Wikipedia talk:Requests for adminship/Archive 147

Archive 140

←

Archive 145

Archive 146

Archive 147

Archive 148

Archive 149

Archive 150

Wikipedia:Requests for adminship/Ironholds 2

A discussion that probably needs better visability and more input-- Wikipedia talk:Requests for adminship/Ironholds 2. Cheers, Dlohcierekim 03:16, 8 October 2008 (UTC)

That's a very interesting RfA. AdjustShift (talk) 15:11, 8 October 2008 (UTC)

Yes, it certainly is, and that's kind of a new idea (at least for most recent times). I might comment around there soon; glad you left a heads-up note about it here, Dloh. Jamie^☆S93 15:35, 8 October 2008 (UTC)

If this is the place to comment on the novel process rather than the candidate, then I'd just like to point out that {{User:SQL/RfX Report}} needs updating - which could have lost some participants in the beta test. ϢereSpielChequers 18:39, 9 October 2008 (UTC)

I suggest transcluding User:Tangotango/RfA Analysis/Report in User:SQL/RfX Report in the meantime, since it isn't updated anyway. --Amalthea^Talk 18:47, 9 October 2008 (UTC)

Wikipedia:Requests for adminship/Raggonix

Would anyone please close per WP:SNOW. Cheers, Dlohcierekim 22:54, 9 October 2008 (UTC)

Done. You could have closed it yourself you know. -- how do you turn this on 23:02, 9 October 2008 (UTC)

He opposed in it. Closing it would've been a conflict of interest. And I added the unsuccessful RFA to the archives. Useight (talk) 23:04, 9 October 2008 (UTC)

Thanks. Dlohcierekim 23:06, 9 October 2008 (UTC)

@Useight, well I was going to but edit conflicted you - both times! :-) -- how do you turn this on 23:06, 9 October 2008 (UTC)

No worries, it happens all the time. Useight (talk) 23:11, 9 October 2008 (UTC)

Question on proposed new format

Sorry if this has been covered, I'm having difficulty catching up on all the reams of conversation that lead to this experiment. I was just wondering what would happen to all of the requests that are generally closed quickly on the grounds of WP:SNOW or WP:NOTNOW. Under the new system would enthusiastic new users with little to no chance of success have to go through five days and dozens of questions? Would early out of procedure closes be tolerated on the basis of discussion rather than voting? Guest9999 (talk) 13:11, 10 October 2008 (UTC)

I think those cases can be picked out pretty easily from answers to questions. Many snowed RfA's have poor answers to questions. I propose that requests to close per SNOW and NOTNOW be made in the discussion section before voting commences. People can then vote for NOTNOW closure but not to support/oppose the candidate. —Cyclonenim (talk · contribs · email) 13:16, 10 October 2008 (UTC)

Cyclonim is correct. There should not be any change: Whether people can vote or can discuss and vote later, RfAs will still be closable sooner than usual, either because of withdrawal or because of SNOW or NOTNOW. After all, people just vote for NOT NOW because they review the candidate and find there is no reason that he/she could ever pass that RfA. That can still be said in discussion... After all, WP:NOTNOW and WP:SNOW closes are just WP:IAR - and you can ignore rules with any system just the same ;-) SoWhy 18:48, 10 October 2008 (UTC)

Trial adminship - timed

OK, one last attempt to salvage something that might meet some of the objectives WITHOUT allowing anyone to get an end run around RfA or the current standards.

Where an RfA would normally fail, but the candidate has a comfortable majority in support, and much of the opposition is due to uncertainly about the candidate, or many opposers indicte a willingness for a trial, the closing bureaucrat may offer the candidate "trial adminship". Trial adminship shall last eight weeks, at the end of which the candidate shall automatically be desysopped. After the trial is over, or at any time during it (advised minimum being six weeks in) the candidate may apply for full status through a second RfA, at which current RfA standards rules shall apply.

Advantages 1) The community and not the crat now gets the last word. 2) The community can look to the trial results to help determine suitability. 3) "Trial admins" have a strong incentive to learn to do it properly (and should be offered mentors). 4) To remain an admin, you still have to meet the current RfA standards - no lowering the bar. 5) People who we might think will /probably/ be good (but might not) can have an opportunity prove themselves - with no long term harm if it turns out bad. 6) No new process is required here. 7) Crats just need to leave a note somewhere for the stewards with the scheduled date for trial end. ((7) We can leave in a provision that a trial period can be prematurely terminated if any two crats certify that the candidate is clearly proving unfit. - this maybe isn't vital))

Thoughts?--Scott MacDonald (talk) 20:21, 10 October 2008 (UTC)

That makes the bureaucrats the judge of a user's conduct (with #7); I can't speak for all my fellow 'crats, but I have more pressing issues at hand than to babysit everyone that passes a "temp RfA" (not to mention it's a major departure for us: currently, we gauge consensus, not the candidate; you're proposing we have a hand in the latter. #4 also proves problematic, as there aren't any "current RfA standards" with which to hold people to. EVula // talk // ☯ // 20:31, 10 October 2008 (UTC)

OK, we can scratch #7 - it was intended as an emergency measure in case of blatant disasters NOT for babysitting (maybe saying any arbitrator would be better, they are there to judge conduct). By "current RfA standards" I just mean go through RfA as it is right now.--Scott MacDonald (talk) 20:37, 10 October 2008 (UTC)

Point 4 (and associated). Doubling of the effort and input required by making candidates run twice. Point 5. "No long term harm". Two bad blocks and two bad deletions driving off editors is long term harm. I've no idea why you think a few bad admin calls are not going to create long term damage but you're simply wrong. Point 7 - per EVula. Pedro :  Chat  20:56, 10 October 2008 (UTC)

Answerable. Any candidate who fails with 65% is only going to certainly run again anyway - so that's not an issue. Indeed if his second run is after when we have more evidence, then it is more likely to be definitive. And that type of harm can be done by any admin, there's no evidence it will be done by a trial admin more. Indeed since they are looking over their shoulder to their next RfA is is probably a good deal less likely that with some other "full status" new admin, who will not lose his status for that type of behaviour. Point 7 is certainly negotiable.--Scott MacDonald (talk) 21:02, 10 October 2008 (UTC)

Good thoughts. However "there's no evidence it will be done by a trial admin more". Yeah, I have to agree because there are no trial admins. Impressed by some of your responses, unimpressed by that argument to say the least. Pedro :  Chat  21:08, 10 October 2008 (UTC)

You made the assumption that there would be more problem with trial admins. I simply pointed out that there's no basis for that. Not just because there are none, but because logical deduction says that they have strong motivation to behave.--Scott MacDonald (talk) 21:28, 10 October 2008 (UTC)

Forgive me Scott, but you've made that argument a number of times on various pages. Are you actually of the opinion that a candidate who tows the line slavishly because they are being "watched over" is a good thing? Your argument is fundamentally flawed - by yourself. The last thing we need is people who will play the game (i.e. have motivation to behave) to get the bit. We need people who can demonstrate WP:CLUE from about day 16 of their editing history. Pedro :  Chat  21:38, 10 October 2008 (UTC)

RfA as it stands is about the worst way of measuring clue I can envisage. I see no evidence of hoards of gamers - but I see plenty of cases when we discover incompetence too late. Trial would out incompetence far better than RfA second prophesying future behaviour based on edit summaries and who your friends are.--Scott MacDonald (talk) 21:47, 10 October 2008 (UTC)

Oppose and here's why: admins execute their actions (under normal conditions) with the trust and power of the community behind them. To gain that trust and power, that community has to grant it. Trial admins have no such trust and power, or it has not been demonstrated by consensus that the community is behind them. This "trial admin" proposal seems to have forgotten about the foundation of adminship: the trust of the community. —kurykh 21:12, 10 October 2008 (UTC)

Exceptionally well stated. 100% agreement. Pedro :  Chat  21:15, 10 October 2008 (UTC)

Well, except can't we discuss things without jumping in with polarising bolded votes. You seem to think the "trust of the community" is binary and somehow inflexible. Trust sometimes needs to be tested - and we are talking about candidates that a clear majority of people would trust with indefinite full adminship, and other people (whilst opposing that due to some uncertainly) might happily trust with a trial run. You argument sounds good, but really doesn't stack up.--Scott MacDonald (talk) 21:28, 10 October 2008 (UTC)

Your arguments betray a fundamental misunderstanding or lack of understanding of both RfA and adminship. First, you start out with "polarising bolded votes". Ever thought about calling it a concise position being put forward? But no, you took it as a vote in an attempt to belittle it. Second, trust is something that the community grants or withholds as a result of past actions. Actions, especially admin actions done with no coherent community trust (not individual trust, but community trust) have no power behind them, making them fundamentally shaky. Third, RfA is not and should not be majoritorian. What is "comfortable majority"? What is "clear majority"? 60%? 70%? Wikipedia operates on consensus, and a consensus of the community should be the arbiter of who gets the bit and who does not. —kurykh 21:48, 10 October 2008 (UTC)

I guess I operate differently. When I put forward a proposal, I am not necessarily sold on it, just willing to enter into a discussion. Point out the difficulties, let's talk about it, can we work out a better way? Can it be improved? Perhaps in the end it will be no good, and I might even agree? But when people jump in with "oppose" we are running to the barricades, and that's not how I see discussion and consensus best arrived at. I don't mean to belittle people who do it, I just find it most unhelpful. Sure wikipedia works on consensus but we could have different levels of that. We could have people saying "oppose this candidate, unless it is just for a trial" - add those opinions to the outright supports and maybe you've got consensus for a trial. You say trust is given for "past actions" - I agree. But maybe the past actions would be enough in one case to trust someone with a trial period (trust can have degrees) - and then judge whether to trust them on past admin actions. Maybe I'm wrong, but maybe not.--Scott MacDonald (talk) 21:59, 10 October 2008 (UTC)

Sigh. You state that you see no evidence of hoards of gamers - you obviously haven't visited RFA recently. Your process will create even more of them. Then you state RFA is related to who your friends are and your edit summaries. Well, guess what - as a trial admin my mates would be plentiful and my edit summaries spectacular. Scott, I admire your efforts, I like your approaches, but you are beating a dead horse and your argument is becoming less cohesive with deeper analysis. You admit you may be wrong. Well, just this once my man I'm afraid you are. Pedro :  Chat  22:05, 10 October 2008 (UTC)

If people are too bothered to read the rationale behind the bolded first word then that's their problem, not mine. But your argument is full of hypothetical statements people might say to back your premise up, and such an argument is on anything but solid ground. With all due respect, I'm not only criticizing the proposal but the entire idea. You say trust has degrees: I'm not saying individual trust here. I'm saying community trust, and I cannot emphasize that enough. There is no "trust if trial, no trust if without" with the community; actions carried out under either banner will be assumed to be equal, but with unequal backing of the community. —kurykh 22:08, 10 October 2008 (UTC)

