Wikipedia talk:Requests for rollback/Archive 4

Archives

 * See also: June 2005 and January 2008 on the wikien-l mailing list

Archiving
Is there a bot that will be able to archive requests? I'm sure we're going to be getting a slew of requests over the next few days. --  tariq abjotu  23:41, 9 January 2008 (UTC)
 * ST47 was going to work on one, I'll get him onto it. Meanwhile, I created Requests for rollback/Archive 1 - feel free to manually archive at will.  Ry an P os tl et hw ai te  23:44, 9 January 2008 (UTC)
 * I've had to block a User:RFRBot because it was not approved. --Rschen7754 (T C) 00:24, 10 January 2008 (UTC)
 * It's a clone of another of ST47's bots so it's ok, I think it got unblocked.  Ry an P os tl et hw ai te  00:29, 10 January 2008 (UTC)

RFR
Does rfr need protection? --Rschen7754 (T C) 23:46, 9 January 2008 (UTC)
 * Yup, likewise the header for the main page.  Ry an P os tl et hw ai te  00:29, 10 January 2008 (UTC)
 * I've fully-protected the template and move-protected RfR itself to match other pages like ANI and RFPP. Acalamari 00:30, 10 January 2008 (UTC)

User rights conflicts
Why does the "give rollback" page not warn you if there's a conflict, or has this not been implemented yet? I'm asking because this just happened. Acalamari 00:33, 10 January 2008 (UTC)
 * It's not the end of the world, it happens with protection - I'll have a word with the devs though.  Ry an P os tl et hw ai te  00:35, 10 January 2008 (UTC)
 * Oh, I know it happens with protections (happened to me earlier). :) It's just that it would be handy for a conflict alert, like with blocks. Acalamari 00:37, 10 January 2008 (UTC)
 * I've fixed this in 29526, in a slightly different way. If you're changing someone's userrights to the same thing, it will just silently fail to log it. —Simetrical (talk • contribs) 03:17, 10 January 2008 (UTC)

Request requirements
Is it required that all requests be made here? If, for example, a user were to request rollback on an admin's talk page, should that admin send them here or are we free to simply review the request and give/deny it as the situation warrants? I'm assuming this is an RfA type situation, and the requests must be made here. - auburn pilot   talk  00:40, 10 January 2008 (UTC)
 * That's an interesting point, and something I'd be willing to look at. It would be a good idea to require requests to come through here.  Ry an P os tl et hw ai te  00:46, 10 January 2008 (UTC)
 * I was wondering the same thing.   jj137  ♠ 01:41, 10 January 2008 (UTC)
 * Does anyone have an answer here? I am very curious to know.   jj137  ♠ 02:12, 10 January 2008 (UTC)
 * I'm all up for discussion on the matter - I say yes.  Ry an P os tl et hw ai te  02:13, 10 January 2008 (UTC)
 * Personally, I don't think we should require them to come through here, its not a big thing, so having it able to be handed out whenever should not be a big problem. Viridae Talk 02:14, 10 January 2008 (UTC)
 * (ec) I don't really think they should have to either. I mean, it isn't as big a deal as something like RfA; the advantage to coming here though is it will probably be fulfilled much quicker.   jj137  ♠ 02:16, 10 January 2008 (UTC)
 * I'm up for discussion as well, by the way.   jj137  ♠ 02:17, 10 January 2008 (UTC)
 * I anticipate this page running smoothest if it functions like WP:RFPP. I.e. You'll get the fastest response here, but there's no problem with asking a specific admin directly as long as you're not forum shopping. It works well. – Steel 02:20, 10 January 2008 (UTC)
 * My thoughts exactly. Viridae Talk 02:23, 10 January 2008 (UTC)


 * Please, if the right is granted outside of a formal request, please ensure the granting summary is adequate and informative. I've been reviewing some of the log summaries and some are, well, poor, and some are good. Some examples: "vandalism fighter"; "Dude, this is *Jimbo*!"; "active vandal fighter in an ignored area"; "as there are no policy yet..." (this was granting User:Willy on wheels rollback rights - self-reverted a few minutes later); "requested for RC patrolling"; "Frequent vandal fighter/rc patroller"; "promotions"; "look ma, i'm changing jimbo's permissions!" (sigh or :-) depending on mood); "'zilla use with great care"; "liek zomg vandalbot!"; "asked me for it via IRC"; and "trollback" - those last two worry me... Carcharoth (talk) 02:30, 10 January 2008 (UTC)
 * Trollback was me - it was a typo :( Viridae Talk 02:31, 10 January 2008 (UTC)


 * I think Steel359 has the right idea, this is a place to get admin attention for this issue, but not the only way. 1 != 2 02:49, 10 January 2008 (UTC)
 * Yes; if someone requests it on your talk page, there's no point in having them bring it here if you are going to grant it anyway.   jj137  ♠ 02:50, 10 January 2008 (UTC)

I think the final choice here should be left up to the deciding admin.  jj137  ♠ 03:46, 10 January 2008 (UTC)
 * Someone changed my rights from "rollbacker to rollbacker." This is because I am "respectable."--The Fat Man Who Never Came Back (talk) 05:10, 10 January 2008 (UTC)

Totally confused...
Who decided there was consensus to implement this? I looked over Non-administrator rollback and the mailing list and I haven't figure out what group decided the vote was a consensus to implement the policy. I'm assuming it was the dev's, but can anyone point me to where the justification for considering the vote a consensus to implement? Apparently I'm blind as I can't find it :P.  Justin  chat 00:41, 10 January 2008 (UTC)
 * If you're blind, so am I, because I don't see where the implementation was decided either. As far as I can tell, the discussion regarding how this should be implemented has barely started on Non-administrator rollback. - auburn pilot   talk  00:44, 10 January 2008 (UTC)
 * The devs looked over discussion for the poll and decided there was consensus - they didn't justidy themselves, they just make decisions and implement or don't.  Ry an P os tl et hw ai te  00:45, 10 January 2008 (UTC)
 * "The devs looked over discussion for the poll and decided there was consensus." That's the generous way of looking at it. --kingboyk (talk) 00:53, 10 January 2008 (UTC)

Well ultimately the decision on adding features lies with the developers. Presumably on review of the relevant discussions, the developer in question decided the feature should be added to the English Wikipedia. I think its taken everyone a little by surprise (the user rights log is a mess already) but I think everyone's doing a good job of creating a decent infrastructure for granting this right so far. WjBscribe 00:47, 10 January 2008 (UTC)


 * The last time this policy was discussed the vote was 216/108 (66.66% approved) and failed. This one is 304/151 (66.81% approved) and passed. Correct me if I'm wrong but don't WP:RFA's typically require 70-75% for approval? This consensus stuff is far too complicated for my brain.  Justin  chat 00:59, 10 January 2008 (UTC)
 * Of note is also I guess that last time it was not really technically possible yet to grant rollback per user, because the developers had not implemented it yet. Therefore last time the developers had nothing "to grant" to the wikipedia users and now that they can, they might have decided that the arguments of the oppose-voters are not strong enough to negate the obvious positive effects this may have on the overall project. I'm taking a wild guess, but I'm pretty sure the reasoning would have been something like that. --TheDJ (talk • contribs) 01:17, 10 January 2008 (UTC)


 * Glad to know I am not the only one who missed something  Alexfusco5  01:11, 10 January 2008 (UTC)

Meh, many people forget that Wikipedia is, above all, a technocracy; the development gods play a pivotal role in all technical decisions. Just hope they don't have Special:Makepillarofsalt implemented yet. :)  krimpet ✽  01:23, 10 January 2008 (UTC)


 * The devs just turned the tool on. It is up to the community to draft policy as to how it is used. 1 != 2 02:52, 10 January 2008 (UTC)

I've added the disputed tag. I disagree with this being added, I disagree with how its been added, and I disagree there was consensus. -Halo (talk) 02:55, 10 January 2008 (UTC)
 * There's more comments on the Administrators%27_noticeboard -Halo (talk) 02:58, 10 January 2008 (UTC)

Consensus was not gained for this policy
See discussion at Wikipedia talk:Non-administrator rollback. Doc glasgow commented elsewhere:


 * Certainly people were determined to force this through no matter what. Six day polls called at Christmas holidays. The poll is closed without warning. We are told that the oppose arguments made no sense. And 65% reached in a hurry is declared a consensus. This is madness and the whole notion of community, discussion and consensus has been thrown out the window. And why should the judgement of a dev which is flawed be a "done deal"?

