on variant A crashed and burned. (and I thought the original intent was actually B anyway, but perhaps that's not true then). If you hate B, please come up with a C?
The way I saw it was that using a template allows one to use normal editing rules. That way, you don't need to learn anything new to make someone an admin (doesn't matter how simple or complex a process, you don't need to learn anything new... big advantage there. Also, wiki-editing of pages is known to scale well, thereby reducing it to the previous problem ).
I don't understand what you mean by "consensus for promotion has to be determined prior to the nomination". Did you mean that such consensus must (potentially) exist? --Kim Bruning 02:59, 14 May 2007 (UTC)[reply]
I'll try to say it more clearly, though I'm not sure it will come out that way.
Variant B revolves around the ability to readd the tag, and the current text says readding the tag should be based on "rough consensus". Since "rough consensus" for promotion is what we're trying to judge through the whole process, the step of "readding the tag and waiting seven days" becomes meaningless. Readding the tag requires that the consensus to promote already exist, and how can such consensus be formed if the user isn't even listed in the category anymore? Dekimasuよ! 03:43, 14 May 2007 (UTC)[reply]
Not really. The idea is that if someone removes the tag, you can discuss with that person. They can then choose to re-add the tag (either that or everyone agrees that the remover was a troll :-) ). It works exactly like page editing. OTOH, maybe a lot of people don't know best practices for page editing? --Kim Bruning 14:02, 14 May 2007 (UTC)[reply]
Let me get this straight, the tag gets removed, but it can be put back if there is consensus to do so? So basically consensus determines if the the tag is there in the end and if the person is promoted. Wouldn't it be neat if there was a centralized location to handle this consensus building? Oh wait, that is WP:RfA. HighInBC(Need help? Ask me) 03:47, 14 May 2007 (UTC)[reply]
RFA has two issues. The first is that it is centralised and therefore cannot possibly scale. So removing the centralization is a good first step. Secondly, RFA is currently judged by majority, it seems. This is very very bad, and either that gets corrected or RFA must die :-/ . Page editing is by rough consensus, and is a previous problem which we already understand. Therefore controlling promotion by using page editing is a Good Idea (tm) :-) --Kim Bruning 14:02, 14 May 2007 (UTC)[reply]
Obviously, if edit warring occurs over the proposed-adminship tag, the proposed person has sufficient controversy that he shouldn't be made an admin "just like that". Similarly, if the tag was applied while other people weren't looking, the person shouldn't be made an admin "by the backdoor". That is why both PROD's "no replacing the tag" clause as well as the central logging page are deal-breakers for several of the involved here. Adminship is just enough of a big deal that people want it out in the open. >Radiant< 13:32, 14 May 2007 (UTC)[reply]
Edit warring? Is everything either flowers and unicorns and ponies, or hell and damnation for you, with nothing in between? :-P --Kim Bruning 14:02, 14 May 2007 (UTC) There are 10 kinds of people ...[reply]
(ps. all central pages MUST go, they're a set of rotten teeth that need to be pulled, sooner or later. None are long term viable. Either they go now, when we are all in control and ready to deal with the situation, or they go by themselves later, when we are not. :-/ ) Kim Bruning
I'm not saying that I object here (as I've said, I welcome a test run); I'm explaining why other people object, and note that there are sufficient such people to make test runs impractical. >Radiant< 14:39, 14 May 2007 (UTC)[reply]
I can see how this can weed out obviously unsuitable candidates, but it adds an unnecessary layer to the process which is prone to abuse if not monitored very carefully and even then, you'll get endless discussions on whether the opposition was in good faith. How is this any more helpful than simply opposing an already listed nomination? - Mgm|(talk) 12:15, 14 May 2007 (UTC)[reply]
I don't get it. Did someone add more than one layer while I wasn't looking? --Kim Bruning 13:49, 14 May 2007 (UTC)[reply]
You've been trying to remove that layer, but I'm sure he's responding to the original setup in which the proposal would just result in a tag added, a tag removed, and then a normal RfA. Dekimasuよ! 13:52, 14 May 2007 (UTC)[reply]
Yes, that was particularly silly. Some people are going "we don't mind reform, but it's a dealbreaker if that reform actually tries to do away with RFA". :-P --Kim Bruning 14:05, 14 May 2007 (UTC) I could be misreading, please don't shoot me ^^;; Maybe folks are saying something subtly different?[reply]