Good or bad, RfA is what it is, and it's clearly not going to change any time soon. An editor unfairly blocked by a trial admin is no more likely to be understanding than if (s)he was unfairly blocked by the real thing. Result? Another pissed off editor. --Malleus Fatuorum (talk) 21:30, 10 October 2008 (UTC)

Oh, I am extremely pessimistic about ever changing RfA - but that's not a reason not to try. A bad admin call is a bad admin call - the question is how we do effective quality control selection. I'd say trial adminship is MORE likely to give us better quality overall.--Scott MacDonald (talk) 21:37, 10 October 2008 (UTC)

I have no reason to be a friend of the present RfA process, quite the reverse. I do, however, see the logic in the argument presented by Kurykh above, notwithstanding my personal view on the value of what is considered to be consensus on wikipedia. --Malleus Fatuorum (talk) 21:46, 10 October 2008 (UTC)

• Slowly unbundle the tools, like we did with rollback IMHO. That's the answer. Trial adminship is a bureaucratic nightmare with a hundred holes in it. Scott - I admire your persistence and belief in the idea. Good work on floating the idea, and I agree we should not be defeatistst about RFA change (by which I mean the culture over the process). But let this "trial admin" idea go. It's unworkable, labour intensive, process wonkery. Please - divert your evident passion and commitment to another idea than this one - it's not going to happen. This is not due to a lack of will on the communities part but because it's, basically, not a very good idea. Sorry to be blunt. Pedro :  Chat 

Unbundling is the only sane option. Will the world come to an end if some non-admin editors are allowed to look at deleted pages? Did it come to an end when rollback was granted to non-admin editors? The way to see who can be trusted with the tools is to trust them with the tools, or at least those tools that the community sees fit to prise from the unwilling hands of the administrators. --Malleus Fatuorum (talk) 22:00, 10 October 2008 (UTC)

I might be willing to support unbundling the protect tool... —kurykh 22:14, 10 October 2008 (UTC)

• • I'm happy to discuss your suggestion, but my initial reaction would be "bureaucratic nightmare with a hundred holes in it" ;). Maybe my idea sucks, but if we tell people to let their ideas go a few hours after they propose them, then we will certainly never change anything. This discussion is good. We all need to listen and reflect. We don't make progress when we argue for our preconceptions, we make progress when I, or you, consider an idea we initially think sucks and then say "oh, wait a minute, if we did it this way, and improved that, it might be worth a try". I think one problem with wikipedia is that for a modern project we are incredibly conservative about our institutions. Someone will resit (if not block) every change. If we keep on like this, Wikipedia will never evolve - and without evolution there is only eventual redundancy. Wikipedia can always improve - we just need to open minds.--Scott MacDonald (talk) 22:07, 10 October 2008 (UTC)
    • The problem with changing is that there is no consensus on what to change, or even if change is needed at all. We do not change for the sake of change. We change to answer a pressing problem that needs fixing, and it is my opinion that the process needs no fixing at this time. Sure, there are novel ideas, but the ideas just don't seem to improve the process. —kurykh 22:13, 10 October 2008 (UTC)
    • The RfA process is really the least of wikipedia's problems I would suggest. Sure, it's a degenerated into a hazing, but that was always inevitable: "I had to have my head stuck down the toilet, so why shouldn't you?" I remain to be convinced that wikipedia is short of administrators anyway. --Malleus Fatuorum (talk) 22:20, 10 October 2008 (UTC)

Unbundling the tools

Unbundelling the tools is worth consideration, but my initial reaction is "how do we do that without a tonne load of extra process?". Of course, there may be ways of doing that. But one observation - someone said "would it be a big deal if non-admins could read deleted edits?". The answer is unfortunately, yes it certainly would - far bigger than allowing non-admins to block people, in fact. Here is why. We get a lot of libels on wikipedia, and a lot of privacy violating information. Whilst a very very small proportion of such edits are oversighted, a fair number are simply deleted. (Attack articles, vandalism with people's phone details, BLP violations.) When a subject complains about a libellous article or that slanderous edits still exist in an edit history, OTRS people will routinely delete such edits. That's generally fine. As long as deleted edits can only be viewed by a relatively small number of fairly trusted users, then subjects will often be content that the information is no longer "public" or "published". However, if you open deleted edits to wider viewing (and we already have 1400 viewers) what will happen is a massive increase in the demand for oversight. And oversight can't easily be reversed, or challenged as a mistake, and really needs restricting.--Scott MacDonald (talk) 22:20, 10 October 2008 (UTC)

Viewing deleted pages is a big no no as far as I'm concerned - it's the one thing that can cause us real harm if things get into the wrong hands. This is especially important when we consider that images are not oversightable at present. The way to go might be to unbundle the protect button - that's generally fairly uncontroversial (well, except for the latest Sarah Palin RfArb) and I'm sure there's plenty of people who could use it without any problems. We could perhaps move onto the delete button as well for people who work in deletion areas such as CSD and AfD (of course, they'd have to show some level of competance first!). Ryan Postlethwaite^{See the mess I've created or let's have banter} 22:26, 10 October 2008 (UTC)

I was thinking a semi-protect button would be the next step after rollback. Gwen Gale (talk) 22:34, 10 October 2008 (UTC)

Agreed. Protect is not a big deal and most decisions (Sarah Palin excepted, obviously) are relatively uncontroversial. Try this on a trial basis with users who are already shown to have good judgement, maybe? This also provides an excellent background for someone who decides to RfA; have they already been trusted with rollback and protect? Make it part of WP:PERM and easily revokable by admins who feel it has been used abusively (as opposed to honest mistakes). Leave it up to admin discretion, same as rollback. Perhaps also an understanding that users granted the right may only use it to semi; full-protecting a page is an instantaneous removal of the right. Prince of Canada ^{t | c} 06:52, 11 October 2008 (UTC)

Simple. We need a software change, but with will and some nice talking to the devs this can be done. Rollback and Account creator are already done so let's look at the rest:

• Rollback - given to all trusted users by admins on review. (in place)
• Account Creator - given to all trusted users by admins on review. (sort of in place)
• View Deleted - given to all trusted users by admins on review.
• Delete - given to all trusted users by admins on review but only with specific reference to accuarte speedy deletion tagging or WP:AFD closes with subsequent request.
• Restore - given as part of the above right.
• Semi-Protect - given to all trusted users by admins on review but only with specific reference to WP:RFPP edits and/or zero edit warring.
• Protect - not given out.
• Block - not given out.
• Mediawiki - given to all trusted users by admins on review to users who have demonstrated proficency through bot / template coding
• Admin - given via WP:RFA by bureaucrats as per current process. Adminship discussion includes the fact that admins have the abilities to grant and revoke these rights.

Pedro :  Chat  22:37, 10 October 2008 (UTC)

Per above Pedro, view deleted is one of the worst we can give out - it's the only one that can have real world implications. Ryan Postlethwaite^{See the mess I've created or let's have banter} 22:39, 10 October 2008 (UTC)

(ec)I'd say this is the wrong way about. Block/unblock should be given BEFORE undelete and viewing deleted edits. Blocking or unblocking only affects users and the project - it has no real world implications. Whereas undeleting the wrong stuff, particularly with BLPs can have massive implications to people well beyond wikipedia. The ability to access or restore deleted edits in the one thing we should guard above all. The rest I'm happy to give out incrementally (even delete buttons).--~~

I'm hesitant about giving out the block/unblock, viewing deleted edits and delete/undelete functions. I am definitely sure, however, that the protect/unprotect button can be given out to others rollback style, with sufficient oversight. —kurykh 22:45, 10 October 2008 (UTC)

(edit conflict) Just a comment after a quick review of your suggestions, Pedro: You can not separate semi-protection from full-protection, as technically that's within the same right (protect). —αἰτίας •discussion• 22:46, 10 October 2008 (UTC)

I don't think there's much to unbundle. Gwen Gale (talk) 22:48, 10 October 2008 (UTC)

To aitias - yep - I did note a software change would be needed. To Ryan - yes, but we have oversight as about 99.999% of deleted stuff is crap. To all, just a suggestion that unbundling does not seem as awful as it may be. Giving trusted users "delete" to G10's would seem a good idea, no? We should not let software and technical issues influence us. The software 'can be changed. The political opinions will be harder! Pedro :  Chat  22:52, 10 October 2008 (UTC)

Unfortunately, that would mean we'd need to oversight every single G10 or BLP page we delete, which would require a hell of a lot of new oversighters! Also, images are currently unoversightable so that would need to be fixed. Ryan Postlethwaite^{See the mess I've created or let's have banter} 22:54, 10 October 2008 (UTC)

Why? Only trusted users could see attack pages. That's maybe 200% of the current editors with +sysop. I'm struggling to see the damage / harm or I may not be making myself clear. Trusted users seeing deleted pages would seem to be academic. Pedro :  Chat  22:58, 10 October 2008 (UTC)

The problem isn't the viewing aspect, it's the potential to release it to the press or other agencies, or even the person that the libel/attack is about - it's the whole reason why viewing deleted material is the only one that can make a real world difference. We have all sorts in the deleted images db - I know that one admin had to deal with child pornography, do we really want non-admins seeing stuff like that? I can just see the headlines, "Wikipedia, the free child pornography collection that anyone can look at". Ryan Postlethwaite^{See the mess I've created or let's have banter} 23:03, 10 October 2008 (UTC)

So we trust admins to look at it, but we don't trust them to let other people look at it?Pedro :  Chat  23:12, 10 October 2008 (UTC)

No, but the fewer people that can see it the better. However, we do trust that admins wouldn't do something stupid with it should they slip across it in some logs. Ryan Postlethwaite^{See the mess I've created or let's have banter} 23:15, 10 October 2008 (UTC)

I'm afraid unbundling view-deleted is out of the question since this comment by MGodwin. --Amalthea^Talk 00:08, 11 October 2008 (UTC)

(EC) The unbundle that I would really like to see is bot flagging given to the BAG. bibliomaniac15 22:56, 10 October 2008 (UTC)

Generally speaking I think all the rights mentioned by Pedro, which are not already unbundled, are (highly) sensitive. They, or at least some of them, can evidently cause serious harm if they are used inappropriately/incorrectly/abusively. Therefore I really believe there is a need to be approved by the community and not just by one single administrator to use them. —αἰτίας •discussion• 23:16, 10 October 2008 (UTC)

I have to agree with this thought, as well - I don't think a single administrator should be responsible for handing out these kinds of tools. Rollback is pretty innocuous, but deletion is another matter. Shereth 23:22, 10 October 2008 (UTC)

(2xEC) I appreciate the intention behind this notion, but the systematic unbundling of all these tools seems liable to create a system that is even more excessively bureaucratic and prone to problems than the current setup to me. The creation of a tiered system has the potential to engender a sense of "rank" among Wikipedians, and this should be avoided at all costs. In any event, deciding which admin "powers" to dole out in such a fashion and deciding which ones are more "sensitive" than others is also subject to considerable debate and the ultimate decision is likely to be arbitrary. At least, with the "powers" packaged, it is a simple question of "do we trust this user with the ability to make controversial decisions" rather than trying to wade through the problem of deciding precisely which controversial abilities they are suited to. Shereth 22:56, 10 October 2008 (UTC)