In no way can this be considered "policy". — Hex    (❝  ?!  ❞)   01:40, 10 January 2008 (UTC)


 * Agreed. At the very least we shouldn't be putting a watchlist notice about this up, and this page should reflect that rollblack status may be taken away, once this is corrected. -- Ned Scott 01:52, 10 January 2008 (UTC)

Semi-protection
Would this page benefit from semiprotection? As a non-autoconfirmed user would be unlikely to get rollback... --Rschen7754 (T C) 01:22, 10 January 2008 (UTC)
 * I've already sprotected it. It is also fully protected from moves.  Nakon  01:24, 10 January 2008 (UTC)

Process for review
So no process for review? Just "decline" and that's it, ever? Nakon said "maybe" on my request and when I answered his concerns, someone else came along and had declined it and I get a bot note on my talk page saying, "sorry!".. so not even a review of my explanation to Nakon's "maybe" ? Geesh. It wasn't even there long enough to be read before the bot removed it. -- ALLSTAR  echo 01:28, 10 January 2008 (UTC)


 * No does not equal never on Wikipedia, just come back in a month, explain you've been previously turned down and we'll look at the situation again. Nick (talk) 01:31, 10 January 2008 (UTC)
 * I just think it's not a good process when a deciding admin leaves a "maybe" with concerns and I'm not allowed to address those concerns.. for a month apparently. -- ALLSTAR  echo 01:34, 10 January 2008 (UTC)
 * You could try politely asking the administrator that turned you down for specific concerns and present your side of the story to the administrator who may decide to grant you the rollback tool. There are other rollback tools out there, this is just a slightly faster and less server intensive tool, it's in no way a reward or status symbol, so there's no reason to be annoyed at having to wait for a month. Nick (talk) 01:42, 10 January 2008 (UTC)
 * You were blocked a little over a month ago for edit warring... Viridae Talk 01:44, 10 January 2008 (UTC)
 * 2 months. November 2007, and it was the only edit warring block I've ever had in all of my history on WP, where I had been trying to keep inaccurate info out of the Houston Nutt article because he had not been officially named the new football coach at University of Mississippi and people kept putting in the article that he was the coach. -- ALLSTAR  echo 01:49, 10 January 2008 (UTC)
 * Actually, a little over 1 - november 30th. (ie 1 day of december) Viridae Talk 01:51, 10 January 2008 (UTC)

At any rate, I'm not looking at this as a reward or status symbol. I do use Twinkle and quite efficiently. I just think there should be some sort of review process. We have Deletion Review, RfC, etc. and there should be one for this as well. Thanks. -- ALLSTAR  echo 01:56, 10 January 2008 (UTC)
 * Requests for rollback review. Daniel (talk) 01:56, 10 January 2008 (UTC)

Thanks. I don't see that anywhere on the main page. -- ALLSTAR  echo 02:13, 10 January 2008 (UTC)
 * Any formal setup (template) for the Review, or do we just post a comment requesting review? El Greco(talk) 01:59, 10 January 2008 (UTC)
 * No idea. I didn't create it or have anything to do with it... Daniel (talk) 02:01, 10 January 2008 (UTC)

To avoid too many pages, perhaps we should just create a single Rollback Noticeboard to replace both Requests for rollback review and Requests for rollback removal and to handle most other discussion on the issue. NoSeptember 02:06, 10 January 2008 (UTC)
 * Sounds good to me. We could add those to the bottom of the current Requests for rollback page? ie
 * Rollback +
 * Rollback -
 * Rollback review
 * Viridae Talk 02:10, 10 January 2008 (UTC)


 * Can I address my request for review on the review page, or wait until this stuff gets sorted out? El Greco(talk) 02:19, 10 January 2008 (UTC)
 * I think in your case El Greco, you should just submit another request because all it's showing on Requests for rollback/Denied is:
 * 
 * It's got yours lumped in with User:Welshleprechaun's request, with no reasons why you were denied or anything. Maybe you didn't request it with the right template, I don't know.. -- <strong style="color:#fff;background:#DC143C;border:1px solid #000">ALLSTAR <strong style="color:#FFF;background:#0F4D92;border:1px solid #000"> echo 02:32, 10 January 2008 (UTC)
 * I fixed it. It was due to the bot changeoverDaniel (talk) 02:33, 10 January 2008 (UTC)


 * Daniel beat me to it. The bot was a little fast. El Greco(talk) 02:34, 10 January 2008 (UTC)
 * Ahh, so you were denied. For a 3RR block 5 months ago?? Where is the policy that says users can't get rollback if they have "insert whatever here"? I'd be curious to see that policy. -- <strong style="color:#fff;background:#DC143C;border:1px solid #000">ALLSTAR <strong style="color:#FFF;background:#0F4D92;border:1px solid #000"> echo 02:41, 10 January 2008 (UTC)
 * Currently the guidelines are "administrative discretion", hence why there's a process for review if you feel that this discretion is applied incorrectly. Daniel (talk) 02:42, 10 January 2008 (UTC)
 * Well, I posted my request for review at the link Daniel post above. El Greco(talk) 02:45, 10 January 2008 (UTC)

I have removed Requests for rollback review as this is far too new to be considering reviewing everything. Review boards to fix problematic review boards are well known not to work well - couple that with these reviews being of the *user*, and it will be a drama magnet. I suggest that people trust the outcomes initially, try to improve the request/removal review processes, and report any major problems to ANI. John Vandenberg (talk) 03:36, 10 January 2008 (UTC)
 * So no process for reviewing grievances or answering admins concerns in relation to this circus other than adding to an already backlogged ANI? -- <strong style="color:#fff;background:#DC143C;border:1px solid #000">ALLSTAR <strong style="color:#FFF;background:#0F4D92;border:1px solid #000"> echo 04:08, 10 January 2008 (UTC)
 * Processes develop when the right process has become established over time. Establishing process for something that has barely left the gates is asking everyone to be a process wonk.  We dont need yet another official venue that only deals with a narrowly defined problem, as the panel will be a narrow segment of the community who like to review users and probably have a strong POV about the way the shiny new toy should be managed - that results in very poor results if real independent review is required.  Ideally, if rollback is given to a non-admin and someone is concerned about it, they should AGF to begin with, and watch for it being used inappropriately.  If you see a user doing something that looks problematic, ask the user about it.  If that fails to address the core of the problem, talk to the admin who granted the permission, or take it to User conduct RfC.  If all else fails, take it to ANI, or escalate to Arbcom.  John Vandenberg (talk) 09:00, 11 January 2008 (UTC)

"I'll have a word with the devs though"
Ryan, I have a list of things that could do with developer attention. You said up above: "I'll have a word with the devs though". Would you be able to have a word with them about other things as well? (In case my point isn't clear, it is extremely annoying when developers, even though they are, like us, volunteers, are more likely to go ahead with things that are possible, instead of developing things that aren't yet possible - in other words, it is not really a matter of what the community wants, but what developers have found the time to work on). Carcharoth (talk) 01:34, 10 January 2008 (UTC)

Basic issues
So what happens when two administrators disagree? --bainer (talk) 01:49, 10 January 2008 (UTC)
 * One asks the other to reconsider? Viridae Talk 01:52, 10 January 2008 (UTC)
 * The universe explodes. Just kidding. In reality, they work it out without any namecalling or stress. Let's hope that can be done here... Master of Puppets   Call me MoP! ☺  01:52, 10 January 2008 (UTC)
 * If they can't work it out, or the user wishes to appeal, WP:ANI would seem to be the place to get admin consensus to grant or deny the rights just as it is for disputes over blocks and other admin actions. NoSeptember  01:54, 10 January 2008 (UTC)

See User talk:Secret for your answer. Viridae Talk 01:57, 10 January 2008 (UTC)

Same thing that happens when admins disagree about other admin tools, we seek a higher consensus or if we revert war each other we get desysoped. 1 != 2 02:54, 10 January 2008 (UTC)

User rights log
I think (correct me if I'm wrong) that the user rights log on en-wikipedia mostly consisted of granting of sysop rights before this process was implemented. What we now have is lots of "rollbacker" rights granting swamping the sysop ones. I can see an advantage in being able to get a simple log of the rollback rights granting activity (who is granting the tools and who is being granted the tools). Is it possible to filter the user rights log by "rollback"? I'm aware that you can get a list of rollbackers from Special:Listusers, but that is different to what I'm asking for, which is sort of Special:Log (type=rights) (ie. this), but limiting the results according to what right was granted (ie. right=rollback). Would this be feasible? Could someone "have a word with the devs"? Maybe we could have a vote on whether this sort of software change needs doing or not? Carcharoth (talk) 02:17, 10 January 2008 (UTC)
 * Good call. You --> Bugzilla and/or IRC, please. – Mike . lifeguard  &#124; @en.wb 02:29, 10 January 2008 (UTC)
 * With only about 30 sysop promotions a month, most will be rollback rights changes. A word search on "sysop" should let you jump to admin promotions. NoSeptember  02:33, 10 January 2008 (UTC)
 * Not if you are looking at the entire log. Could we note down somewhere the exact date of the first rollback request, just so people can divide the log at that point (or their copies of that log). Carcharoth (talk) 02:39, 10 January 2008 (UTC)
 * Bugzilla and IRC are black boxes for me. Can the devs whistle up something so I can request things like this without using those processes? Like, oh, I don't know, the way requests for rollback allows people to avoid the black boxes of scripts and TWINKLE and AWB? My point is that those who know the right devs to talk to find it much easier to get their ideas even looked at! There should be a "request for bugzilla" noticeboard for those who would like to communicate their ideas to the developers, but don't want to register yet another account, or hang out at bugzilla. The developers I have the most respect for are those who (openly) come over to Wikipedia and take part in discussions and provide useful information, as opposed to those who (sometimes unintentionally) build up barriers between themselves and the community they are writing the software for (at least I think that's what they are doing, as opposed to writing the software because they like writing software). Carcharoth (talk) 02:39, 10 January 2008 (UTC)

I've chatted a bit with Simetrical: very hard to do, but a client-side script could do it (according to Splarka). I've added 12571, but I have no high hopes. I image someone will do a script to get a similar effect. – Mike . lifeguard  &#124; @en.wb 03:27, 10 January 2008 (UTC)


 * Are you looking for this → Special:Listusers/rollbacker, or something different? -- Ned Scott 03:44, 10 January 2008 (UTC)
 * Wait, that's not a log.. never mind. -- Ned Scott 03:44, 10 January 2008 (UTC)

There's a script for everything :)   jj137  ♠ 03:45, 10 January 2008 (UTC)