That is only silly if you think RfA is not doing its job, I personally see people who should getting promoted, and people who should not failing. I see no issues with scalability at this point, or any of the other problems people say are there. So I don't think that is a silly thing to consider a deal breaker at all. HighInBC(Need help? Ask me) 14:12, 14 May 2007 (UTC)[reply]
I expect/intend to be placing people on RFA at a rate of 10-100 per day in the near future. (I'm awaiting statistics first though). Will it hold? --Kim Bruning 14:24, 14 May 2007 (UTC)[reply]
If you think RFA is perfection itself, then surely you would not be worried by competing methods, for they will obviously fail to compete in the field. <predatory grin> --Kim Bruning 14:24, 14 May 2007 (UTC)[reply]
In that case, the scalability problem is not in the process itself, but in the small number of screeners, unless you can think of a way to ensure reliability without review. If the number of nominees exceeds the number of reviewers, the reliability will fall apart. Dekimasuよ! 14:28, 14 May 2007 (UTC)[reply]
Hmmm yes. You only need a small number of reviewers per candidate, if each reviewer carefully scrutinizes the person in question. The current level of review isn't too great, frankly. People just check edit counts and maybe a couple of edits, we get a lot of quantity of reviews, but not much quality. :-/ On the gripping hand, the admin flag is supposed to be no big deal. Right now admin-flag holders are turning into an elite (which must then be watched, etc etc etc... bureaucracy on bureaucracy). So I wouldn't even care if we had reduced standards... not that this method would result in causing reduced standards. We just need more admins. Hmmm, using formal requirements for adminship might be interesting to meet your requirement of ensuring reliability with (less) review? --Kim Bruning 14:36, 14 May 2007 (UTC)[reply]
Just because being an admin is no big deal that does not mean how we choose admins should not be a big deal. The reason for this is that de-admining someone is a big deal. If we lower standards then we will have more de-admining(and the disruption that precedes them), which none of us wants. HighInBC(Need help? Ask me) 14:41, 14 May 2007 (UTC)[reply]
10-100 people a day? If you really have that many people that you think will be good admins, perhaps you can nominate them at a slower rate, otherwise it is unlikely each will get the attention they deserve(no matter what system we use). HighInBC(Need help? Ask me) 14:32, 14 May 2007 (UTC)[reply]
We may have an (as yet invisible) backlog (awaiting statistics). The 10-100 figure is what we might need to catch up with that backlog. A system that gives *all* these people the attention they deserve is what we define as being called a "scalable system". --Kim Bruning 14:40, 14 May 2007 (UTC)[reply]
So now you're making unprovable claims about invisible backlogs, supported by nonexistent statistics from who-knows-where. I think I'm with nae'blis on this one: it doesn't seem like there's much sense talking to you, because you either ignore the points you can't answer or just make things up to suit you. I (and many others) have already given perfectly good reasons for rejecting this proposal, and none of them have been addressed. I was willing to discuss it, but I'm not willing to be railroaded. So I'll just say that I will remove this tag wherever I see it, and that's the end of it for my part. KafzielTalk 14:50, 14 May 2007 (UTC)[reply]
Dekimasu found relevant up-to-date statistics, and else I'd have gotten the up-to-date numbers later today. (I was waiting because else people could go "ohhhh, but those are OLD numbers... and now they can't ;-) ). So much for unprovable claims. O:-)
Note that violation of WP:POINT is a blockable offence, of course :-).
For the record, could you summarize which points of yours have not yet been addressed? --Kim Bruning 15:17, 14 May 2007 (UTC)[reply]
It's not a violation of POINT to oppose someone's adminship for any reason.
Can you summarize which points of mine have been addressed? I said this should never replace RfA. You ignored me. I said the purpose of this proposal is not to find unsuitable candidates for admin coaching. You changed the subject. I mentioned that BRD is not policy. You ignored me. I said that I see no need for RfA reform in the first place. You ignored me. You imply people are ignorant and our positions are insane, and because you put a little smiley in every post (three in the one above) we're supposed to pretend you're being civil and cooperative? You're being intentionally obtuse, so there's no point in arguing anymore. I will oppose everyone who uses the tag, on the grounds that it shows a fear of criticism and a distaste for due process. I dare you to block me for it. KafzielTalk 16:03, 14 May 2007 (UTC)[reply]
Point by point
Disrupting experiments because you feel like hurting other people's work is a violation of WP:POINT, and WP:DICK. You haven't done anything yet, so this is not actionable anyway.