Illogical captain. Open your windows, smell the fresh air, smell the coffee. And then when you've recovered your senses tell me in all honesty that there isn't already a sense of "rank" among wikipedians. --Malleus Fatuorum (talk) 23:03, 10 October 2008 (UTC)

Hah! I don't mean to imply that there isn't already a sense of rank (there most certainly is). My concern is that this will only enforce an already existing misconception and add new "ranks" for editors to have their eye on. Shereth 23:05, 10 October 2008 (UTC)

Yeah, that a fair point, although I haven't seen many boasting about their rollbacker rights. It's difficult though, to think of a less impressive chat-up line than "Would you like to come back to my my place and look through wikipedia's deleted pages?" :lol: --Malleus Fatuorum (talk) 23:13, 10 October 2008 (UTC)

Just wait till I show you my huge, long block log ... oh baby! Shereth 23:14, 10 October 2008 (UTC)

The only one I like the sound of there is the ability to edit the interface. The rest don't sound at all appealing. -- how do you turn this on 23:06, 10 October 2008 (UTC)

User pages with stacks of colourful little user boxes (in sundry styles) linking to the rights log entry for each trusted tool. Gwen Gale (talk) 23:08, 10 October 2008 (UTC)

Why this before/after business? The tools can be handed out in any order. Of course, different levels of trust (though that term is very vague) may be required for different abilities, but requiring possession of the deletion tool before the blocking one or vice versa is simply silly. Nousernamesleft (talk) 23:19, 10 October 2008 (UTC)

In my humble opinion these tools are all equally sentitive, and the way I see it is when I trust someone enough with the block button, I'd trust them with the delete button and the protection button and all the rest, hence I don't see a need to unbundle. Shereth 23:22, 10 October 2008 (UTC)

Same here. The only admin tools that I'd agree with are those that can be replicated - that is, rollback can be replicated through various scripts. Other "rights" such as IP excemption and account creation aren't, in my view, a part of the mainstream idea of what admin tools are (when was the last RFA you saw that said they'd work on account creations?), and IP excemption is simply something that admins have because they're "trustworthy", and has been split off and a good use has been made out of it. I cannot see why any non-admin would have any need for things like delete/block/protect (which are the 3 main tools most admins use). -- how do you turn this on 23:30, 10 October 2008 (UTC)

Let me suggest a different dimension. There are clearly a number of administrators who use the block button, or its threat, incompetently, but who may be quite competent in discerning consensus in an XfD. Why should all administrators have access to tools that some of them are not competent to use? --Malleus Fatuorum (talk) 23:31, 10 October 2008 (UTC)

That's a good question. If you do have an issue with a certain admin, have you tried talking to them about it, or bringing it up somewhere? Or is this an imaginary scenario? -- how do you turn this on 23:37, 10 October 2008 (UTC)

Not imaginary at all. Check my block log. --Malleus Fatuorum (talk) 23:47, 10 October 2008 (UTC)

I see. -- how do you turn this on 23:53, 10 October 2008 (UTC)

You are correct. I suppose there is a certain disconnect here between those who believe trust is the qualifier for whether someone should be made an admin (get the tools) or whether competence is the qualifier. That said, I believe the number of admins who make "incompetent" decisions in one "admin area" while shining in others is relatively low; most of us who are not comfortable with something like blocking will avoid it in favor of what we are comfortable with. In the rare cases that an otherwise trustworthy admin totally bombs it with a bad block, generally the mistake is not repeated. In short, I view the un-bundling of tools (in this context at least) as a solution in search of a problem. Shereth 23:38, 10 October 2008 (UTC)

I agree with you that further unbundling is highly unlikely. I only joined in this discussion for a bit of light relief, not under any illusion that anything would be likely to be changed by it. --Malleus Fatuorum (talk) 23:47, 10 October 2008 (UTC)

Is anything productive ever done as a result of a discussion on this page? ;-) -- how do you turn this on 23:54, 10 October 2008 (UTC)

In a word, no. This is probably one of the most sapping, demoralising, and destructive pages on wikipedia. One that I'd suggest you unwatch, as I've just done. Let the RfAs take place in secret, as custom dictates. --Malleus Fatuorum (talk) 00:06, 11 October 2008 (UTC)

I don't think there's a disconnect between trust and incompetence. The community trusts an editor to be competent with the tools. Instead of a dichotomy, they complement each other. —kurykh 23:59, 10 October 2008 (UTC)

New page

As I tried to point out above I don't think unbundling the tools is generally senseless. My major concern is that a single administrator should decide whether someone is granted the relevant flag (i.e. right). Thus, what about a new page where one requests particular flags and the community is allowed to voice their opinion (analogous to WP:RFA). The discussion there is closed after a certain period of time by an administrator either granting the flag(s) if there is an consensus for doing so or not granting the flag(s) if there is no consensus for granting them. In this way the community could decide whether someone should get a particular right. Any thoughts? —αἰτίας •discussion• 23:39, 10 October 2008 (UTC)

This came to my mind as I read the above. (ROFL @ Malleus as an aside.) In a discussion several weeks ago on a certain "BADSITE" that everyone loves to hate, I mentioned that a good alternative, in my mind, was to have different types of admins. Those that work on content and those that work with people. Content admins having (un)protect/(un)delete/view deleted rights; people admin, or whatever they may be called, (un)block/acct creation. In cases where tools from each type of admin are required, it forces teamwork, two sets of eyes on the issue and thus a little bit of oversight. So this proposal is a bit different, obviously, but it's got very similar points. So with all that said, I support such a change. I don't think this would enforce a sense of "ranks", rather I think it would be much like rollback. Surely editors, some of them, will be proud of their achievement in receiving the extra rights, but I don't think it would further divide the community. Quite opposite, actually, I think it would bring adminship back down to "no big deal" in a lot of ways, and allow users who could do a lot of help with one set of extra rights—but otherwise may not be able to pass RFA because of mistakes related to other rights that come with adminship—to do so. I further agree, however, that allowing admins the sole discretion to make the call on who should be granted what, as is done with rollback, is a bit too risky, thus an RFA-style process seems completely appropriate and will, in my opinion, benefit the project and our backlogs tremendously. Jennavecia ^(Talk) 02:42, 11 October 2008 (UTC)

Then people would have to RfA twice if they wanted to be both content and people admins? Ling.Nut ^{(talk—WP:3IAR)} 03:20, 11 October 2008 (UTC)

Well, I'm not proposing my idea. I was just commenting that it's something I'd thought about in quite a bit of detail, and I think it's a good idea to bundle them in groups. The proposed idea is different than my own and I think it's a good one. But really, I think one would have to go through the process more than once to get multiple bundles, but I think it would be a bit less intense than RFA. That would be the hope. لennavecia 15:33, 11 October 2008 (UTC)

• The problems are first that being a ah good admin for a couple of months isn't hard. While there are some exceptions most de-adminings are driven by unusal events that can either be avoided for a couple of months or are not common enough to be likely to come up (how many admins have actualy blocked another admin for example?). Second is that telling admins that there is an admin out there that they cannot entirely trust is not a good thing (obviously we know there are such admins but it helps to pretend there are not).Geni 00:21, 13 October 2008 (UTC)

While RfA is under reform...

Head over to Wikipedia:RfA Review and participate if you have not done so yet. The recommendation stage will close on 00:00 UTC on 15 October 2008. While I understand that some editors have already started putting forth proposals for RfA reform, my opinion is that the results compiled would give us a better picture on what reforms are more likely to work. I'm Mailer Diablo and I approve this message! - 01:56, 11 October 2008 (UTC)

No thank you. That effort is an exercise in misplaced priorities that has already diverted far too much of the community's time and attention from productive activities, such as training. This website has large numbers of people who would like to become administrators, a good share of whom would become good ones with proficient mentorship and preparation. Please, if you care about getting more good administrators for the site, go to any of the following places instead. Introduce yourself to some of the people there and lend a helping hand. Durova^Charge! 06:21, 11 October 2008 (UTC)

• • Category:Wikipedia administrator hopefuls (1142 names)
  • Wikipedia:List of administrator hopefuls (hundreds of potentially qualified people)
  • Wikipedia:Admin coaching (64 people waiting for help)

Actually, if you really want to engage in productive activities, please go edit the encyclopedia instead. (This is not directed at Durova, but to everyone in general.) —kurykh 06:26, 11 October 2008 (UTC)

Every site needs administrators and Wikipedia doesn't have enough of them. Creating content is important too, but if everyone followed your advice, Kurykh, and abandoned this in favor of content, then normal attrition would soon make it impossible to keep up with vandalism, CSD, and other necessary functions. Durova^Charge! 06:38, 11 October 2008 (UTC) spoken by an editor who has 152 featured content credits

I personally think that it is more productive to find some solutions for a better RfA in the long term, rather than to leave it be and then having to see through several dramas in future RfAs of these hopefuls, resulting in attrition and more resources being diverted away just to deal with these issues. - Mailer Diablo 07:07, 11 October 2008 (UTC)

I agree with Mailer. It's worthwhile and the more input the better. If you don't like the idea of it Durova, good for you, but no need to be insulting to those who do, or those who attempt to draw attention to it. لennavecia 15:50, 11 October 2008 (UTC)

The ability to block vs. "no big deal"

Discussion moved to Wikipedia:Village Pump (proposals). - jc37 23:40, 11 October 2008 (UTC)

Is there really a problem?

Lets firstly say that I do appreciate a lot of the hard work and thought-provoking discussion that has taken place on this page. However, much of it seems to be founded on the assumption that RfA is "broken", that not enough editors are choosing to take the plunge, that the pass rate is not high enough, etc etc.

I'm not so sure that there actually is a problem at the moment though.

Looking at the current Tangotango readout, there are currently five discussions in progress that are on track to pass easily. Four of them have unanimous "Support" votes. One has over a hundred supports and not one oppose. There are two further that are easily within the "pass range" on 88% (although one of those is less than a day old). There are two discussions that appear to be on track to fail, but one of those is the controversial Ironholds one, and the other is a candidate that in my opinion, was always going to struggle to get up to the threshold.

So, we currently have five candidates that will almost certainly make it, two that will probably make it, paired against one unorthodox discussion and one discussion that will probably not end in success. Discounting the Ironholds RFA, which is not being held to our usual standards and procedures and that I feel would distort the statistics, we have 7 out of 8 looking to pass, or 87%. Is an 87% pass rate really cause for concern that the process needs radical reform or that the community is somehow being unduly harsh?

(Note, this is the current version of the RfA report that I am looking at. A new entry popped up just as I was typing this in.) Lankiveil ^{(speak to me)} 07:57, 11 October 2008 (UTC).