Request time
I really want to have like a 10-15 min time-limit set for everyone in this page, so we could closely look at their edits, to check if there are fit for a powerful tool like rollback. Any thoughts Secret account 02:37, 10 January 2008 (UTC)
 * I agree fully. Daniel (talk) 02:39, 10 January 2008 (UTC)
 * But what about the requests being granted outside this process? Carcharoth (talk) 02:40, 10 January 2008 (UTC)
 * They shouldn't really be being granted outside this to be honest. But I favour a 10 minues waiting time.  Ry an P os tl et hw ai te  02:41, 10 January 2008 (UTC)
 * If an admin trusts the user to not abuse the tool, I don't see why there should be a "formal" request here. It's already in the logs and it can be easily removed by any other admin. This shouldn't be such a big deal.  Nakon  02:42, 10 January 2008 (UTC)
 * I like the way people are playing with Jimbo's permissions. :-) Carcharoth (talk) 02:43, 10 January 2008 (UTC)
 * "It can be easily removed" - what, after a five day discussion to prevent it being called "wheel warring"? Daniel (talk) 02:44, 10 January 2008 (UTC)
 * Or the admin could just talk it over with the other admin on their talk page. Nakon  02:44, 10 January 2008 (UTC)
 * I added the 10 minutes to the main page, I doubt anyone would reject. Secret account 02:45, 10 January 2008 (UTC)
 * I object. I understand you want people to take their time, but there's no reason to take five minutes checking the person's history and then another five twiddling my thumbs. If it's a foregone conclusion, it's a foregone conclusion. --  tariq abjotu  02:48, 10 January 2008 (UTC)
 * There needs to be a notice that says there will be some delay. Some people I know are trustworthy while others I need to do more research on them. Acalamari 02:51, 10 January 2008 (UTC)
 * "A powerful tool like rollback." Yeah... okay... if you say so. It really doesn't take that long to take a reasonably close look at someone's block log, contribution history, and AN/ANI posts. But, I would concede that for some of the requests, I'm not sure how the admin had time to even do that. --  tariq abjotu  02:45, 10 January 2008 (UTC)
 * It takes me longer to load the edit count page than the time its' taking to approve most requests here. Daniel (talk) 02:46, 10 January 2008 (UTC)
 * I feel that I can make a reasonable judgment in a few moments. Block log? Check. Sufficient # of contribs? Check. Contribs that suggest efforts to revert vandalism? Check. Those are my criteria. --Merovingian (T, C) 02:50, 10 January 2008 (UTC)
 * But check the vandalism, if it's obviously edits in good faith, and it was tagged as rvv, then it becomes an major issue. Secret account 02:56, 10 January 2008 (UTC)
 * Obviously, but I highly doubt we get vandals that sneaky often enough. Also, there should be plenty of other edits to check in order to confirm overall good-faith editing. --Merovingian (T, C) 03:00, 10 January 2008 (UTC)

If an admin makes a foolish decision call them on it, but insisting they take X amount of time when they may very well already know the person seems like rule cruft to me. 1 != 2 02:55, 10 January 2008 (UTC)
 * (ec x2) That's about the same criteria I use. I agree with 1! = 2; there is no point in waiting. There are a couple tools I use (this one, specifically) to quickly gather an edit count, by the way.   jj137  ♠ 02:57, 10 January 2008 (UTC)
 * Also I think two users should accept the request, rather than one, already, me and a couple of others complained about one editor, who's rollback was given a min after request. User:LaraLove removed his rollback. Thanks Secret account 03:03, 10 January 2008 (UTC)

Not-mean-spirited sarcasm: I like the blinking note on Special:Userrights now. Nice touch, seriously. --Merovingian (T, C) 03:08, 10 January 2008 (UTC)
 * It seems it takes me about three minutes to check user and user talk pages, most recent contributions, the noticeboard search feature, the block log and the edit count. Of course, if something was amiss it'd take me longer, but three seems to be the going rate. Daniel (talk) 03:15, 10 January 2008 (UTC)
 * It's not that I disagree that there should be some sort of note, just poking fun at good ol' . --Merovingian (T, C) 03:17, 10 January 2008 (UTC)

Comments
At the rate we're going, any editor who puts his name on this list will get the rollback button. A total of 8 requests have been denied out of 84. 90% approval is going to make this new access level as abundant as Twinkle itself. Ripe for abuse, if you ask me. Someone jumped the gun here; We need to either 1) shut down this process immediately until we define some sort of stable criteria or 2) slow down the promotions. There's no deadline to get this all sorted out, so we can take our time. Sean William @ 03:42, 10 January 2008 (UTC)
 * Slow down the promotions, I would even suggest having a minimum of a few hours until a request is granted.  Ry an P os tl et hw ai te  03:44, 10 January 2008 (UTC)
 * Totally agree with slowing down the promotions — even waiting one hour is plenty for any concerns to be aired and discussed, and if it needs further discussion we can postpone making a decision for as long as needed. Daniel (talk) 03:45, 10 January 2008 (UTC)
 * A minimum of one hour I'd suggest, probably worth noting that in bold in the header.  Ry an P os tl et hw ai te  03:46, 10 January 2008 (UTC)


 * I don't know, but the rate of new requests is slowing. --Merovingian (T, C) 03:46, 10 January 2008 (UTC)

Oh yeah, while we're on the point of slowing down promotions, can we also slow down the archiving so that admins can double-check and review whether the grants/denials of the tools is proper? — Kurykh  03:47, 10 January 2008 (UTC)
 * I agree: I'm glad I'm not the only one who thought the archiving was too fast. Acalamari 03:48, 10 January 2008 (UTC)


 * Is there a real difference as long as there's a "paper" trail? --Merovingian (T, C) 03:50, 10 January 2008 (UTC)
 * Yes. Discussion is faster on the main page. Posting probable objections on an archive page isn't a place where you might get a response. — Kurykh  03:52, 10 January 2008 (UTC)
 * Well of course you wouldn't open up discussion on an archive page. You'd take it to ANI, no?  --Merovingian (T, C) 03:54, 10 January 2008 (UTC)
 * Well, if you just slow down the archiving, you won't have to further bloat ANI with such matters, right? — Kurykh  03:56, 10 January 2008 (UTC)
 * Okay, that's true, ANI has enough to deal with. --Merovingian (T, C) 04:02, 10 January 2008 (UTC)
 * There's no need to rush things. Look at RFPP, which is something more important than this, and more pertinent to the encyclopedia. You don't see a rush to archive. — Kurykh  03:57, 10 January 2008 (UTC)
 * I've added it into the header that we should leave requests for one hour.  Ry an P os tl et hw ai te  03:51, 10 January 2008 (UTC)
 * An hour really seems unnecessarily long. I agree there should be some kind of delay, but an hour will simply cause a large backlog with little benefit. - auburn pilot's   sock  03:55, 10 January 2008 (UTC)
 * I agree, an hour is too long, backlogs are a concern, also I want it that we could reject users without waiting an hour. Secret account 03:58, 10 January 2008 (UTC)
 * Don't you think that you're basically writing policy as your going along is a sign that this policy wasn't well thought out? Why 1 hour? Why hasn't the community decided instead of you arbitrarily changing it? -Halo (talk) 04:00, 10 January 2008 (UTC)
 * It can be changed, very easily - let's come to a consensus for a time frame, I'm really not fussed.  Ry an P os tl et hw ai te  04:01, 10 January 2008 (UTC)

I've boldly changed it to fifteen minutes, with it clearly worded to allow for longer if concerns are raised (to ensure the applicant time to respond and have a discussion). Comments appreciated. Daniel (talk) 04:02, 10 January 2008 (UTC)
 * I believe this is fair. Acalamari 04:05, 10 January 2008 (UTC)
 * Most of the promotions, with the exception of a couple I don't have an issue with Secret account 04:06, 10 January 2008 (UTC)

Question
I’m not quite sure about something. If I was to be denied rollback because I didn’t give an edit summary on a few reversions, how much “proof” should I give to let you know that I won’t make that mistake again in order to have said rollback functions? The pages are vandalized quite a bit, and I apologize for the edit summary issue, and I promise I won’t do it again, but if I don’t have it, it would take me longer. Isn’t there a way for admins to remove the function if I didn’t keep up to my word? When can I re-apply (for a lack of better words) for the rollback function? BlackPearl14 Pirate Lord-ess 04:03, 10 January 2008 (UTC)
 * You can "reapply" as soon or as late as you like, although administrators may not take too favourably to sooner as it may be considered admin shopping. I'm sure a fortnight of solid work followed by a reapplication citing your efforts to address the concerns would be ample. Daniel (talk) 04:06, 10 January 2008 (UTC)


 * I'll say a couple of weeks Secret account 04:07, 10 January 2008 (UTC)

Thank you both so much for taking the time to help me out with this one! I didn’t think that doing it right away made sense, but you’ve really helped me out a ton! Again, thanks! I will be sure to cite my work :) BlackPearl14 Pirate Lord-ess 04:08, 10 January 2008 (UTC)

It might be nice to have this issue to be fleshed out on more than a case-by-case basis. --Merovingian (T, C) 04:09, 10 January 2008 (UTC)

List of requirements
Maybe a list of requirements would be a good idea? It was frowned upon, but in retrorespect would probably be a good idea. I propose;
 * 500 edits
 * A clean block log in the past 6 months
 * No sign of edit warring in the past 3 months
 * Then, admin discretion if the meet these

Thoughts?  Ry an P os tl et hw ai te  04:12, 10 January 2008 (UTC)


 * Adding requirements after this has already been granted (and still is being granted as I type this) doesn't make sense. Would everybody who has already been granted the right going to be reviewed now, to be sure they meet the requirements? - Rjd0060 (talk) 04:19, 10 January 2008 (UTC)
 * I think most people would still have the tool.  Ry an P os tl et hw ai te  04:20, 10 January 2008 (UTC)
 * I have no issue with the 500 edits requirement, but I'd drop "A clean block log in the past 6 months" and "No sign of edit warring in the past 3 months" for a less stringent and subjective "No blocks for edit warring and/or 3RR violations with the preceding 3 months". A lot can happen on this site in a month; 6 of them can be an eternity. - auburn pilot's   sock  04:23, 10 January 2008 (UTC)
 * you, sounds better.  Ry an P os tl et hw ai te  04:24, 10 January 2008 (UTC)
 * I agree with the requirements, also add no evidence of misusing twinkle, undo, etc to the requirements Secret account 04:23, 10 January 2008 (UTC)