Should not replace RFA This is an experiment, we learn new methods to deal with RFAs. Replacement of RFA is not on the table in the first place. Therefore this is not actionable at this point in time (or ever).
This method may or may not find people for admin coaching as a side effect. You don't have to coach them. This side effect would perhaps be somewhat desirable. However, desirable or not, we won't know if this effect will actually occur at all. That's what we need to experiment for. Therefore if this point isn't irrelevant, it is still not actionable at this point in time.
BRD is not policy. Correct point of fact. No associated action.
There is no need for RFA reform. Measurements show this to be factually incorrect. If I can't tell someone something politely, I shut up.
Implying people are ignorant. Did not! All I said was that Experiments have become uncommon. Recognizing that people are unaccustomed to them seems prudent.
Implying certain measures are insane. If the intent is to simplify a process and make it scalable, how seriously should one take a proposal that tacks the entire old process onto some new set of hoops, thereby making the combination twice as complex?
Dare me to block you. No. :-P
Do you have any actionable points remaining? --Kim Bruning 20:34, 14 May 2007 (UTC)[reply]
I've been saying from the very beginning that I could find a reason to oppose each and every person nominated this way. It's not just to hurt other people's work; it's how this proposal is supposed to work. I'm not going to be bullied into silence so people can become admins for the sake of your little experiment. Since removal of the tag is the only way to stop inappropriate people from becoming admins, I will do so. And, yes, I already have.
No one is going to become an admin based on this experiment. We do not promote people or in fact take any kind of actual irreversible action based on an experiment. It won't happen. That's stupid. Why would we do that? Why would you think we would do that? No way is that going to happen. --Kim Bruning 00:02, 15 May 2007 (UTC)[reply]
You're darn right it's not going to happen. And I'm not going to wait for this experiment to get forced through into a live trial before I start to do something about it. KafzielTalk 00:28, 15 May 2007 (UTC)[reply]
Um. It's not happening like that period. Not sure what you're going to do about something that won't happen. --Kim Bruning 00:46, 15 May 2007 (UTC)[reply]
I'm making it known that now and at any future stage, I will remove the tag from anyone who uses it. Whether it's at the stage of an experiment, a proposal, a test run, a way to augment RfA, or the replacement of RfA. If consensus means anything at all it won't even get past this stage. If it does, I'll still be around. KafzielTalk 12:09, 15 May 2007 (UTC)[reply]
Also, again, note that neither WP:POINT or M:DICK are policies. Although if you do want to start pointing the finger at others, which one would be more disruptive: my removal of tags from userpages one at a time and leaving explanations per the instructions, or your plan to nominate 100 unsuitable (even unwilling) candidates at RfA each day? KafzielTalk 21:08, 14 May 2007 (UTC)[reply]
If you say so. Arbcom has banned for violation at times. --00:02, 15 May 2007 (UTC)
Note that you completely ignored the part where I said your mass nominations would be far more disruptive than my by-the-book removal of the admin prod tags. Par for the course: if you know I'm right, just ignore me. KafzielTalk 00:28, 15 May 2007 (UTC)[reply]
Ah, that was in this section. But weren't you saying that RFA was perfect and has no issues? If we can't use PROA, and RFA is perfect anyway, why not just use RFA? Shouldn't be disruptive at all, I'm only using RFA for what it's meant for. --Kim Bruning 00:46, 15 May 2007 (UTC) Actually that's exactly the kind of reasoning that leads to a WP:POINT violation. I need to somehow avoid violating WP:POINT, and one way to do so is to suggest changes far in advance. At the very least then I can say I made a good faith best effort[reply]
You could have said you made a good faith effort if your suggestion went unnoticed. But we're calling you on it and telling you it would be counterproductive, and you're aware of that, so now that defense doesn't hold up.
RfA has no issues if used properly. But quality nominations are important. You can nominate all the editors you want, provided you do all the necessary homework beforehand. Your mass nominations would be counterproductive because a) you don't intend to vet any of the candidates properly, and b) we all know there aren't actually that many suitable candidates waiting in the wings anyway, so 90% of the people you sent up would be snowball failures and a waste of everyone's time. KafzielTalk 12:09, 15 May 2007 (UTC)[reply]
Your patronizing tone to Dekimasu (accompanied by yet another of those tiresome emoticons, because you knew it was patronizing) did not go unnoticed. You can split hairs all you like, but you were implying that he was ignorant.