In asking this, I'm not trying to be obtuse, but is this merely another example of the classic: "RfA is broken" vs. "RfA is not broken"? (Which even has it's own RfC : )

If it's not, would you clarify? - jc37 07:59, 11 October 2008 (UTC)

I tend to lurk on this page a fair bit, and the Ironholds RFA just got me thinking. Lots of people are pouring lots of effort into coming up with solutions to this "problem", but that all seems to be predicated on the fact that there actually is a problem. I used to think this was the case, but when I looked at the raw facts and figures, I'm no longer 100% sure that that is the case. I figure that coming back and asking the question on one of the original premises that a lot of the discussion here might generate some interesting discussion or insight from others into the whole state of affairs. As much for my own benefit as for the benefit of others =). If we can improve the process from it, then so much the better.

So, no, it's not a loaded question, and I'm genuinely curious what others think. Lankiveil ^{(speak to me)} 08:05, 11 October 2008 (UTC).

I'd point you both to WP:RREV and ask you to read through. Although the success rate is one thing, there are a number of other reasons that people have problems with the current process. Having said that, any attempt to change the process is met with it's own strong oppposition, even in the face of evidence collected at the RfA review. Personally, I'm of a mind to walk away from the RfA process until such time as there is strong agreement that change is both needed and will be backed and implemented. At the moment, it feels like I'm a lone voice and that consensus is against me. Gazimoff 08:08, 11 October 2008 (UTC)

I've been following it. My main problem is that it's been taking me longer to answer the questions than the deadline has allowed : ) - jc37 08:12, 11 October 2008 (UTC)

I have been a contributor to WP:RREV, although it seems to be a very "heavy" process that is dragging a little. And truly Gazi, how could anybody not be impressed by the amount of work and effort that you've put into that whole affair? I'd agree certainly that the process is far from perfect, but to my mind any changes to the process will be "improvements" rather than "fixes". Lankiveil ^{(speak to me)} 08:24, 11 October 2008 (UTC).

My own opinion: Dispute resolution (at least involving admins) is broken. Arbcom burnouts are rampant, case delays are in excess of four months, the community is becoming heavily factionalized with lots of wikipedia-as-a-battleground behavior. Once an admin bit is bestowed, it is virtually impossible to undo that change except in the most egregious circumstances, and that leads to widespread demoralization of the community. The overall system is broken, and if view RFA as the overall "deciding who should be an admin" process, then yes, RFA is broken.

But on the other hand, if we look at RFA just as more-narrowly define subprocess that "bestows admin rights", then no, I'd don't think I'd call RFA 'broken', per se. By and large, RFA does its job. But I do think there's plenty of room for improvement. If processes were articles, I think I'd call RFA a "Good Article", but not yet a "Featured Article", if that makes sense. It works, it does the basic job it needs to do, but i think it could do things better.

Just my two cents though. It'll be interesting to see what the comment process turns up. --Alecmconroy (talk) 08:16, 11 October 2008 (UTC)

RfA is not "broken" by any means, it still does it's job. The community though may/has lost confidence in the process, in my opinion because adminship is extremely difficult to take a way, and If the community has lost confidence in the process, then it needs to be "reformed". Just my opinion though. RockManQ ^(talk) 16:26, 11 October 2008 (UTC)

Unbundling makes sense, but only like this

It would require a small software change. A couple people have floated this idea before (I wish it were mine since I think it's clever) on both WR and outside sites. A progression like this...

• Rollbacker
• Leads to being able to turn on Semi-protection, but not turning it off
• Leads to being able to turn on Full protection, but not turning it off
• Leads to doing moves over redirects, stuff like that
• Leads to another gnome thing that needs tools for housekeeping
• A high end of deletion as the unbundling "cap", but have this one turned on by a crat instead of an admin--see below.

Stuff that would be excluded would be blocking (any sort), turning off full protection (no way, you want the Main Page and key templates unprotected?), reading deleted copies, automatic IPexempt, editing through full protection, history merges, restoring deleted copies--none of that.

No new process needed. Just a technical fix to let people set these "safer" admin tools, the ones that can cause the least damage. None of the behavioral ones like blocking. Just the bulk of the gnome ones. No reading deleted, anything like that. Thats like handing a 16 year old kid the keys to the Dodge Viper. Let him learn on the Ford Escort first. Again, NO new process need. User:Iliketoedit has like 2000 edits, 800 to WP space? No edit wars, not a scumbag? He'd have rollbacker no problem. If he can handle that, why can't he handle semi protting a page that's under vandal attack? He can handle that, why not turning on full prot if he sees an edit war? If he Iliketoedit screws up, it's as trivial as taking away rollbacker is today. So, for example, if a dedicated vandal fighter/RC guy went to the current existing request for permissions page--this hypothetical fellow doesn't want admin, he just likes RC--he could ask for the ability to turn on semi-protection to help in his RC work. Likely to abuse the tool? Benefits WP? The admins there turn it on, and off he goes to protect WP.

Our guy a month later asks for turning on Full Protection power. Has he been good with the semi? Yep, correctly used it 195 times in RC work. He gets Full access turned on. He gets into content support work a month or two later. He asks for the gnome-like tools, gets them. Another month or two later, he's done fine. Maybe he does a n00b admin mistake here or there, and just fixes them himself. Who doesn't? Everyone learning to drive that dinky little Ford HAS to jump the curb parallel parking to get good at it. I sure did on Commons; my bumped is appropriately scuffed. Everyone else's here is, too. In any event, our hypothetical fellow has had all the non-behavior and super-high-end admin tools now for 4-6 months, and has been just fine with him. Assuming he hasn't blown up at any point along the way, he'd likely be a shoe-in for RFA to get the rest (if he ran, but he may not want to or need to). And so on. Hell, some people may just want SOME of the tools to begin with. Maybe they just want to help on CSD. Or image migrations. Or RC.

Wouldn't Wikipedia as a whole benefit with an extra army of trusted gnomes with power tools running around, as opposed to just having armed soldiers? I mean, think about it: The United States Army doesn't send in someone armed to the teeth to do the maintenance on their airport runways. They send in the technical specialists. That's what unbundling and just using the existing RfP (is that what its called, Requests for Permissions?) system would do. All it requires is some technical changes. Maybe we don't need another 1000 admins a year--maybe another 200-300 a year is fine, but we do need another 1000 people that can do the trivial crap that requires a tiny bit of extra button to do. rootology (C)(T) 14:49, 11 October 2008 (UTC)

Some sense in this, but can you show a need? Is their a constant backlog at the protection requests page? And, why can't we let any auto-confirmed user move over a redirect?--Scott MacDonald (talk) 15:26, 11 October 2008 (UTC)

Move over redirect basically deletes the redirect page in order to move the article page. It's because it requires the deletion that it's limited to admins only.Gazimoff 19:02, 11 October 2008 (UTC)

More than some sense. Excellent idea, well-thought-out. Good way to split the tools: there's gnome stuff and behavioural stuff. Gnoming doesn't much need the support of the wide community, but behaviour-modification does. Absolute support for this if it comes to a !vote. Prince of Canada ^{t | c} 15:31, 11 October 2008 (UTC)

This proposal seems like quite a good one, and it good be given out at per say, Request for permissions. (Except for the last one which would be given out by crats). RockManQ ^(talk) 15:54, 11 October 2008 (UTC)

Again, I've said above; that I would like a blocking provision in the plan (no way would this be given out by Request for permissions). I guess I could compromise if have to. RockManQ ^(talk) 15:58, 11 October 2008 (UTC)

I think having this at WP:PERM is brilliant. It would also (I think?) streamline RfA... "User:Iliketoedit has had Protect and Delete access for three and five months, respectively, and has shown good judgement with both. We should grant the rest of the tools blah blah." Prince of Canada ^{t | c} 16:06, 11 October 2008 (UTC)

I agree that this kind of split makes perfect sense. However, I am not sure how WP would change if such a continuous career path was established. I would prefer it if the gnome subset of the admin tools could only be turned on and off at once. (Perhaps these tools can simply be given to all rollbackers?)

However, as a consequence of such a split I suspect that admins will spend less time on jobs that are closer to what ordinary editors do (like moving over redirects). This will reinforce the appearance that there are two major, distinct groups; an appearance that already leads to tensions between the groups. (For some of the reasons, see ingroup bias, outgroup homogeneity bias, and especially ultimate attribution error.) --Hans Adler (talk) 16:12, 11 October 2008 (UTC)

I think it would be the exact opposite effect, actually. By making admins not "have" to be the go-to guys for every stupid or silly thing that happens or needs to get done it would make 95% of what an admin is today not a big deal by not having to lock away the various non-critical tools. By doing this, admins that have gone through the vetting process of an RFA would basically be still what they are today, but would only be specifically "needed" to act as moderators of the website, since they would be the only ones with the behavioral tools. The whole point is to fix RFA while benefitting Wikipedia. Great! This idea would give access to the tools to people that can use them, for the tools that are a substantial portion of what needs doing--supporting content, and building an encyclopedia. I think everyone associates the borked up nature of RFA with dramaz and blockingz and all that. Thats not what people should be admins for. You only should be an admin to support content with the tools. Not that you're trusted to not bot-nuke half a million articles, or not block good people, or as some reward. This idea would empower admins to some degree, sure, vs the Power Users it would create. But so what? I could care less if someone gets a bit of an ego boost for passing RFA. Good for them, everyone thinks you're keen, and here, have some extra cookies and buttons. If unbundling drops the rate of RFAs by 50%--great!! As long as we get that many more people using the tools they want to use to support WP, whats the big deal? If someone wants to go the whole way later with the practice and experience, it would also make future RFAs trivial--if someone is great at using their smaller toolset, why would we expect them to not be good with the rest? And the bad people would already be filtered out, since loss of the lighter tools would be logged forever in user rights logs. It's a win-win for WP, for admins (less mundane crap to cover--admins can go back to writing articles with their free time!), for RFA (less drama, the system would pre-vet most applicants), for content (a ton more people doing things) and for the content users that want to just help on lighter gnomey admin stuff without getting into AIV, ANI, RFAR, and all that. rootology (C)(T) 16:29, 11 October 2008 (UTC)

*applause*. Verily, indeed, forsooth, etc. Prince of Canada ^{t | c} 16:41, 11 October 2008 (UTC)

No, the ability to fully protect articles is the last thing that should be given out willy-nilly. Semi-protection, maybe, but with a max duration of 4 days or whatever the "autoconfirmed" threshold is (the point where semi-protection becomes moot for bad-faith editors, while remaining an impediment to good faith editors). — CharlotteWebb 18:13, 11 October 2008 (UTC)

That was just an example, going for a linear progression on the surface, but that actual "progression" can be sorted out, and should probably be the very first bit of any discussion at all of unbundling admin powers. Which is the least possibly harmful to give out without a formal RFA? Which is on the opposite end? And then line up every current admin ability on that scale, and then determine which should only be given out with an RFA. That right there would be the probable starting point for the idea of implementing unbundling via WP:PERM. rootology (C)(T) 18:35, 11 October 2008 (UTC)

How about:

• Rollbacker / Accountcreator
• Apply (time-limited) Semi-protection, but not remove
• Other gnome tools (define?) incl. moves over redirects
• Deletion (approved by admins at WP:PERM, but implemented by crats)
• RfA for all other tools

Prince of Canada ^{t | c} 18:48, 11 October 2008 (UTC)

How about you define what "other gnome tools" are first. "Moves over redirects" is the same as deletion, since anyone could first change a page to redirect to whichever page they intended to move there (if the redirect has no history, it won't prevent a move for anyone). That tool can easily fuck up several pages in a way that's not easy to fix, so you'd want to have more than one person to look at a candidate before handing it out. And the same obviously goes for normal deletion. -Bobet 19:01, 11 October 2008 (UTC)

Alrighty then Joe, for a second I thought you were implying that "being able to turn on Full protection, but not turning it off" (i.e. set-and-forget) was less abusable than "doing moves over redirects, stuff like that" which would of course, be complete bullshit .