I don't know, I already fight alot of vandalism with popups and a slow connection (bad combo). Since I only have 200 edits, my rollback button would be removed. What about the checking the contribs thing that was proposed originally? With the 500 edits thing, some really good intentioned "new users" would be snubbed. Like I said above, why not grant based on track record? Burner0718 (talk) 04:40, 10 January 2008 (UTC)


 * I think we can use a bit more discretion than that, the no block in 6 months seems a bit odd. A block can occur for a variety of reasons, and I think it reasonable that an admin look at those blocks and make a decision. I also don't like X number of edits, there need only be enough edits to decide what type of behavior the user is prone to and that will vary from user and admin. I do agree completely that it should not be given to those with a history of edit warring. 1 != 2 04:57, 10 January 2008 (UTC)


 * I dont think it is such a bad thing to have a minimum number of edits. Nearly every "powerful" tool has a minimum number of edits requirements, like WP:AWB.  Its not so much to prove they can hit the edit button, but the fact that they've had some time to learn the rules, and the way things work.  6 months might be a little long on the blocks though.  (Note: I am one of the "hundred or so" that got through before this closed).  -- Shakata Ga Nai  Talk 05:04, 10 January 2008 (UTC)

On a unrelated note. Might you consider me trustworthy of rollback.  Marlith  T / C  05:10, 10 January 2008 (UTC)

Page protection?
Wouldn't it be wiser to fully protect the page until all people present here agree about the process at least. -- FayssalF  -  Wiki me up®  04:14, 10 January 2008 (UTC)
 * What problems are there about the current process? — Kurykh  04:15, 10 January 2008 (UTC)
 * (ec)That thought crossed my mind, but I figured I'd be too rogue to do so. Sean William @ 04:16, 10 January 2008 (UTC)
 * Kurykh, the process reeks of complete and total chaos. Sean William @ 04:16, 10 January 2008 (UTC)
 * "What problems are there about the current process?" Seriously? There are so many differing opinions on how this should be done, it's staggering. - auburn pilot's   sock  04:18, 10 January 2008 (UTC)

I'm very much for this proposal, but yes, I agree to protection. We clearly need to work out some issues. --Merovingian (T, C) 04:20, 10 January 2008 (UTC)


 * We may expect someone who would not be granted the tool arguing about someone else who would get it under different conditions/circumstances. More than 100 users have got it already and the next 100 would expect the same rules. Add to that the fact that some people are still talking about presenting this to the ArbCom while others have not contacted the devs yet, etc... -- FayssalF  -  Wiki me up®  04:21, 10 January 2008 (UTC)
 * In that case, are we able to revoke the tools for those who will not meet the requirements? — Kurykh  04:22, 10 January 2008 (UTC)
 * That's the question and my idea was to avoid seeing it happening or at least keep the dammage (if there would be any) at a low level. -- FayssalF  -  Wiki me up®  04:28, 10 January 2008 (UTC)
 * As far as ArbCom, that is pointless. They cannot do anything about this.  And for contacting the Developers, that has been done elsewhere, where it was said that Dev's will not revert eachother, and the 2 specific ones that implemented this are unavailable. - Rjd0060 (talk) 04:24, 10 January 2008 (UTC)
 * ArbCom might not be able to control what features are turned on or off, but the use of those features is clearly within en.wiki's domain. -- Ned Scott 04:53, 10 January 2008 (UTC)
 * I've protected the page for now. There's a long way to go before the process should be re-opened, with several key points needing to be addressed, including a criteria, duration, and overall procedure. Let's stop the making-it-up-as-we-go-along and draft something. Sean William @ 04:27, 10 January 2008 (UTC)
 * Thanks. One more issue that needs to be addressed is that of granting rollback without a formal request (as opposed to discussing it with an admin).  --Merovingian (T, C) 04:30, 10 January 2008 (UTC)
 * Based on the protection, can I suggest it is removed from MediaWiki:Watchlist-details to prevent more confusion? -Halo (talk) 04:36, 10 January 2008 (UTC)

No protection needed. I think we are imagining problems where they do not yet exist. We should watch how things are working out and learn from what examples of abuse of rollback turns up. Over time we will develop a complicated policy, but that should be based on real results. Liberal granting of rollback should expose the real areas we need to be careful about. Right now setting strict standards would just be guess work, and therefore, not very useful. NoSeptember 04:34, 10 January 2008 (UTC)
 * This protection is a solution to a problem that doesn't exist. Unless someone can provide a reason, I think I'll unprotect.  People  are made admins because the community trusts their judgment and if we can't trust admins to give rollback based on commonsense then there is a bigger problem than  lack of requirements. Rollback is just as easily removed as it is given so I think we should doll it out liberally and deal with the problems as they arise. John Reaves 04:46, 10 January 2008 (UTC)
 * No objection from me on this one. Things have slowed down, and it really depends on who applies for rollback rather than the criteria.  --Merovingian (T, C) 04:52, 10 January 2008 (UTC)


 * I agree we can wait for a problem before resorting to protection. 1 != 2 04:53, 10 January 2008 (UTC)


 * Uh, clearly there is a problem that does exist. I was actually about to post that I was going to sleep on the idea of arbcom since RfR had been protected for the time being, and now it's been undone. This is a half-baked, highly disputed, process that need, at the very least, a little bit of thought before giving hundreds or thousands of users rollback. -- Ned Scott 05:05, 10 January 2008 (UTC)

Personally I was expecting a RfA like process.  Marlith  T / C  05:11, 10 January 2008 (UTC)

Proposal
Given the trend that I'm beginning to be see here, I propose that three admins must explicitly support the granting of rollback, with the third admin granting it. Feedback welcome. — Kurykh  04:30, 10 January 2008 (UTC)
 * I, for one, like that a lot better than a waiting period. --Merovingian (T, C) 04:31, 10 January 2008 (UTC)


 * Plenty of time to discuss whether this is a permanent solution, later; for now, this sounds like a good idea, to me. – Luna Santin  (talk) 04:33, 10 January 2008 (UTC)
 * Yes, it might take a while, but time isn't really of the essence.  Ry an P os tl et hw ai te  04:34, 10 January 2008 (UTC)


 * * facepalm* Remember the old, long-since disregarded saying "adminship is no big deal?" Well, rollbacker is really not a big deal.  Seriously.  &#10154; Hi DrNick ! 04:34, 10 January 2008 (UTC)


 * It seems a lot of people disagree. Until something is decided, should somebody blank or alter Watchlist-notice so people will stop coming here expecting to receive the rollback? - auburn pilot's   sock  04:36, 10 January 2008 (UTC)


 * I quite agree. --Merovingian (T, C) 04:37, 10 January 2008 (UTC)


 * Well, true, but remember that here on Wikipedia, anything can be turned into pointless drama. Then it's a big deal, so I would rather grant this tool with agreement rather than aimlessly. — Kurykh  04:39, 10 January 2008 (UTC)
 * Do we trust our admins' judgement or not? I see nothing here to indicate that the granting of these rights has been less well considered that the granting of page protection requests etc. NoSeptember  04:45, 10 January 2008 (UTC)
 * Page protection involves articles, which can be diverted to talk pages. Rollback involves a tool, and experience tells us that we rarely revoke tools without drama. — Kurykh  04:46, 10 January 2008 (UTC)
 * Honestly, I don't think the first rollbacks granted when this process came out are any less good than the most recent. --Merovingian (T, C) 04:50, 10 January 2008 (UTC)
 * We block well established users for 3RR all the time. Revoking rollback for cause is not going to be any more drama producing than that. Let's trust our admins to act sensibly as they do overwhelmingly in all the other things they are trusted to do. NoSeptember  04:53, 10 January 2008 (UTC)
 * I agree with your basic premise. — Kurykh  04:56, 10 January 2008 (UTC)
 * Look, I didn't propose this because I wanted more bureaucracy and instruction creep. Given that we're not going by any sort of precedent, I would want to proceed more cautiously instead. — Kurykh  04:44, 10 January 2008 (UTC)
 * Yes, during recent proposals people kept saying "We'll figure it out as we go" -- it'll be much easier to do that if we actually discuss things as we go, at least a tiny bit. Once some general standards evolve, we can revisit such processes. The opening flood is going to be bizarre in any process of this nature (I especially remember WP:USURP going live), so I don't see much harm in treating a unique situation uniquely. – Luna Santin  (talk) 04:48, 10 January 2008 (UTC)

Perhaps a combination of the 2 ideas: X number of admins agree (2 or 3) OR one admin can do it if X amount of time passes with no objection (15-30 min). While there are lots of admins active here now, people will give it less attention when it is less of a novelty and there are certain times of day when fewer people are active. Mr.  Z- man  04:46, 10 January 2008 (UTC)
 * No objection. — Kurykh  04:47, 10 January 2008 (UTC)


 * I would also be amenable to that. – Luna Santin  (talk) 04:48, 10 January 2008 (UTC)
 * This seems like a wiki version of "sign here, here, here and here, initial here, and file the form in triplicate". John Reaves 04:50, 10 January 2008 (UTC)


 * Sounds good. --Merovingian (T, C) 04:50, 10 January 2008 (UTC)
 * Probably the best idea so far.  Ry an P os tl et hw ai te  04:53, 10 January 2008 (UTC)