Dekimasu can speak for himself. I certainly don't intend to be patronising. --Kim Bruning 00:02, 15 May 2007 (UTC)[reply]
More fake data? What "measurements" provide absolute proof that RfA is broken? Seems to me we're promoting new admins just about every day, and I'm not seeing any giant backlogs.
More fake data? What fake data was there before? --Kim Bruning 00:02, 15 May 2007 (UTC)[reply]
Your completely unfounded claim (waiting on "statistics") that there is some massive invisible backlog of administrator-hopefuls. Fake data. KafzielTalk 00:28, 15 May 2007 (UTC)[reply]
I just wanted up to date data. I knew there was a backlog. Dekimasu has provided a link to current-ish data. --Kim Bruning 00:46, 15 May 2007 (UTC)[reply]
And, inevitably, you come back to scalability. If all we want is an easily scalable system, we can have a straight vote. That allows clear and fast input from millions of people at once. This proposal was not written simply to solve the scalability issue; it was written to make things easier for editors who are either too shy or too lazy to go through RfA. It takes just as much work to review an editor under this proposal, with none of the visible input one finds on RfA. This can never replace RfA because—another unaddressed point made by HighInBC, me, and others—it is based on the lack of vocal opposition, rather than the presence of vocal support.
Straight vote sucks for too many reasons which won't fit in this margin. It also doesn't scale in the case of many voters, many candidates. --Kim Bruning 00:02, 15 May 2007 (UTC)[reply]
Yeah, no country has ever had a lot of voters and a lot of candidates. (Note that you are still ignoring the problem of silent consent.) KafzielTalk 00:28, 15 May 2007 (UTC)[reply]
Typically not that number of candidates, no. This is more like handing out drivers licenses. --00:46, 15 May 2007 (UTC)
And no country has ever handed out millions of drivers licenses? Of course they do. And the more people get cars, the harder the tests become. They don't just change the rules to say "hold up this sign and, if nobody notices you, you automatically get a license, because all this testing is just too time-consuming". KafzielTalk 12:09, 15 May 2007 (UTC)[reply]
If you have a better idea, make a new proposal. As far as I'm concerned, this one has failed.
This is not a proposal, this is an experiment. We do a series of experiments to find out what works. Note that experiments do not get real world consequences. That's what they're experiments for.
So it stops being a proposal when you decide to tag it otherwise? I see. KafzielTalk 00:28, 15 May 2007 (UTC)[reply]
That was the wrong tag. We should not make a proposal at this point in time. --Kim Bruning 00:46, 15 May 2007 (UTC)[reply]
Says you. What about the next guy who comes along and decides to change the tag again? You can call it what you like, but it's still a proposal. KafzielTalk 12:09, 15 May 2007 (UTC)[reply]
replies interleaved --Kim Bruning 00:02, 15 May 2007 (UTC)[reply]
I was severly edit-conflicted. I'll bite by putting my answer here: Either way, we'll only have 20 bureaucrats, and my RfA is seven hours overdue. I don't think the human side can catch up to a large increase in the speed of the process. And I'll take the bait one more time... I dispute the idea that the number of admins we need is increasing exponentially (or at some other high rate). Despite the increase in the number of overall user accounts and overall articles, the amount of administrator actions needed at any given time is actually a function of the overall number of edits to the site, and we haven't seen large increases in those (outside of bot edits and AWB) for a while now. Nominating 100 people a day doesn't make sense when the number of "very active Wikipedians" (over 100 edits in a given month) has hovered around 4000 total for the last 6-8 months of available data (here). Or you want your work to be done in a few weeks? Dekimasuよ! 14:50, 14 May 2007 (UTC)[reply]
Now this is the kind of discussion that's going someplace. <grin> There is a constant coming and going of people on wikipedia, so we constantly need new people with admin flags to replace those that leave. Indeed I would like to get rid of the current backlog in a couple of weeks, and then continue at a more friendly pace (though that pace might still be a bit higher than folks are used to) . I think that the important number is the number of admins per user, because it is users who ask for admin assistance, rather than edits. I do believe the number of registered users is still growing. It's ...interesting... to hear that number of edits has become constant, I would have thought that that number should still be growing atm. That's something that needs looking at elsewhere. --Kim Bruning 15:29, 14 May 2007 (UTC)[reply]