If I was going to do something like this I'd start with viewing deleted pages, which would enable the user to non-ignorantly participate in a deletion review (and to know when one should be filed) without being able to independently "wheel-war" over it.

As for redirects that's a serious software flaw to begin with. If A gets moved to B, and then to C, some bot will come along and change A so it redirects to C instead of B, making it impossible for C to ever be moved back to A (I didn't look closely but I think this what happened with Dr. Strangelove a few turns ago). If every revision of A is a redirect, anybody should be able to move a page over it, if they have the ability to move pages at all.

I don't have much opinion about whether balkanizing admin rights is a good idea in principle. On the surface it seems a bit like some game where you learn new spells but there might be some format that functions other than as a complete joke. I might write up a spin-off of Jennavecia's idea and post it in a little while. I think she's onto something. — CharlotteWebb 19:26, 11 October 2008 (UTC)

Let me think more about this. Was there a full list anywhere of all the bells and whistles we get for admin tools on MediaWiki? I don't think I've ever seen one either here, Commons or Meta. rootology (C)(T) 19:32, 11 October 2008 (UTC)

mw:Manual:User rights#List of Permissions. — CharlotteWebb 19:40, 11 October 2008 (UTC)

Move over redirect falls in the same category as delete, as it requires the ability to delete (the redirect) before the move can take place. In addition, any unbundling of the tools would be a development or bug request. I think it's been mentioned before that the developers aren't keen on implementing this. Gazimoff 19:02, 11 October 2008 (UTC)

The first time I ever felt I could use the admin tools was when I needed to edit a fully protected template--not when I needed to protect something (in my experience posts to RFPP are handled quickly). However, if done maliciously, this can cause harm. Thus, it is required to establish the community's trust before being given that ability. Has anyone else had the same experience? That is: you needed to edit something fully protected; not the other way around? What I am saying is this: there is a need for more trusted editors to be given more user rights; but, those needs are not in page protection (rather they're in unprotection, which can be damaging on templates). Lazulilasher (talk) 19:12, 11 October 2008 (UTC)

A little late, but in reply to the comment above: I, as a Gnomish non-admin, have definitely found several occasions when I wanted to edit a fully-protected page - usually to make a minor edit like a spelling or formatting correction. I like the idea of splitting up the admin tools, for the various reasons explained above - and if it does ever happen, I'd hope that the ability to edit protected pages (not templates!) would be one of the first rights made available to non-admins. It's not really any more dangerous than rollback, in my view, and if it could be removed as easily as that is, would pose no danger at all. Terraxos (talk) 00:26, 12 October 2008 (UTC)

We have enough problems with admins continuing to edit through protection without extending the power.--Scott MacDonald (talk) 00:38, 12 October 2008 (UTC)

Right, I was hoping to illustrate how complicated the matter is. If one wants to do edit a protected template -- regardless if the edit is non-controversial -- then he must gain the community's support (I wanted to add a new taskforce, and the change was approved by the WikiProject). This is because the possible danger is large. Concomitantly, we need to encourage those who are not likely to cause any harm in these areas to be able to work in them. However, there is a great degree of trust involved even when it comes to page protection; how does one demonstrate his trustworthiness? Lazulilasher (talk) 01:54, 12 October 2008 (UTC)

If unbundling were to happen, really the only way to gauge this would be to try to trust people first with applying protection correctly, and the other abilities. That was why I focused in on the ability to place protections as being "earlier" in the whole scheme. Putting down a protection is a utility thing. A non/lower admin putting down protection, say a full, to stop an edit war, is going to lock himself out of it too. It's like the idea Misza13 told me to allow any admin to desysop an out of control admin--at the cost of both their sysop bits to immediately stop trouble, with the Arbcom then judging the situation. A Power User/Trusted User placing a full protection would have to rely on good judgement, and someone roguely placing bad protections will be stripped of that (and probably other special tools) pretty darn quick. The damage they can cause BY protecting is minimal and a decent starting judge of resonsibility. rootology (C)(T) 02:17, 12 October 2008 (UTC)

Note: I've proposed for the unbundling of protection in the past (it would be great for certain bots, and experienced template coders), but learned that to do what we would propose, it would need to be entirely re-written. So (for now), it's apparently a "wontfix". Compare to most other user-rights which can now be bundled/unbundled depending on consensus. - jc37 00:40, 12 October 2008 (UTC)

New Idea

As far as I can tell, there has been a bunch of proposals on this page. All of them are garnering different supports, opposes, and comments. What if we compromised and combined elements of two different "plans" into one? RockManQ ^(talk) 16:01, 11 October 2008 (UTC)

Let's take a look:

• Many Editors have expressed the concern that people have lost confidence in the process
• Many Editors have expressed that it needs to be changed
• Some Many Editors have expressed an interest in unbundling the tools
• What, however, is a good way to bring all of these plans together?

Thoughts? RockManQ ^(talk) 16:32, 11 October 2008 (UTC)

You will find that although "many" have done all of these, there will not actually be a consensus that RfA needs changed, and ever if there is there will certainly not be any consensus as to how or indeed why. Some people think it is too lenient, some people that it is too harsh. I've enjoyed the discussion here, but realistically sometimes we just need to admit defeat.--Scott MacDonald (talk) 19:42, 11 October 2008 (UTC)

I guess so :( RockManQ ^(talk) 20:18, 11 October 2008 (UTC)

I betcha it works out in the end. Change is slow, but it'll come. Unbundling has already started, and I think it looks like it's going to continue. And that is one of many forms of "trial adminship", on instead of time-based, it's tool-based.

As for removing tool access, Wikipedia is an ecosystem. Empower more users to use tools and you will automatically empower SOME to abuse tools. More tool abuse will lead to a better ability to revoke tool access. Pushing through molasses, to be sure, but the project will get there, just wait and see.

Every time I've lost my faith and concluded the Wiki process just can't handle something, its proven me wrong. Just wait-- the project will get a handle on this-- just maybe not today. --Alecmconroy (talk) 00:44, 12 October 2008 (UTC)

Yep, Wikipedia is constantly evolving. RockManQ ^(talk) 05:18, 12 October 2008 (UTC)

The only concerns though is that the process would be quite slow, as many admins might see this as a big "fuck you" or something, and would fight it tooth and nail (see the many admins opposing here). RockManQ ^(talk) 06:09, 12 October 2008 (UTC)

Just like the real world, reforms require a lot of political will to see any form of tangible success. - Mailer Diablo 10:24, 12 October 2008 (UTC)

Yep, most certainly. RockManQ ^(talk) 02:04, 13 October 2008 (UTC)

Question period preceding vote, perhaps

One thing that I think would help this process would be to have a question/answer period prior to the vote-taking. Have a one-week interval where people are examining the record, asking questions about what they think are important aspects, and allowing the candidate to answer. Once that's been over, then open up the voting floodgates. One thing that is apparent to me going through this is that the comments of the early voters set the tone of the whole RFA, and if there was a chance for discussion prior to people being able to vote, those votes might be a bit more informed.—Kww(talk) 14:51, 4 October 2008 (UTC)

I Support that idea :P. Ironholds 15:06, 4 October 2008 (UTC)

As do I. I think that the discussion might help a lot. iMatthew (talk) 15:08, 4 October 2008 (UTC)

Fully agree with the concept. Not sure the discussion needs a full week and then the voting another week; I think we could compress it all into 1 to 1.5 weeks. --barneca (talk) 15:12, 4 October 2008 (UTC) addendum: I spammed this link up higher in the thread, but discussion seemed to have already past that thread by, and it meshes well with Kww's post. I promise this is the last time I'll link it in this discussion, and I'll do it in tiny font to avoid being too aggresive with my linkspamming: User:Barneca/The Problem With RFA --barneca (talk) 15:22, 4 October 2008 (UTC)

I agree entirely, with Barneca's caveats. The candidate should retain the right to withdraw at any time of course, so a bruising discussion period need not be compounded by a punitive vote. the skomorokh 15:15, 4 October 2008 (UTC)

I agree that two weeks is definitely too long. I would suggest 48 hours for this question session before going into the week-long voting period. Candidate must retain right to withdraw during question period. I also think that WP:NOTNOW and WP:SNOW should still be options during the question period. Useight (talk) 16:04, 4 October 2008 (UTC)

Approve of that idea. How exactly would this be put into practice? Would we have to go through a whole policy change proposal, include it in the RfA review, what? Ironholds 16:05, 4 October 2008 (UTC)

The RFA Review is something unofficial, we would not be required to slip this idea into it; plus the Review is possibly too far along to try to add another idea. As for getting this idea put into practice, it's going to be near impossible to get consensus (but we're off to a good start) because Wikipedia fossilized a couple years ago. Change is extraordinarily difficult. Useight (talk) 16:09, 4 October 2008 (UTC)

We need not adopt a defeatist attitude: perhaps editors are so exasperated with RfA that they may be prepared to try a slightly different approach. If the Matt Brit RfA was allowed, perhaps a guinea pig candidate can be persuaded to try this method? the skomorokh 16:30, 4 October 2008 (UTC)

If there is consensus to change, we implement. We do not need to consult the author of RfA review, crats etc..etc. I actually think this is a splendid idea. Although, the only downfall I see is just an overabundance of questions causing the head of the candidate to spin mercilessly. We've all agreed in the past that the number of questions far exceed the necessary amount. What would this encourage? Wisdom89 _{(T / ^C)} 16:38, 4 October 2008 (UTC)

Editors discussing with each other rather than asking the candidate, less vote-orientated assessment, measured consideration of the candidates merits instead of the Rfa focusing overwhlemingly on issues raised in the first few opposes—to name a few. My previous comment was to suggest we did not need to wait for consensus-arrived-at-through-tortuous-month-long-discussion, but could just alter the format of an upcoming Rfa, politely asking editors to withhold from voting for two days, and going from there. Run it up the flagpole and see who salutes. the skomorokh 16:46, 4 October 2008 (UTC)