 * Good idea, but I prefer we wait at least a week and decide with experience. We really don't know if this rule is needed or not. 1 != 2 04:54, 10 January 2008 (UTC)
 * Well, it was a proposal based on what is already happening. So, it won't be much of a change. I got the idea from Merovingian's tagging of requests with his /her approval. — Kurykh  04:57, 10 January 2008 (UTC)
 * 15-30 minutes seems fast; 3-6 Hours seems appropriate. I understand this will backlog in the short term but it will ease after 2 weeks. Admins will begin to forget to check this weeks from now, and sly types will apply at 0700 UTC. And this is now coming from a guy who was just granted approval in less than 5 minutes.--CastAStone//(talk) 05:04, 10 January 2008 (UTC)
 * Um, there were complaints that 1 hour was already too long, let alone 3-6 hours. Btw, 0700 UTC is 11:00pm for me, so it's not really much of a concern. — Kurykh  05:06, 10 January 2008 (UTC)


 * Wikipedia ran fine for years with only a small population of editors having rollback, and no Twinkle. Why is it important the everyone who wants rollback get it right this minute?  Would it be so bad to leave a request up for twenty-four hours?  Really?  If I'm going to say 'this person shouldn't get rollback right now', I'd like to be able to back it up with a couple of diffs, and to be able to say it as politely as possible.  A fifteen-minute window makes for fast, sloppy work.  I can also foresee situations where nationalist/ethnic edit warriors just wait until their opponents are asleep to post a request.... TenOfAllTrades(talk) 05:05, 10 January 2008 (UTC)
 * Given enough attention, the background can be checked within minutes. 24 hours is not really needed for a routine check of contribs. — Kurykh  05:08, 10 January 2008 (UTC)
 * I think it'd work better if was similar to the PROD system were it happens as long as there is no opposition. John Reaves 05:09, 10 January 2008 (UTC)
 * I was thinking the same thing. --Merovingian (T, C) 05:12, 10 January 2008 (UTC)
 * Would frivolous oppose rationales be permissible? Because prod does somewhat work on that principle. — Kurykh  05:12, 10 January 2008 (UTC)
 * If you can determine the difference between frivolous opposition and bad-faith opposition, then I suppose so. --Merovingian (T, C) 05:19, 10 January 2008 (UTC)

Requiring support is bureaucratic. We should be able to trust admins to use common sense. John Reaves 05:26, 10 January 2008 (UTC)
 * If this page can maintain as least as much admin traffic as WP:RFPP, requests that sit for more than a couple hours with no comment should be fairly rare. Mr.  Z- man  05:29, 10 January 2008 (UTC)

This is probably going to get lost in the shuffle
...but I didn't even get a chance to defend myself before the page was fully protected. The admins protected the page, then called me out on a couple things, then removed the whole deal. The page history shows this. It may be moot since it looks like this whole "feature" is about to be thrown out the windows anyways, but I wanted to voice my concern in hopes that someone hears me out. Coreycubed (talk) 04:32, 10 January 2008 (UTC)
 * For the time being, admins can listen and respond to your concerns here. Apologies. -- FayssalF  -  Wiki me up®  04:34, 10 January 2008 (UTC)
 * Not a problem; I understand there are bigger things on the plate here than me and my ability to revert edits (I've been following a lot of the discussion, polling, etc). For the record, as a follow up -- I'm guessing that it has to do with my lack of edit summaries and/or using Twinkle to revert edits that aren't technically vandalism. Well, most of the time when I'm making those edits, I don't have time to do anything more than click a button, so even though the edit is legit, I'm not summarizing it (no one's said anything about it so far). Furthermore, if I'm just editing it to go back to the way it was, I use Twinkle, which technically is supposed to be for removing blatant vandalism only (?). I know I'm not the only one editing articles on my watchlist that uses Twinkle to revert edits that aren't blatant vandalism (such as poor/unsourcable edits). If this is a problem, I'd appreciate someone putting me straight -- or I'll go on blindly using Twinkle to wreak havoc with new editors... Coreycubed (talk) 04:39, 10 January 2008 (UTC)
 * When using Twinkle, if you go with only the edit summary you gave in those edits, it's assumed to mean that you were reverting vandalism. If it wasn't vandalism, there should be additional information saying to that effect. What I objected to was what appeared to be reverting as vandalism when it isn't blatant or apparent. — Kurykh  04:41, 10 January 2008 (UTC)
 * Thanks, I need to either stop making those edits at work, or pay enough attention to put a little edit summary in before I click. I appreciate the feedback! Coreycubed (talk) 04:47, 10 January 2008 (UTC)
 * Thank you for taking it so well. :)  --Merovingian (T, C) 04:54, 10 January 2008 (UTC)

Other stuff
Just to remind everybody that Wikipedia isn't shut down for this discussion and admin attention is also needed elsewhere (here for example). - Rjd0060 (talk) 05:19, 10 January 2008 (UTC)

Full stop needed
User:Random832 brings up the best point I've heard in all of this: "EVERYBODY CALM DOWN. ok, are you calm now? listen to this: The developers did not implement the proposal that was voted on. They implemented a single technical feature. There was a wide consensus, including many of the oppose voters, that we should have some form of this feature. If you don't like the proposal, you're free to suggest alternatives for what the policy or process or whatever should be for giving rollback. We are not bound to the proposal. —Random832 02:53, 10 January 2008 (UTC)"

He brings up a very good point. This whole time, certain users have been using the fact that the feature was activated as a declaration of consensus. Most of us were too worked up over the fact that a dev made the change to notice the core of the issue, that rollback was going to be enabled in some way (strong support for bots, and indications of strong support for scripts), but the process on this page, as we see it, was not it. Not even among those who supported rollback. They quickly acted on their own, too excited that we had the feature turned on, to even stop and think that we were still far from being ready to implement anything.

So would a sensible admin please re-protect this page. -- Ned Scott 05:51, 10 January 2008 (UTC)


 * A very sensible post. The fact that a dev enabled the feature did NOT make this a free-for-all. After reviewing the bugzilla entry the dev didn't claim that a consensus was reached. He simply enabled a feature. And I tend to disagree with the assertion that simply enabling a feature makes USE of that feature a consensus approved by the dev. I agree, a protect until all of this is sorted out is a VERY good idea, however, I don't see it forthcoming without involvement by either ArbCom or some higher force of nature.  Justin  chat 05:58, 10 January 2008 (UTC)


 * What it means is that we need to work out a Rollback policy. Here is a page we can use for that purpose: Rollback policy. I invite you to work on it. Perhaps it should be worked out a bit before we move on to ArbCom, since the community is the maker of policy, not ArbCom. NoSeptember  06:17, 10 January 2008 (UTC)


 * As a concerned editor that was just granted rollback, if a full stop is done on the process, what if anything becomes of the editors that have already been granted this feature? What I propose is this: editors that have been given rollback up to the point of process stop be allowed to retain the feature, with the standing proviso that it be used wisely, in the efforts to rollback obvious vandalism. (I understand that this creates a "slipped through the door before opening time" situation, however what is done is done, do we really want to do a "full-reverse", and undo all the grants to this point? (I personally will go with whatever decision is arrived at...) Edit Centric (talk) 06:23, 10 January 2008 (UTC)
 * We can and should work on the policy without stopping the running of the project, just as we tweak all our policies on the run. You should show the community how well non-admin rollback works by using it. :) NoSeptember  06:31, 10 January 2008 (UTC)
 * Will do, at least I shouldn't have to sift through diffs, copy-pasting and manually editing anymore, which is a beautiful thing. Edit Centric (talk) 07:21, 10 January 2008 (UTC)
 * There is a great deal of argument about whether or not this was implemented with a true consensus, and to suggest it should "continue on" is to suggest that we can now start ignoring consensus, and more importantly, minority opinion. We don't implement policy before consensus then improve. We improve a proposal to find a consensus and then implement.  Justin  chat 08:52, 10 January 2008 (UTC)

If you want to slow things down or stop things, you should remove the message from users' watchlists, that says "Users can now acquire rollback on an individual basis." Even if there were full consensus on the implementation of this feature, it would have been better to allow requests to trickle at first rather than notifying every editor of the new feature's availability and having a flood.--Srleffler (talk) 06:31, 10 January 2008 (UTC)


 * You know, that might not have been such a bad idea. Then again, there are a lot of us that have been following the progress of this proposal... Edit Centric (talk) 07:21, 10 January 2008 (UTC)

What's the Difference?
Forgive me, but I'm a little confused. I was introduced to Twinkle by a friend just a couple of hours ago and "installed" it onto my account. There are three rollback features on that one. I'm a little puzzled by this: what's the point in creating this admin-installed rollback if there's a twinkle? No worries, I'm following the Edit Summary rule as I was asked to :) But I'm still not quite clear: what's the difference between the two, and if there is, could you get this particular rollback in addition to Twinkle? BlackPearl14 Pirate Lord-ess 06:12, 10 January 2008 (UTC)
 * This rollback is completed server-side with a single request rather than client-side which requires multiple page requests. Nakon  06:13, 10 January 2008 (UTC)
 * And Twinkle can't be used by everybody as it requires good javascript support. BJ Talk 06:25, 10 January 2008 (UTC)

Right, ^^, so you can use both Twinkle ‘’and’’ the admin-“locked” rollback? Sorry, just a little confused. Thanks for all your help, though, that cleared up some things :) BlackPearl14 Pirate Lord-ess 06:31, 10 January 2008 (UTC)
 * If you have rollbacker rights you can use both. BJ Talk 06:42, 10 January 2008 (UTC)

Lighten up!