What sort of criteria does one use to create a "backlog" of potential admins? Also, for it to be a scalable system by my definition, it would also need to give the correct attention to the reviewers, and Wikipedia's expectation of admins. HighInBC(Need help? Ask me) 14:42, 14 May 2007 (UTC)[reply]
I'm asking someone on toolserver for a list with everyone with 2000 edits/4 months. This does seem to be the predominant RFA requirement. Even if I disagree, at least it is a measurable objective criterion. This would be sufficient for me to nominate. People then may need to judge subjective criteria (if you insist on having them), or semi-objective criteria needing human checks. --Kim Bruning 14:50, 14 May 2007 (UTC) caveat: awaiting statistics, note: it is sometimes nescesary to modify requirements slightly to meet a scalability requirement., note 2: wikipedia is growing. Scaling is happening. Subsystems either scale along with the growth of wikipedia or they break.[reply]
Yikes. That's a long and somewhat misleading list. I'd nominate all not-bold, not struck-through people on that list, and simply see what happens... except some of the strikethroughs (such as "not interested at this time"), actually should actively be nominated first :-/ . I'm kind of POed at the acceptance requirement atm, btw... I'd try to get that tossed out before I start nominating :-) --Kim Bruning 15:14, 14 May 2007 (UTC)[reply]
Nominating a very large number of editors without at least a look into their block logs and a short scan of contribs and talk page would be counter productive. 2000 edits/4 months alone doesn't really mean very much either way. RxS 17:15, 14 May 2007 (UTC)[reply]
Doing RFA's work for them eh? So how do you propose proceding, knowing that I wish to nominate such a large number of people? It would be rather tricky to do as you suggest on my own. --Kim Bruning 20:37, 14 May 2007 (UTC)[reply]
No, not doing it's job for them, it's not wasting their time. Bulk nominating editors for RFA without any consideration of their suitability is no better then Boothy opposing every nominee without any consideration. I don't see how tricky it is if you're really interested in the end result (rather then making a point about process). You could check a 100 editors in a days work...block log, scan the talk page and a quick scan of contribs, you'd give everyone plenty to talk about. RxS 21:40, 14 May 2007 (UTC)[reply]
I have a real life. :-/ Any other approach? --Kim Bruning 00:03, 15 May 2007 (UTC)[reply]
See? Right there. You blew off RxS's whole point with nothing more than a sarcastic comment, but there's that little emoticon to make everything all better. Unbelievable. KafzielTalk 00:13, 15 May 2007 (UTC)[reply]
I actually do have a real life, and this sounds like I'd need to spend 10-40 full-time days at some point. That's just not practical. So I figure we need to split that out across multiple people somehow, or maybe use some sort of heuristic. Either that or we drop or reduce the requirement. All this based on RxS's time assessment. Do you have any constructive suggestions how to approach that? --Kim Bruning 00:24, 15 May 2007 (UTC)[reply]
No, but apparently neither do you so I don't see why you keep asking everyone else. In my book, RfA is fine. So I'm not the one who needs to come up with the great ideas. You can't make a suggestion and then announce that you don't have time for it and propose we all split the work amongst ourselves. If you don't have time to vet those candidates yourself, it would irresponsible to nominate them and unreasonable to ask anyone else to do the work for you. Rx StrangeLove's point is that bulk nominations are not the answer, and time spent disrupting RfA with half-assed nominations would be better spent investigating more thoroughly. KafzielTalk 00:36, 15 May 2007 (UTC)[reply]
"RFA is fine but there is a huge backlog". Ok, so if I submit the entire backlog at once, there shouldn't be a problem.
That was my reasoning before, but now it turns out that I need to vet all of them first, all on my own.
But I can't do that, there's too many. So that idea needs tweaking.