(edit conflict) (response to Wisdom) Yeah, and that's the issue with this proposal. It would only add more questions, which I as a candidate enjoyed answering, but the number of questions does often become excessive. With this implementation, two days of questions (plus probably some more added during the normal week-long period), would only increase the number of questions. If we only allow questions to be asked and answered, I see this as possibly a solution looking for a problem, but if we used the rarely-used Discussion section during that 48 hour period it should decrease the amount of "badgering" that occurs. Useight (talk) 16:51, 4 October 2008 (UTC)

Longer discussion need not necessarily mean more questions; I would imagine with the "badgering" stigma removed, nominators and others would be free to advocate and research on the candidate's behalf. the skomorokh 16:57, 4 October 2008 (UTC)

←Would questions still be allowed to continue during the voting phase or cease straight away? The latter would be a bad idea.. —Cyclonenim (talk · contribs · email) 16:50, 4 October 2008 (UTC)

Questions definitely need to be permitted during the voting phase. Useight (talk) 16:53, 4 October 2008 (UTC)

I agree with Useight. -- RyRy (talk) 16:56, 4 October 2008 (UTC)

I'd be prepared to act as a test case to see what happens. After 3 previous RfA's It's safe to say I dont follow the "I didn't pass? I quit!!" attitude, so it should work out alright. I was thinking of applying anyway, but if I can apply+help out future candidates then it's a win-win for everyone. Ironholds 17:00, 4 October 2008 (UTC)

This also increases the likelihood of more nitpicky trap questions where there is only one right answer..or designed to produce the wrong one. Wisdom89 _{(T / ^C)} 17:00, 4 October 2008 (UTC)

If people would be OK with me going ahead under this format, I guess we'll find out! Some questions will always be two wrong answers; open to recall comes to mind, for example. Ironholds 17:02, 4 October 2008 (UTC)

This has been discussed before. Limiting any part of the phase to less than a week prevents users from participating, because some only log on during the weekend, for example. 48 hours before voting isn't really much of a time-frame either. That said, if such a change were made, the questions should be required to be specific to the candidate, no general "go look at another RFA for the answer" questions. No trick questions and no stupid "why are bananas yellow", time wasting questions. Admins and 'crats should be prepared to trim out the stupidity. Jennavecia ^(Talk) 17:20, 4 October 2008 (UTC)

Ok, well just so it's clear. I support such a change to the process. Wisdom89 _{(T / ^C)} 17:28, 4 October 2008 (UTC)

If you can work something out i'll be the test-case. We could try a week of questions to avoid people missing out, and if that turns out to be too long with mine but the process works we can trim it. Think of me as the local Redshirt.Ironholds 17:30, 4 October 2008 (UTC)

I'm not sure I agree with someone involved in the change being a test subject. It'd be better to institute the change upon an uninvolved editor to prevent any bias that could possibly arise. Additionally, would you want the omgz drama over at your RfA? It's gonna be hell to keep track of. —Cyclonenim (talk · contribs · email) 17:48, 4 October 2008 (UTC)

I'm fine with any involved drama, as long as its made clear that "I dont like the new format" isn't a valid reason for oppose. My involvement with the change so far has just been "hey, i'd like to be stuck up on the firing range, but I dont think format X is the best way I should be shot". Ironholds 17:55, 4 October 2008 (UTC)

(e/c) On the contrary, Cyclonenim, I don't think we should spring a new format on a candidate that wasn't 100% behind the idea. I fully support an experiment, and salute Ironholds' offer. I think a reasonable Crat would respect a slightly longer than normal RFA once, as an experiment, to see what happens, without requiring a full-blown multi-month "official" discussion at RFA Review or somewhere. Since questions and discussion would still be allowed during voting, there's no need to have 1 full week of discussion, then a longer voting time period for discussion. How about 5 days of discussion only, followed by 5 days of voting? Only 3 days longer than a normal RFA, and we'll know at the end whether it should be longer or shorter.

Also, fully 100% agree with Useight, I see the possible benefit more in everyone using the discussion section than in the asking of questions.

Frankly, if Ironholds is really up for this, that he should just create an RFA, delete the Support and Oppose and Neutral sections, make a note of the altered format in some way so people aren't confused, and as Wisdom says, raise it up the flagpole and see who salutes. --barneca (talk) 18:03, 4 October 2008 (UTC)

This sounds interesting; We have tried this over at ArbCom elections before, so we can roughly see the likely outcome. Note that however, this will also mean the longer the question period before the vote, the more screen time for anyone putting controversial questions forth. (meaning the excessive/ridiculous questions problem is going to magnified a number of times). - Mailer Diablo 18:06, 4 October 2008 (UTC)

I'll get started on this right away. Give me 24 hours to contact some users who've been interested in nomming me before and some time to answer the basic three questions and format it all and i'll put it up. Ironholds 18:09, 4 October 2008 (UTC)

I applaud you for volunteering to be the "guinea pig" for this proposal, and all the best of luck! - Mailer Diablo 18:14, 4 October 2008 (UTC)

I have to echo MD - that epitomizes WP:BOLD if anything ever could. I'd be shivering in my RfA boots to be the first to test a new process. I'm still eager to see how this goes. Wisdom89 _{(T / ^C)} 18:28, 4 October 2008 (UTC)

One thing I'm not clear on, how long for the questions and how long for the voting? Not everyone supports 48h/1 week and I think two weeks (1 week each) is too long. — Rlevse • Talk • 19:58, 4 October 2008 (UTC)

For my one i'm sticking with three days; long enough for anyone getting back from a weekend away, short enough that it isnt overly onerous. As you said, a week is too long, not everyone likes a 48/h period, so i've picked something that should appease both parties. Ironholds 20:02, 4 October 2008 (UTC)

I just threw 48 hours out there as an idea; 72 hours sounds fine to me, too. Useight (talk) 20:51, 4 October 2008 (UTC)

72 hours should more be than sufficient. Enough time to catch regular Wikipedians. Wisdom89 _{(T / ^C)} 20:53, 4 October 2008 (UTC)

The problem with 72 hours, is the same one that has existed everytime we've brought this discussion up in the past. 72 hours is NOT sufficeint for the users who only have access on certain days---such as weekend. I can't think of a single process on Wikipedia that isn't on a 7 day process, for this reason.---Balloonman ^PoppaBalloon 15:15, 5 October 2008 (UTC)

Apart from AFDs, MFDs, DRV, etc etc. -- how do you turn this on 15:35, 5 October 2008 (UTC)

Just to add my voice to the area, I support this initiative, as it's one I've thrown around a couple of times before, but doesn't seem to have caught on. Having said that, it's not without it's own drawbacks, and I completely agree with the idea of having a few test cases before making a decision one way or another.

I like the idea because it deliberately breaks the process into two phases - one where questions are asked, evidence is gathered and the pros and cons of each candidate are weighed up and analysed. This provides an ample forum for people to discuss concepts while encouraging them to check back and refresh their opinions, something that the current process does not really suit. It also neatly matches the concerns raised at the Debate and Election sections of RfA review by providing a structure more suitable for civil debate and discussion, while still allowing the community to support or oppose based on their own view of the weights of the statements. I do feel that, as in Arbcom cases, that we should consider drawing a line under the debate before moving on to the election except in exceptional circumstances (crat discresion), otherwise I am concerned that the debate will turn into a free-for-all if the vote is not heading in the direction groups of contributors feel it should. While I feel everyone should have an opportunity to air their views, I also feel that once the window has closed that should be it.

I do have a couple of concerns, though. Firstly, I think that one test is not enough - I'd hope to see a minimum of three, including one candidate who has been through the process before (such as Ironholds) and one candidate who has not previously submitted to RfA. Secondly, after going through the process myself, I know that a week can feel like an agonising eternity in itself. While I completely agree with Jennavecia that we need to allow as many people as possible the chance to participate, I also feel that we have to be fair on the candidate as well and not draw out the process too much. I would suggest a two-week period to start with, followed by a review at the end of the test cases. It may be that either the debate or the election phase could then be shortened without disenfranchising contributors or materially affecting the outcome, especially if the test RfA phases were organised to switch over the course of a weekend.

I realise that this is a bit of a mind-dump. I'm happy to work on a framework, templates etc in short timescales as I suddenly have a free weekend, so please shout if there's anything I can add that would be of use. Many thanks, Gazimoff 21:12, 4 October 2008 (UTC)

• I'll support this change. Protonk (talk) 23:29, 4 October 2008 (UTC)

STRONG OPPOSE Considering that I think the question period is A) the part of the RfA process people hate the most and B) the most meaningless part, having a question/answer period would only make a bad process worse! Candidates should show their policy knowledge and application of policy via their edits, not some essay exam. I don't really want to pass people who are book smart, but lack any street smarts/expertise.---Balloonman ^PoppaBalloon 14:58, 5 October 2008 (UTC)

Balloonman, maybe you should cool it a bit? Writing your opinion in all caps bold text doesn't make it any better than anyone elses opinion; in fact, it would probably make people think worse of you. -- how do you turn this on 15:04, 5 October 2008 (UTC)

ER, the only part that is in bold caps is the !vote where caps/boding is perfectly acceptable.---Balloonman ^PoppaBalloon 15:18, 5 October 2008 (UTC) PS, I have to wonder if your attempt to pose my rationale as making people "think worse of" me is really in response to this discussion. This is the only response to your attempt to divert the oppose. If you want to comment about the problem of the proposal, I'll be happy to respond, but your comment is an attempt to make it personal. The proposal only makes a bad process worse.---Balloonman ^PoppaBalloon 15:23, 5 October 2008 (UTC)

Bolding is, capitalizing isn't. I'm simply asking you to cool it, since capitalizing is equivelent to shouting on the internet. (And no, this has nothing to do with the "advice" on my talk). -- how do you turn this on 15:34, 5 October 2008 (UTC)

In the body yes, but not in the !vote. Perhaps if you'd been around longer you might realize this.---Balloonman ^PoppaBalloon 15:42, 5 October 2008 (UTC)

Please don't patronize me. I've not once seen anyone capitalize their vote. It's pointless, looks incredibly childish, and it looks like shouting. All perfect ways to heat up discussion completely unnecessarily. -- how do you turn this on 15:48, 5 October 2008 (UTC)

Again, in the !vote, it is accepted to emphasize one's strength of argument... it happens all the time. In fact, there are RfA's currently in the work with capitalized bolded OPPOSES. As for patronizing, I apologize, but I don't appreciate your patronizing me.---Balloonman ^PoppaBalloon 15:58, 5 October 2008 (UTC)

Back on topic: Balloonman, thanks for your input on the idea. I've volunteered for this as a "will it work/wont it work" test, although it is a serious request for adminship. Street smarts and so on will still come up during the Q&A period; for example, if user A asks "what would you do in situation X?" and there is street evidence of the candidate, in situation X, doing something completely different, that can come up, either in the opposes or in a "I've found a difference similar to situation X. You seem to have handled it very differently to the way you said you would in your reply to user A's question. Why is this?" question. For those of you who are fans of the idea, I've finished my RfA page. Suggestions on posting: Now? when all the planets are in alignment? Wait, as Gazimoff said, for more candidates for this process variant? Ironholds 19:03, 5 October 2008 (UTC)