(care of Gurch - see history of project page)

Viridae Talk 06:29, 10 January 2008 (UTC)
 * PS It was too funny to waste! Viridae Talk 06:30, 10 January 2008 (UTC)


 * I saw that hit. I can't stop laughing. -- Shakata Ga Nai Talk 06:33, 10 January 2008 (UTC)

That's the exact same thing I thought when I realized I could assign myself rollback. :) --Merovingian (T, C) 06:44, 10 January 2008 (UTC)

edit summary question
Hello, I've heard of the non-admin rollback, but what about edit summaries? I tend to use more verbose edit summaries, so that even when undoing vandalism, I'll normally add "rm personal attack; undid..." or something similar. Does this rollback allow that? Thank you. -- Kyok o  06:32, 10 January 2008 (UTC)
 * It shouldn't, you'll have to use a script for that. John Reaves 06:44, 10 January 2008 (UTC)


 * It was one of the things we were supposed to discuss before hand... -- Ned Scott 06:46, 10 January 2008 (UTC)
 * Rollback does allow edit summaries (both when used by administrators and by others), it just requires an extra parameter to be passed in the URL. This is easy for scripts to implement – Gurch 06:48, 10 January 2008 (UTC)
 * This rollback would pretty pointless if it required edit summaries. John Reaves 06:48, 10 January 2008 (UTC)
 * Not requires, allows. A script utilizing rollback and passing a custom edit summary is far more efficient than one reverting manually – Gurch 06:50, 10 January 2008 (UTC)
 * Good point. It should be made clear that those wanting to reduce server load and still use informative edit summaries should: (a) get this right turned on; and (b) use a script that uses the new right as well as allowing custom edit summaries. At least I think that is what Gurch is saying? Carcharoth (talk) 13:35, 10 January 2008 (UTC)

Instructions needed.
Is there a page with instructions on the use of this feature? If so, it should be linked from this project page. (Yes, I know it's not that complicated a feature. That just means the instruction page should be short. It should cover both how to use the feature and when to use it.)--Srleffler (talk) 06:51, 10 January 2008 (UTC)
 * Nobody has got things organized yet:
 * Rollback feature
 * Help:Reverting
 * Rollback policy
 * Requests for rollback


 * all have in many cases duplicated or wrong info, but info nonetheless – Gurch 06:52, 10 January 2008 (UTC)
 * Well Help:Reverting has been modified at the source (Meta) but not updated for enwiki yet. But yes, all these pages need to be worked on. NoSeptember  06:55, 10 January 2008 (UTC)
 * Ah, I see my post lower down duplicated this section. Carcharoth (talk) 13:36, 10 January 2008 (UTC)

Any updates to documentation should include the point made here, namely that custom edit summaries can be included using scripts. This removes a large part of my objections. I feel that a mass rollback should have a tailor-made edit summary attached explaining the mass rollback, and I still feel uncomfortable that people feel it is OK to rollback edits with an automated edit summary - being able to select a reason for the rollback shouldn't be considered an onerous requirement. Carcharoth (talk) 13:43, 10 January 2008 (UTC)

Can someone pleas slow down the archive bot?
It archives very quickly - can we at least give it 15-20 minutes from request done/not done so if someone can object to the ecision? Viridae Talk 07:01, 10 January 2008 (UTC)
 * While we are at it, it cant handle multiple rejected requests on the same day. It does not add the new request. Viridae Talk 07:13, 10 January 2008 (UTC)

Flawed to the hilt
An editor, who had a 3RR block a little over 2 months ago gets ok'ed but me who had an "edit war" block almost 2 months ago, gets denied? Flawed process for sure. -- <strong style="color:#fff;background:#DC143C;border:1px solid #000">ALLSTAR <strong style="color:#FFF;background:#0F4D92;border:1px solid #000"> echo 07:05, 10 January 2008 (UTC)
 * For the love of god, stop bitching. You were blocked for edit warring some 1 month and 11 days ago to be exact, hardly almost two months. Viridae Talk 07:07, 10 January 2008 (UTC)
 * It may or may not be flawed, but that fact does not follow from your comment. Administrators should exercise discretion in deciding whether to grant rollback; if they feel one editor's conduct has improved over two months (which is after all quite a long time) but another editor's conduct has not, then surely this would be the right thing to do? Not everything has to be automatically approved/not approved based on fixed critera; this is not RfA – Gurch 07:09, 10 January 2008 (UTC)
 * Oh and to get into semantics, the other block was 1 month and 9 days before yours, making it closer to 3 months ago than 2, and almost double the length of time has elapsed. Viridae Talk 07:10, 10 January 2008 (UTC)
 * And yet you tell me to come back in a couple of months? Yeah, I see. -- <strong style="color:#fff;background:#DC143C;border:1px solid #000">ALLSTAR <strong style="color:#FFF;background:#0F4D92;border:1px solid #000"> echo 07:11, 10 January 2008 (UTC)


 * Even if it wasn't perfect, no system would ever be. Not having the process at all would have some relative downsides somewhere too. Argument from lack of perfection tends to generally just be "throwing a fit".  Voice -of- All  07:13, 10 January 2008 (UTC)
 * Agreed. However, other people, admins to be exact, see my request differently as they stated on the page. The only reason Viridae was so quick to deny it and not allow input from other admins, as other users seem to be getting but I didn't in my first go at this, is because he denied the first one. Look, I was told by other admins to repost it again and I did. It's flawed regardless. Thanks. I'm done now. -- <strong style="color:#fff;background:#DC143C;border:1px solid #000">ALLSTAR <strong style="color:#FFF;background:#0F4D92;border:1px solid #000"> echo 07:15, 10 January 2008 (UTc)

A 3rr block is irrelevant unless the user wasn't using edit summaries or somehow abusing a rollback script. John Reaves 07:17, 10 January 2008 (UTC)
 * Mine was actually 3RR but the blocking admin put it as edit warring. Same thing. And mine had edit summaries to boot. -- <strong style="color:#fff;background:#DC143C;border:1px solid #000">ALLSTAR <strong style="color:#FFF;background:#0F4D92;border:1px solid #000"> echo 07:19, 10 January 2008 (UTC)

Man, I'd hate to see Allstarecho fail RfA - I'm sure the regulars at WT:RFA would warm nicely to his comments that RfA is broken. Seriously, per Viridae. Daniel (talk) 07:37, 10 January 2008 (UTC)
 * Luckily, this isn't RfA. This is rollback, without rules and regulations. ;] -- <strong style="color:#fff;background:#DC143C;border:1px solid #000">ALLSTAR <strong style="color:#FFF;background:#0F4D92;border:1px solid #000"> echo 07:43, 10 January 2008 (UTC)
 * It has both. The rules are only administrators can grant or decline requests, based on a discretionary evaluation. All users are regulated by other Wikipedia policies, such as civility and our general standards of behaviour. Daniel (talk) 07:52, 10 January 2008 (UTC)
 * Yes but that's no reason to not give rollback. John Reaves 07:55, 10 January 2008 (UTC)
 * I don't disagree, and I never implied that perfect civility was required to be approved. I meant that all users are expected to behave in a civil fashion when interacting in and around this process, refuting the statement that this process has no regulations. Daniel (talk) 07:58, 10 January 2008 (UTC)
 * I won't be applying for this feature. I feel that the rollback tool is a privelage for admins who have gained the trust of the whole community and not just a select few who think they know better. I will wait until I become an admin (IF ever) to use this feature to honour my committment to the community who should be able to discuss who gets the tool or not. I certainly would not have condoned an editor blocked so recently for 3RR getting the tool even if they'd nearly 3 months, but obviously it's not the community that counts. That's all I will say and will not comment further on the topic. I've made my protest/boycott. Cheers, Spawn Man (talk) 08:34, 10 January 2008 (UTC)
 * It isn't like they couldn't already do it utilising more bandwidth and server resources using scripts. No-one is banning those scripts that I know of. Why don't you want people who don't need delete/undelete privileges to smooth out the way wikipedia works in some small way? Ans e ll  08:45, 10 January 2008 (UTC)
 * Well, its true that this can be done with scripts. But as a user of Popups, and now Rollback (yes, I applied earlier).  This is SOOOOO much faster.  If I was a malicious user, I could easily to 10 times more damage with Rollback and with popups -simply because of how fast it works.  Speed is a dangerous thing.  Now for me doing WP:CVU, this feature is wonderful, and I never want to have to do with out it ever again. -- Shakata Ga Nai  Talk 08:53, 10 January 2008 (UTC)

Rollback and sysop
Can admins have both groups, and would doing so make any difference? Also, if a rollback user passes a RFA, do they have rollback removed? --Rschen7754 (T C) 07:35, 10 January 2008 (UTC)
 * Technically they can, although it seems moot. With regards to removing it with a successful RfA, I guess the bureaucrats could do it in one hit when they add the +sysop, but it seems to be a slightly irrelevant point :) Daniel (talk) 07:39, 10 January 2008 (UTC)
 * It'd only become relevant if the user is then desysopped for some reason. Presumably, the desysopping would also include the rollback tool, right? Whomever desysops should also remember (or be reminded, or be explicitly instructed) to also remove other relevant rights, such as the rollback tool. UltraExactZZ Claims~ Evidence 16:03, 10 January 2008 (UTC)
 * De-sysopping is a meta issue. Rollback is an en.wiki issue. If a de-sysop is voultary the former admin comes here and requests the rollback tool (if wanted) and they get it automatically. If it is a controversial de-sysop then the request is reviewed here but may not be granted depending on the reason for the de-sysop. Pedro : Chat  16:06, 10 January 2008 (UTC)
 * I think the point was that some admins have both rights (sysop and rollbacker), and if they are desysopped, who decides whether they keep the rollback right? You can't expect stewards to remove the rollback tool at the point that they perform a desysopping. Carcharoth (talk) 16:23, 10 January 2008 (UTC)
 * Sorry, yes, with you now. Seems silly for admins to go giving themselves rollbacker rights, but there we go. Pedro : Chat  16:25, 10 January 2008 (UTC)