So what can we do to deal with the backlog? --Kim Bruning 00:50, 15 May 2007 (UTC)[reply]
Again, you're the only one saying that there is a huge backlog. I don't see it. I also don't see a need for a new system, because the current system works just fine. Vandalism gets fixed. Vandals get blocked. Nonsense gets speedied. Images get deleted. Articles get protected. People who should be admins get nominated, people who shouldn't don't. Everything is being properly maintained, and good new admins get the buttons every day. Wikipedia is not falling apart. Your whole premise is based on the idea that there is some imaginary giant backlog, and that is simply not true. KafzielTalk 12:09, 15 May 2007 (UTC)[reply]
Don't have time to do a hundred? Do 10 a day, or do 5 a day. That'd be better then none a day, no? Get some like minded editors together and do 15-20 a day and by the end of the week you could have 2 or 3 hundred. That's even better then none. But to throw out hundreds of nominations that you haven't even looked at creates such a duplication of effort that it's disruptive. Why ask 10 or 20 or who knows how many editors to do the same vetting that you could have done once? If you can't at least cull out those editors that clearly are not suitable for RFA then it's just a bad idea. RxS 03:37, 15 May 2007 (UTC)[reply]
I thought the whole point of RFA was to do the vetting. RFA is not there for the sake of RFA, it's there for the sake of figuring out who should be an admin. If we have to institute a new vetting procedure anyway (because that's what you're basically proposing), couldn't we skip RFA entirely? <grin> Why should we settle for merely better than nothing? Isn't that almost the same as giving up? --Kim Bruning 08:13, 15 May 2007 (UTC)[reply]
No, the point of RFA is to determine whether an editor can be trusted with the tools. The vetting I'm talking about is (as I've said a couple times already) is to cull out editors who are clearly not suitable for RFA. That vetting is being done now by nominators. You're proposing removing that intial vetting and making each editor do it on their own, which of course is a complete waste of time if you're really interested in improving the process. And I would think that with all the hand waving about catching up and backlogs 100 new nominations in a week would be a good thing rather than "giving up" RxS 14:52, 15 May 2007 (UTC)[reply]
If you don't like a particular experiment, typically the best solution is to create a modification that does meet your requirements. This is something people who are only used to a proposed process kind of process might not be used to ? :-)
Anyway, could people list their requirements here?
--Kim Bruning 14:11, 14 May 2007 (UTC)[reply]
I think one of my questions about the modification is this: If the experiment is going to be changed beyond recognition, why are we still discussing it here at Wikipedia:Proposed adminship? We have proposals for changing or eliminating RfA up the wazoo, but the specific one that was originally proposed here does not seem to be the solution. Dekimasuよ! 14:25, 14 May 2007 (UTC)[reply]
Well, you start out with an existing proposal that seems the most viable, and then conduct experiments to see where there are remaining issues with it. We then tweak the system until it works. It might be that the final outcome is totally different from the original prototype.
However, in the mean time, we will have learned a lot about what works and what doesn't. Ideally, we might even end up with a working prototype which can be used for real in the field :-). If not, we can pick up a different concept and try again, applying what we learned in the previous experiment. At some point, we'll find something we really like.
--Kim Bruning 15:03, 14 May 2007 (UTC) And thank you for fixing the refactor :-) [reply]
Since even Kim is agreeing that we're no longer working with the original 'proposed adminship' proposal/experiment/trout anymore, I have marked the primary page as rejected. New discussions on new p/e/ts should ideally take place somewhere else (Wikipedia_talk:Requests for adminship/Reform perhaps?), but we've got a collaborative effort going here right now so no sense wasting bureaucracy. It can always be refactored to a new page if we get anywhere. -- nae'blis 21:08, 16 May 2007 (UTC)[reply]
Scalable (must work even if we start adding many more people)
Reliable (must have similar results in similar situations)
Correct (must have the correct outcome, based on objective parameters)
Transparent (must be easily navigable and the workings of the process must be readily apparent to the uninitiated)
Beneficially attractive (must attract good candidates in sufficient numbers without providing an open door to poor candidates)
Must include open discussion where the merits and weaknesses of a proposed nominee can be seen by others for further investigation and comment.
Must be maintainable from the start (taking 15-30 minutes to do research per RfA is going to be a problem when you get 700 nominations the day the proposed system goes live)
Someone who knows more about statistics and population growth stuff might be able to explain this, but I found the comment somewhere above about the number of edits to Wikipedia remaining roughly constant over several months interesting. This suggests that Wikipedia's exponential growth (maybe not strictly exponential, but you know what I mean) is starting to plateau. First question: how can this be determined more accurately? Second question: what does this mean for the "scale problem"? Carcharoth 16:09, 14 May 2007 (UTC)[reply]
Let me get back to you on the questions. --Kim Bruning 00:29, 15 May 2007 (UTC)[reply]