I wish you luck... but I still don't like the of increasing the time focusing on the part of RfA that I think is the biggest waste of time. I don't think questions should be asked unless they are specific to the individual in question, but you are inviting a quiz---which is, IMO, entirely the wrong way to "fix" the process. I hope people prove me wrong.---Balloonman ^PoppaBalloon 04:06, 6 October 2008 (UTC)

If this turns into a set of canned questions that are the same with every candidate, it's useless, especially if it's more of the An AFD just reached a consensus that the United States is a non-notable country, and all articles pertaining to its citizens should be deleted from Wikipedia kind of thing. I just know that in my own case, I wish that I had had a chance to be asked questions and respond to some of the impressions my record made on people before I was staring at a rack of 8 "Oppose" !votes all referencing the same diff.—Kww(talk) 04:18, 6 October 2008 (UTC)

It should be interesting at least. The plan, from my talk with Gazimoff, is to have three RfA's (no experiment is valid with one participant), one after another, with a 6-day gap in between each one which will be used for post-mortems. The idea is that, after the first one is over, we have 6 days to debate over what went right, what went wrong, what needs to be changed, and so on, and we implement those changes for the second one. We do the same after the second one for the third and, should the process work, it will hopefully be implemented in a more widespread fashion. To allow wikipedians who might be away in the weekend to take part in both halves the question period will start on Tuesday, switching to the voting period on Sunday morning and continuing until the end of Wednesday. At the moment we need two more candidates for the further RfA's, preferably one who has participated in the "old-style" process and one green candidate; I'm open to any volunteers, although I appreciate this is the internet equivalent of asking for people to act as targets for "this nice new firing range we've got". Ironholds 04:57, 6 October 2008 (UTC)

Right, the "prototype" as it were is Now out. I guess that would make this the beta, heh. Ironholds 17:58, 7 October 2008 (UTC)

Where do I comment on the new RfA format? Axl ¤ [Talk] 08:57, 12 October 2008 (UTC)

After reading the entire RfA and this discussion, I absolutely oppose this change. The optional questions section has been and always will be a relatively irrelevant part of the RfA process. Its where people with axes to grind go to try to get someone they don't want as an admin to make a mistake and then flog them to death with it. It's where I suspect a great deal of people go to make themselves look important in the RfA process by asking no end of truly inane questions whose answers can be surmised just by properly vetting the candidate. All in all, this looks like yet another idea brought to you by Wikiproject:Bureaucracy. (Yeah, I know it doesn't exist...at least not officially). Trusilver 07:48, 13 October 2008 (UTC)

New proposal - provisional adminship

Originally proposed here.

Where an RfA would normally fail, but the candidate has 60%+ [a clear/comfortable majority in] support, and where in the opinion of the closing bureaucrat the opposition is largely due to uncertainly about the candidate, rather than informed opposition, then the bureaucrat may offer the candidate "provisional promotion". If the candidate accepts provisional promotion, then during the provisional period (determined by the bureaucrat), any bureaucrat may instruct a steward to desysop if at least two bureaucrats agree the candidate has proven to be unfit. Candidate appeals are made to the community through a new RfA.

This would hopefully lead to the promotion of more administrators, while reducing the risk of errors. Jehochman ^Talk 08:17, 10 October 2008 (UTC)

This could be helpful, although I wouldn't want to see a bureaucrat being opened up to criticism for having taken a discretionary risk, either way. So, perhaps "provisional adminship" could be granted following any majority (but not consensus) support in an RfA, say for 3 months, duirng which any bureaucrat could tell a steward to desysop if at least two bureaucrats agree the candidate has been shown to be unfit. Also, participants in RfA discussions could give their support only to provisional adminship. Gwen Gale (talk) 08:50, 10 October 2008 (UTC)

So all I have to do is lay low for three months, then I'm in? Why not just lower the hurdle to 60% and introduce more robust means for desysopping? Franamax (talk) 08:54, 10 October 2008 (UTC)

A few might lay low, but even that's a sign of self-discipline and forward-minded thinking. Gwen Gale (talk) 09:04, 10 October 2008 (UTC)

No, it's a natural reaction to being on probation. See my comment to PoC below. Franamax (talk) 09:10, 10 October 2008 (UTC)

Gwen's got a great idea, but how about we just make all adminships provisional for three months? Probation ends automagically if there are no issues raised by crats about the use of tools. The well-supported admins will have no difficulty, and it gives the less-supported time to prove that they can use the tools responsibly and deserve the trust they're given. That way it's the same rule for everyone. By analogy, here in Ontario there's a 90-day probationary period after starting any new job; your employer may terminate at will and does not have to show cause. Same concept here; it's no hardship on good workers, and lets questionable workers know that they need to be on their toes. Thoughts?

(ec) Franamax raises an excellent point, though. But.. Archtransit aside, are there really many admins who plan that far ahead to abuse the tools? Prince of Canada ^{t | c} 08:58, 10 October 2008 (UTC)

Interesting. You should update the article about at-will employment to explain what goes on in Canada (see template {{Globalize/USA}}). I always thought that high probability of being sacked for no reason was a predominantly Amerikan thing. — CharlotteWebb 12:28, 10 October 2008 (UTC)

AT is a classic case, but not the best example. Any "provisional" admin would be extra careful, just like your probationary workers. Once their rights are vested though, they revert to form. In general, abusers don't plan to abuse, they just do. Franamax (talk) 09:08, 10 October 2008 (UTC)

Or, another thought, what about making admin coaching/mentoring/school/training/fill-in-the-blank a mandatory part of becoming an admin, whether before RfA or during the (potential) probationary period? Guarantees that all admins begin with the same understanding of policy/procedure, which then adds to a robust desysop procedure if they blow it (honest mistakes aside, of course). Prince of Canada ^{t | c} 09:04, 10 October 2008 (UTC)

Continues the problem of admin-coached gaming of the system. Good admins aren't made; they're born. You can't teach character and maturity. Those who have character and maturity will learn to use the tools correctly on their own, without coaching requirements. SandyGeorgia (Talk) 09:11, 10 October 2008 (UTC)

How would you feel--assuming a probationary period were implemented--about passing through some sort of mentoring/exam/whatever after a successful RfA, as one of the conditions for probation ending? So, one negative condition: don't be a twerp! one positive condition: pass this exam/go through mentoring/whatever. I'm mostly just noodling, I don't even know what I think about it. Prince of Canada ^{t | c} 09:50, 10 October 2008 (UTC)

To avoid excessive rocking of the procedural boat, it might be good not to automatically impose provisional on all candidates-- but if we had a provisional status, don't you imagine it would be sufficiently useful that we might start !voting to steer all qualified admin candidates that way? I think such a change might be very useful. --Alecmconroy (talk) 11:07, 10 October 2008 (UTC)

While I'm a fan of coaching, I don't suggest coaching for everyone... in fact, I've rejected a fair number of coachees... I have to see a legitimate need (that isn't personality based---eg policy knowledge/experience) before I accept a coachee. To make it a requirement, would be A) clog the system significantly as there aren't enough people to do it and B) wouldn't add any real value as some coaches don't really do anything for their coachee. Yes, I support coaching, but I'm also one of it's biggest critics.---Balloonman ^PoppaBalloon 14:12, 10 October 2008 (UTC)

Seriously, 60 % ? And I would suggest that we simply solve problems but not sysopping anyone below 80% to begin with. Goodness, we already have enough problems breeding problems. Maybe 75%, and provisional for a year and I'd support. (POC, yes; not necessarily plan ahead to abuse, but behaved long enough to convince the community past issues had been overcome.) SandyGeorgia (Talk) 09:02, 10 October 2008 (UTC)

Understood--I hadn't thought of that angle. My guess would be that users who took that route would quickly start misbehaving again as soon as the sysop bit is added--they no longer have anything to worry about, as it's much harder to desysop than to get it in the first place. I think; am I off base here? Prince of Canada ^{t | c} 09:04, 10 October 2008 (UTC)

Yes, that's the issue. I think we've have a lot less problems if we *raised* the bar, not lowered it. Set provisions on anyone between 75 and 85%, no sysops below 75, or just stop sysopping at all below 85%. We don't need more admins; we need more good admins. SandyGeorgia (Talk) 09:08, 10 October 2008 (UTC)

I wonder. It would be interesting to ask whether the problem admins are ones that got lower support percentages. Personally I wonder that there may not be a correlation, and thus lowering the bar (with some safeguards) may not see a reduction in quality. You make an assumption, it would be interesting to see whether it hold up under scrutiny.--Scott MacDonald (talk) 09:17, 10 October 2008 (UTC)

Well, I wouldn't be saying it if I didn't know of such cases :-). Of course, that doesn't preclude admins who pass with 90 to 95% also causing problems. SandyGeorgia (Talk) 09:22, 10 October 2008 (UTC)

Sure, but we "know of some cases" of RfAs with 100% causing problems and we don't extrapolate. I know of some cases of an admin being Scottish and that admin being a problem - but that doesn't mean there's a correlation. Post hoc, ergo propter hoc?--Scott MacDonald (talk) 09:52, 10 October 2008 (UTC)

What bar are you raising though? By strict attention to percentage, you're condoning supermajority voting - can we discard that !vote idea then? The real problem instead lies in how to weight the votes. !supports with the rationale of "sure, why not" and !opposes with the rationale of "self-noms are prima facie evidence..." are equally problematic. They shouldn't necessarily have full weight, neither should they have zero weight. The problem lies in accurately weighting the comments. To that extent, I support the intent of this proposal, and thereby contradict myself within 2 kilobytes of text. The probationary timeframe aspects bother me - but maybe I'm OK with a year. :) Franamax (talk) 09:24, 10 October 2008 (UTC)

I already put in a different proposal for how to address that several threads up (that 'crats should have far more discretion in passing noms, as at FAC, where vote stacking doesn't fly). SandyGeorgia (Talk) 09:42, 10 October 2008 (UTC)

I Oppose this. Though I understand it's well-meant. My main thought is that we need more quality of admins, not more quantity of admins. - jc37 09:06, 10 October 2008 (UTC)

Let's discuss, rather than jump in with bolded and polarising "support and oppose", eh?--Scott MacDonald (talk) 09:11, 10 October 2008 (UTC)

I'm happy to discuss. Though I'll note that this discussion has been posted now to the WP:VP, so, I may have been under the mistaken apprehension that a poll was beginning. - jc37 09:30, 10 October 2008 (UTC)

Is the admin selection process so broken that it needs to be changed? Truth be told, I'm starting to think of some unintended outcomes which might get stirred up by this and I'm always wary of ever-more complicated bureaucratic processes. Gwen Gale (talk) 09:12, 10 October 2008 (UTC)