(unindent) For the record, at the time of writing, eight administrators also have the rollback flag (out of around 200 rollbackers - log summary in brackets): User:Archtransit (Users up for RfA.. Worthy for a test) the user was later sysopped, User:Camaron1 (Adding myself to rollback group to make clear I have rollback tool access via adminship.), User:Krimpet (always low prices. always.), User:Misza13 (liek zomg, will my rollback work twice as fast now?), User:Sean William (abuse!), User:The Placebo Effect (rollbacker), User:WilyD (Evidently trusted), User:Zscout370 (promotions). Several administrators also had the flag added and later removed (both actions variously by themselves or others). Seven of the eight were admins adding themselves, probably to test the system. One case was a rollbacker having just passed at RfA. Carcharoth (talk) 17:03, 10 January 2008 (UTC)
 * As one of the eight that tried it, all I can say is it does not make any difference to how the rollback feature works for you (not that I was expecting otherwise). I have removed myself as it was a little pointless, and I now know how to use Special:Userrights well enough. Camaron1 | Chris (talk) 21:19, 10 January 2008 (UTC)

Slowing it down
I increased the approval from 15 to 60 minutes, although I favour a day. I doubt that an hour (not to mention 15 minutes) allow enough people to critically review a nomination; having no opposition, or countering support to opposition in case of rejections, in that brief time frame is no indication of anything. And the bot is archiving closed nominations too fast, even as discussion remains ongoing (is it picking up on the icons?). El_C 08:34, 10 January 2008 (UTC)
 * Bots not doing as they're told? Lol, have you seen The Terminator? ;) Spawn Man (talk) 08:37, 10 January 2008 (UTC)
 * Yeah, it's really a very small step from archiving closed RFR nominations too quickly to nuclear war – Gurch 09:04, 10 January 2008 (UTC)
 * I'm not going to take 60 minutes for users I'm already familiar with. That's ridiculous.  Majorly  (talk) 09:04, 10 January 2008 (UTC)
 * Other people are allowed to have a say about those nomiantions, not just Majorly. El_C 09:07, 10 January 2008 (UTC)
 * The same can be said for protection/blocking etc. We don't require an hour's wait before we block a vandal or protect a page do we?  Majorly  (talk) 09:11, 10 January 2008 (UTC)


 * Ssssh, don't give them ideas – Gurch 09:11, 10 January 2008 (UTC)
 * This permission request is much closer to RfA than RfP. El_C 09:14, 10 January 2008 (UTC)
 * Hardly. It's more like AWB or VPRF approval - and that's by one person only.  Majorly  (talk) 09:16, 10 January 2008 (UTC)
 * I'm having second thoughts about a fixed time frame now... But the reason I added it is also, crucially, because it appears to be proliferating very fast as of its activation date (today). El_C 09:25, 10 January 2008 (UTC)
 * It proliferated fast because for 6 years users have not been able to use rollback and now they can, thus there is a backlog of people who wanted it but did not have it. That will have mostly emptied by the end of today and then the request rate will be much lower – Gurch 09:27, 10 January 2008 (UTC)

Right I've just done one :). Took me 6 minutes to check logs, talk and user pages and random sample of 1,000 contributions. Have I done wrong? Pedro : Chat  09:13, 10 January 2008 (UTC)
 * No.  Majorly  (talk) 09:15, 10 January 2008 (UTC)
 * I'm looking at User:MPerel - taken me four minutes - I can ascertain an clean log, sensible user page, talk looks good, uses pop-ups to undo at the moment - has participated across the wiki - user will benefit and can be trusted with the tool. This does not need 60 minutes. It really doesn't. Pedro : Chat  09:18, 10 January 2008 (UTC)
 * On second thoughts, use your discretion. But I don't want to see conversations cut short and people who don't get a chance to comment because applications are virtually instantaneous. El_C 09:22, 10 January 2008 (UTC)
 * I agree totally. How can I put this.... I haunt WP:RFA far too much. I pretty confident that I can detect any issues that need further debate or consideration within a five minute review of contributions / talk pages / logs etc. . I agree that this does not need to be done lightly or without due regard, but 60 minutes just seems to be without value. Pedro : Chat  09:25, 10 January 2008 (UTC)
 * spend 55 minutes doing other stuff and then come back and do the request – Gurch 09:28, 10 January 2008 (UTC)
 * Exactly. Or spend 5 mintues doing the review and 55 doing the other stuff. Whichever way round the end result is the same. Pedro : Chat  09:29, 10 January 2008 (UTC)
 * Just don't approve things after one minute, like earlier. Please. There were three users who I wanted to object to and got edit-conflicted with people saying "Done" less than sixty seconds after it was added to the page. As I said above, it takes me at least three minutes to do a basic check of all the required pages, and even then this isn't exhaustive. Daniel (talk) 09:34, 10 January 2008 (UTC)
 * Maybe we need a guide for approval that's more than a one-sentence tag. El_C 09:36, 10 January 2008 (UTC)

Fine, I removed the entire thing. Use your discretion and we'll see what happens. Either there's a fixed time, like RfAs, or we treat this as AWB. El_C 09:36, 10 January 2008 (UTC)

The way I see it, we can't afford to let rollback rights become a big deal like adminship did. The longer the processes to get it or remove it are, the bigger of a deal it becomes. Leave it for a few minutes to see if anyone objects, review the user, approve/decline. If someone brings up an issue later, it should not be exponentially harder to remove the right based on that than it would have been to decline the request if they had brought up the concerns before it was approved. Perhaps the bot could be configured to leave requests on the page for a while after they have been done, or move them to a temporary "processed requests" section on the bottom of the page like WP:RFPP in case people have concerns after the request has been processed. Mr.  Z- man  09:38, 10 January 2008 (UTC)
 * (ec'd)Discretion and sense. There's no competition to grant the rollback button, same as there's no competition to plough through CSD. I for one do not want to be the first person taken to WP:ANI as here's the stupid amdin who granted rollback to a POV warrior. That's enough to hold me back and make sure I give a fair and critical assesment. Pedro : Chat  09:41, 10 January 2008 (UTC)
 * I think there should be a delay this user was granted while I was looking into his edits, IMHO I have concerns after looking through 2000+ edit primarily awb edit I found no vandalism reverts, and now concerned about the way in which awb is being used to add and remove spaces. Gnangarra 13:54, 10 January 2008 (UTC)

Slower is better. I just saw someone being given rollback who had an already established history of inappropriate reverting. Now it's done. Leave at least a few days for review on these things. Friday (talk) 15:40, 10 January 2008 (UTC)
 * No, that's the attitude we have to prevent: "it's done, too bad" - if you have an issue, take it up with the admin who granted it. Mr.  Z- man  23:59, 10 January 2008 (UTC)

Documentation tidying up
Can we tidy up the Help and Wikipedia pages about this, please? See Wikipedia talk:Rollback feature. Thanks. Carcharoth (talk) 12:42, 10 January 2008 (UTC)

1 minute assessment
I must ask are these actually being checked this dif it was 1 minute from request to approvale with another editor approved by the same person in between. Gnangarra 14:18, 10 January 2008 (UTC)
 * If I know someone well, and know I can trust them, I grant them it without waiting anyt time at all. But I would agree, you should be absolutely position they aren't going to abuse the tool in order to grant it, so checks should be made.  Ry an P os tl et hw ai te  14:20, 10 January 2008 (UTC)
 * (ec'd)Agreed. As noted above, I just can't see how you can really do justice in less than a few minutes. To check the basics (IMHO) of block log, user and talk pages and a few random samples amongst the last 1,000 contribs (identifying some instances of reversion using whatever tool the editor prefers e.g. undo, TW, pop ups or just by hand) must take four / five minutes. Pedro : Chat  14:21, 10 January 2008 (UTC)
 * Approved in 1 minute, archived in 7 minutes this process has a serious lack of transparency and no ability for for any discussion. IMHO if you know the editor then isnt that a WP:COI as such an admin should refrain from the approval. Gnangarra 14:25, 10 January 2008 (UTC)
 * Heh, I had a feeling that this might come up. I can assure you that in that time that I was able to check everything you said and more, Pedro (block log, normal log, user talk page with history, user contributions, and Interiot edit count).  I extensively use Firefox tabs and Lupin's popups so that I can sort of queue up things I know I want to look at to let them load while I examine other things.  I think it saves me a lot of time, and I stand by my decisions in those RFR approvals.  On a sidenote, I had no previous dealings with any of these editors before responding to their requests on this page.  Ioeth (talk contribs friendly) 14:29, 10 January 2008 (UTC)
 * And in fairness, the approvals you have made I would have approved as well. You're just too darn quick for me! Pedro : Chat  14:31, 10 January 2008 (UTC)
 * the COI was directed at Ryans If I know someone well, and know I can trust them, I grant them it without waiting not yours, Gnangarra 14:34, 10 January 2008 (UTC)
 * No problem! I just wanted to make it known in the interest of transparency, which I am always amicable to with regard to any of my actions on (or off, really) Wikipedia.  [[Image:Smile.png|16px]]  Ioeth (talk contribs friendly) 14:47, 10 January 2008 (UTC)
 * I should clarify, I nominate quite a lot of people for adminship, so I go through a lot of peoples contributions in depth through that. This is where the trust comes in, it's nothing to do with someone being my buddy or anything!  Ry an P os tl et hw ai te  15:26, 10 January 2008 (UTC)

I opposed this on the basis of too much process. I actually have never feared people having rollback: Editwarriors will war anyway, and anyone who vandalises with it will be blocked. Rollback is not only "no big deal" it is no deal at all. Thus, until and unless there is demonstratable and stable consensus otherwise, I am willing to grant rollback to any user (exempting obvious trolls) on request. My objection is to any process or instruction creep here.--Docg 15:45, 10 January 2008 (UTC)
 * It clearly states in the instructions that administrators should look for signs of misuse, such as edit warring, you do really need to vet the people who you give the tool to.  Ry an P os tl et hw ai te  15:48, 10 January 2008 (UTC)
 * Oh, I will vet them. But my threshold for granting will be extremely low. I do not view mild edit warriors having rollback as alarming. We block edit warriors, if they use rollback, we block them faster.--Docg 15:52, 10 January 2008 (UTC)