More debundling is the way to go; find ways for non-admins to be more useful, such as rollback. Or maybe even yearly provisionals, but not as low as 60%. SandyGeorgia (Talk) 09:24, 10 October 2008 (UTC)

• The proposal looks like a really bad idea for many reasons. First, this would take away too much power from the community and give it to the bureaucrats. The decision for final promotion is, right now, made primarily based on the consensus as expressed by the community in the RfA vote. Under this proposal, if I read it correctly, a provisional admin would automatically become a permanent one at the end of the provisional period, without a new RfA. I find that to be objectionable as a basic principle. A crat's decision that the original lack of consensus was based on "uncertainty" rather than "informed opposition" would be fairly subjective anyway, and one would not know, without a new RfA, if this uncertainty had been dissipated, if new issues had arisen or if the community really changed its mind. A decision to promote to adminship is and should be based on the expression of community's trust, not on subjective decisions of a single crat. If we do require a new confirmation RfA for provisional admins, then the entire idea of a provisional adminship becomes largely redundant. In 4-6 months after an unsuccessful RfA people usually reapply anyway and more often than not pass, if the issues had been non-major and had been addressed. Second, there is too much arbitrariness and uncertainty in this idea in terms of a crat deciding when a provisional adminship is warranted and how long a provisional period should be. With something like an RfA more clear and uniform criteria and rules are necessary, to maintain the credibility of the entire process and its fairness. Third, I think that the proposal is misguided in what it actually tries to accomplish. The current RfA system, in spite of various gripes, actually does a pretty good job in terms of the final results that it produces: by and large, people who should be approved get approved and people who should not be approved are not approved. The main complaints regarding the RfA system are because of its percieved acrimonious nature, and this proposal is not likely to change that. If anything, there will be more complaints and wrangling about who did or did not get provisional adminship and why, and, knowing that in effect the RfA bar had been lowered, many people are likely to oppose in RfA more frequently and vigorously to counteract the "grade inflation" this proposal introduces. Overall, a bad idea. Nsk92 (talk) 09:26, 10 October 2008 (UTC)
  • It doesn't "take power away from the community" - because anyone the community's majority opposed would not ever be sysopped. What it does is reduce the power of a minority, who are always fearful of taking a calculated risk, blocking the will of the rest of the community.--Scott MacDonald (talk) 09:32, 10 October 2008 (UTC)
    • Yes, it does, by effectively lowering the bar too much (whereas in my opinion the bar actually needs to be raised a bit). To say that only those who did not get a majority (or, say, 60%) in an RfA fall into a non-discretionary zone who will not be offered adminship (permanent or provisional), certainly diminishes the role of community's vote. An RfA where 40% of the votes were opposes is not exactly a ringing expression of community's trust. This is especially true since, psychologically, it is always harder to cast an oppose vote anyway. Nsk92 (talk) 09:48, 10 October 2008 (UTC)

I think the 60% number should be dropped. It may not sound so bad from a certain point of view until you remember 50% represents "the least possible consensus". 50% is a complete hung-jury, with the community providing no information one way or the other. 60% is pretty minimal support. And since we don't decide RFAs based on percentages anyway, it's sort of a dead end anyway.

The other concern I would have is that this could "lower the bar" to the point where large amounts of oppose votes would be granted very little weight. That's not the role I would see for provisional admins. Provisional adminship to me is ideal in a situation where we would says "Come back later-- everything looks good so far, but we just don't have enough information yet". Provisional adminship is NOT a good idea when we have someone who's both strongly supported and strongly opposed. The proposal recognizes this, but I think its implementation might be troublesome.

The other concern I'm hearing, which I share, is that the current proposal gives / requires too much discretion on the part of the 'crats. Consensus is hard enough to assess-- how would they go about assessing opposition "largely due to uncertainly about the candidate, rather than informed opposition".

But for all the concerns, I think there's a great kernel of an idea here, and basic idea is a good one. Right now, the RFA process has to approximate precognition-- foreseeing future admin performance based only on past editor performance. Being able to, in some cases, decide adminship based on actual use of the admin tools seems like a great idea.

So, here's what I'd propose. Drop the 60% and leave things at "consensus". Drop leaving it up to the 'crats to assess the "nature" of the opposition, and instead let the community speak for itself. Let editors weigh-in by either saying "Oppose", "Support", or "Support provisional adminship". If there's a consensus for full adminship, just go with that. But if there's consensus of provisional adminship, we could offer that to a candidate. This way, we don't "lower the bar" by tinkering with percentages, and we don't ask the 'crats to do more than in RFAs than what they already do-- assess community consensus.

Proposing this in general below: --Alecmconroy (talk) 10:42, 10 October 2008 (UTC)

As a data point, Commons recently did something somewhat like this. See the candidacy of SterkeBak which had some controversy around a flurry of late opposes, some by users with little or no community standing. The 'crats talked about what to do here Commons:Administrators/Requests and votes/SterkeBak/Bureaucrats discussion, with community input as well Commons talk:Administrators/Requests and votes/SterkeBak/Bureaucrats discussion and the decision taken was to promote provisionally, with a 60 day period in which any three 'crats could call for a rerun of the RfA. ++Lar: t/c 20:05, 10 October 2008 (UTC)

Proposal: Basic Framework for Provisional Adminship

So, what it seems like I'm hearing, and what I'm feeling, is that there is some merit to this line of thought, but maybe some problems with this particular implementation. For the purposes of discussion, I might suggest the more vague "basic points":

• 1. If an RFA would pass normally, adminship will be granted as per usual.
• 2. Alternatively, if there exists some level of community support for 'provisional adminship', then provisional adminship will be granted.
• 3. If there isn't sufficient community support for either, no adminship will be granted.
• 4. Provisional Adminship can be revoked through some as yet unspecified process if the provisional admin isn't working out.
• 5. Provisional Adminship can be upgraded to full adminship through some as yet unspecified process if the provisional admin is working out.

How does that sound, as a basic framework? --Alecmconroy (talk) 10:23, 10 October 2008 (UTC)

You are hearing this wrong. As I said above, the entire proposal is a really bad idea. Nsk92 (talk) 10:36, 10 October 2008 (UTC)

Sorry, I should have maybe made a new section-- I wasn't summarizing you personally, I was kinda trying to distill the entire section of commentary down into a single clause. The idea has some support but plenty of concerns too. --Alecmconroy (talk) 10:45, 10 October 2008 (UTC)

You've just added two processes to what was my original simple idea. We need to start trusting crats not inventing processes.--Scott MacDonald (talk) 11:15, 10 October 2008 (UTC)

If your idea will fly with others, I'll support it-- I think it would be an improvement over the present RFA process. I only made this version of the proposal because it seemed like people were getting mired down in the more-controversial specifics, such as percentage levels and 'Crat having to assess the "nature of opposition". But i bet when we boil your idea down to its essentials, the fundamentals of your proposal are going to meet with wide support. --Alecmconroy (talk) 11:25, 10 October 2008 (UTC)

I suspect I'm going to end up agreeing with Franamax on this one; I doubt this will be the best way to "fix" RfA. Just lower the bar from 3:1 to 2:1, and make it easier to desysop people who, in hindsight, shouldn't have been promoted, whether they passed at 2.01:1 or 100:1. However, because I'm just guessing here, and in an attempt to break the inertial logjam around here, I'm all in favor of trying a provisional adminship a couple of times, seeing what happens, and make a decision after we have something, anything, to base our opinions on. Come on people, don't be afraid to try something new. It would be dumb to have a discussion on whether to make this the way to do things from now on, because people will never, never, never agree to such a big change with no particular reason to think it would be an improvement. I wouldn't agree to such a change either. But it would also be dumb to not be willing to give things a try, so we can stop guessing whether something has some merit or not. --barneca (talk) 12:30, 10 October 2008 (UTC)

If you mean above where Franamax says: "Why not just lower the hurdle to 60% and introduce more robust means for desysopping?" I totally agree-- the project would be better served if adminship were easier to bestow and easier to remove. But such changes are very hard to get started-- the project's inertia to resist change can be quite substantial. (and it is an important issue, so we certainly don't want to be hasty).

Adminship as "No big deal" and as "Easy come, easy go" is a double-sided coin. Provisional Adminship is a way to try to make it easier to bestow adminship. For the other side of the coin, see Wikipedia talk:Removing administrator rights/Proposal. Since provisional adminship is a 100% positive-- giving us admins we wouldn't ordinarily have and not jeopardizing the bits of any existing admins, it will probably be easiest to implement. --Alecmconroy (talk) 12:58, 10 October 2008 (UTC)

If we had a real means of desyspopping, then I would support lowering the bar to 75%, not 60%, and applying provisional to anyone between 75 and 85%. The bar is already too low with 'crats passing people in the 75% range, and I'll fight 60% no matter what. There is no reason to sysop someone who doesn't have the trust of 40% of the community, even if provisional. SandyGeorgia (Talk) 18:32, 10 October 2008 (UTC)

pXk's idea for provisional adminship

• 1. If an RFA would pass normally, (75%) adminship will be granted as per usual.
• 2. If at least 60% but less than 75% are in support, a provision run off vote will occur lasting a pre-determined amount of time (let's say 5 days)
• 3. If there isn't at least 60% in the provisional vote, the RfA is considered failed. If there is, the provisional period of 6 months begins.
• 4. Provisional Adminship can be revoked through some as yet unspecified process if the provisional admin isn't working out.
• 5. Provisional Adminship can be upgraded to full adminship through some as yet unspecified process if the provisional admin is working out or be considered a full admin once the period is over.. PXK ^T /_C 15:39, 10 October 2008 (UTC)

No. That's process running wild.--Scott MacDonald (talk) 15:54, 10 October 2008 (UTC)

I think this would work nicely. I might suspect that in practice this would devolve into people just stating upfront which of the three categories they are in-- full support, provisional support, or total oppose. You could do an "instant runoff", or you could go ahead and wait an additional five days for people to consider provisional by itself-- either way's good with me. --Alecmconroy (talk) 16:51, 10 October 2008 (UTC)

Simplify..

1. >X% adminship will be granted as per usual, with 90 day probation for all
2. Crats can desysop during probation per original suggestion, with community review to follow desysopping (similar to 'please review my block' on AN/ANI)
3. We need a better process for recalling full admins as necessary.

Prince of Canada ^{t | c} 17:03, 10 October 2008 (UTC)

Percentages

What's with all these percentages all of a sudden? Do we really want to move away from consensus based decisions and turn RfA completely into a vote? I'm not so sure that taking away bureaucrat discretion is a good thing at all, which is what we'd be doing with the proposals so far. We need some ability to move away from set percentages given the controversial RfA's that sometimes come up, along with the RfA's where there's probable socking and other serious concerns. Ryan Postlethwaite^{See the mess I've created or let's have banter} 22:21, 10 October 2008 (UTC)

• RFA is already a vote. Deal with it :( Stifle (talk) 08:34, 13 October 2008 (UTC)