 * Whether it's right or wrong, some editors will see this as a badge of honor. So, it may be mildly harmful to go handing out rollback to known edit warriors - it may give them (or other editors) the impression that the community has somehow approved of their inappropriate behavior.  Friday (talk) 15:55, 10 January 2008 (UTC)
 * Uhh, I seriously doubt that will happen. John Reaves 16:03, 10 January 2008 (
 * Also, I too think almost anyone should get rollback. Even the most prolific edit warriors should get it unless they have a history of abusing scripts or undo. John Reaves 16:06, 10 January 2008 (UTC)
 * Whilst agreeing most editors should get the tool, only an analysis of logs and contributions will reveal abuse of scripts or the undo feature. Therefore a proper "vetting" is important. It's not CREEP just being sensible. Pedro : Chat  16:09, 10 January 2008 (UTC)


 * (Reply to John above) It has already happened. Whether this is really much of a concern is debatable.  Friday (talk) 16:12, 10 January 2008 (UTC)
 * "some editors will see this as a badge of honor" - only if its hard to get and harder to remove. Otherwise it is just a tool. Mr.  Z- man  00:00, 11 January 2008 (UTC)

Removal
So, what about removal of rollback rights? Are we really supposed to go to WP:AN/I if someone uses rollback in an edit war? That doesn't sound sensible to me. We're not going to the noticeboard when someone edit wars either (assuming it's a clear case of edit warring), we warn them. And, if that doesn't work, we block them (or protect the article in question). Shouldn't it work like that with rollback, too? Warn them if they use it disruptively, and take the right away if the problem continues. WP:AN/I should only be used in cases that aren't as clear, IMHO. --Conti|✉ 15:22, 10 January 2008 (UTC)
 * Logical argument: An admin can remove someones right to edit (block button). That's way higher up than the ability to rollback. Every block does not have to go to WP:ANI. So I would say removal is uncontroversial (either for a short period of time or permanently) as another admin can override it. All that has to be remembered is that when granting rollback to check the logs in case it was previously revoked. Contentious or bad removals of access are treated like bad/disputed blocks and are taken to ANI. Pedro : Chat  15:32, 10 January 2008 (UTC)
 * As there's no formal review/complaint process it does need to go to AN/I, even if its just an after the event review. If there are too many for AN/i then it provides some basis for either establishing a process or removal of the tool all together. Gnangarra 15:41, 10 January 2008 (UTC)
 * I feel that we probably need a bit of policy regarding the granting of these rights in order that the removal of them can be justified.... Pedro : Chat  15:44, 10 January 2008 (UTC)
 * Why don't we establish a process for the removal of rollback rights right now? Personally, I think it should be handled just like we handle blocks. Uncontroversial removals of rollback rights don't need to be mentioned on AN/I, not so clear cases can be discussed there, just like appeals. --Conti|✉ 16:13, 10 January 2008 (UTC)

Slow down archiving
I think that closed requests should remain on the page for at least 12 hours. —Random832 15:37, 10 January 2008 (UTC)
 * I agree, but maybe later; at the moment there are so many requests that the page would become huge very quickly. Let's get the tidal wave sorted to the best of our ability, and then worry about the archives. Pedro : Chat  15:42, 10 January 2008 (UTC)
 * I had already raise the issue with the User:RFRBot think this should act similar to WP:RFPP Gnangarra 15:44, 10 January 2008 (UTC)
 * Yeah, good thinking. Pedro : Chat  15:46, 10 January 2008 (UTC)

This might as well be automatic
People are running around approving willy-nilly; there's nothing resembling consistent standards yet. If this is going to be the case, why not just turn on rollback for everyone by default, and then this process will be used only to decide when to remove it? This would put the discussion time where it needs to be- on those cases where someone has a concern. Friday (talk) 16:06, 10 January 2008 (UTC)
 * Agreed, with exception.. not default for IP and newly registereds. -- <strong style="color:#fff;background:#DC143C;border:1px solid #000">ALLSTAR <strong style="color:#FFF;background:#0F4D92;border:1px solid #000"> echo 16:09, 10 January 2008 (UTC)
 * No, many people have been rejected. Have a look at the archive.  Majorly  (talk) 16:10, 10 January 2008 (UTC)
 * Sure, but it's random. If you happen to have your request processed by one of the "everyone should have this" crowd, it'll get approved.  If you happen to have worse luck, they may try to review your contributions first.  Something a bit closer to consistency would be better.  I do think that our time is better spend discussing cases where someone has a concern than every case, particularly when it's often a pure rubber stamp.  Friday (talk) 16:14, 10 January 2008 (UTC)


 * We have thousands of admins to grant these rights, anyone who wants it will get it easily enough. Since vandals can mass create user accounts for vandalism, the positive granting of the rights is still the right way to go, no matter how liberal we are in granting it. I agree that the Rollback policy should mostly focus on removal issues, which we can work on. NoSeptember  16:13, 10 January 2008 (UTC)
 * How can we have consistant standards when the people strongly opposing this are running around giving every single person rollback to make a point? and they've started over-ruling "not done" decisions on the board - you just can't win.  Ry an P os tl et hw ai te  16:15, 10 January 2008 (UTC)
 * Ryan has a good point - is there any procedure to deal with admins using their tools disruptively? DuncanHill (talk) 16:17, 10 January 2008 (UTC)


 * Obviously there is. The sticky wicket here is that it's arguable how disruptive it is to hand out rollback to anyone who asks for it. Friday (talk) 16:18, 10 January 2008 (UTC)
 * Friday has it basically right. Grant the right, and on the first misuse of it, an admin can revoke it or issue a warning, and we can then discuss whether that user should have it or not. That would save a ton of time, limiting debate to only a very few cases. NoSeptember  16:19, 10 January 2008 (UTC)


 * Yes.. I suspect even then the debates would be relatively few. I assume, for example, that if an admin blocks someone for excessive reverting, that admin might also remove rollback at that time, and it would rarely be controversial.  Friday (talk) 16:22, 10 January 2008 (UTC)

That's the best suggestion I've heard in this whole circus.. everyone has it but when a user is warned/blocked for 3RR/edit war, as an additional step to the warn/block, just remove rollback as well. As if you admins don't have enough things to do, this only added another. A good example was last night.. there was a user harassing and attacking another user. All the admins were here bickering over what guidelines and rules and regulations should be used with this feature that in all honesty, shouldn't have been released with a note at the top of every page screaming "come and get it now!" without guidelines and rules and regulations already in place. Because all the admins were here, no one saw where multiple users posted about the offending user on AN first and then on ANI second when an admin couldn't get summoned via AN. -- <strong style="color:#fff;background:#DC143C;border:1px solid #000">ALLSTAR <strong style="color:#FFF;background:#0F4D92;border:1px solid #000"> echo 16:31, 10 January 2008 (UTC)
 * I disagree that the process should be automated, or that rollback comes at (say) autoconfirmed as an automatic thing. But whilst we have this mess it may as well be. Anyone declined may as well now go admin shopping and ask for it of those who feel fit to just grant it anyhow. Frankly, I feel like I've spent a day reviewing with a similar level of commitment that I would give to WP:RFA, and I'd have been better doing it for editor review and encouraging some contributors instead of this. This is turning into a miserable situation. Pedro : Chat  16:22, 10 January 2008 (UTC)

Two reasons this is better than automatic:


 * 1) Several people are being refused for good reason, which would not happen with automatic
 * 2) If someone granted the tool abuses it, we will know which admin made the decision to do it. If an admin abuses their grant tool to make a point or for any other reason then there will be a record that can be used to address the problem.

I think we need more than a day to decide how this is working. Also, there was a clear consensus against giving it to everyone at the poll. 1 != 2 16:33, 10 January 2008 (UTC)
 * Considering that this feature didn't get a clear consensus, I hardly doubt that. -- <strong style="color:#fff;background:#DC143C;border:1px solid #000">ALLSTAR <strong style="color:#FFF;background:#0F4D92;border:1px solid #000"> echo 16:36, 10 January 2008 (UTC)


 * Agreed that there's no reason to rush. People are being refused for good reason, but they're also being denied for bad reasons, and approved when there are good reasons to refuse. It's not going to be easy to approach consistency when people ask, they get the tool and are removed very quickly, with no archive . The consensus against giving it out to everyone was based on fear, right?  Give this a little while to operate (this process is, at some points in time, almost like giving it to everyone) and that fear will go away when people see that even giving it out wrongly isn't such a big deal.  Friday (talk) 16:40, 10 January 2008 (UTC)

So, is it a wheel war...
...if
 * one admin reviews a candidate, shares his reasoning, and rejects the request ;
 * another admin comes along and approves the request, and blanks the previous admin's comment:
 * when the thread is restored by a third admin, again posts an overruling comment:

Seriously. This is not a good trend. TenOfAllTrades(talk) 16:34, 10 January 2008 (UTC)


 * If an admin decides not to use their tools, and then another one does then that is not wheel warring, as choosing not to use your tool is not an admin act. Changing comments are not admin actions either.


 * The behavior may be seen as disruptive(especially if it is to make a point), but it is not wheel warring. 1 != 2 16:37, 10 January 2008 (UTC)


 * Looking at the logs in question, at the time of writing no admin action has been undone. There has been edit warring on the page, but that is about it. Carcharoth (talk) 16:40, 10 January 2008 (UTC)
 * Doc disruptively blanked the whole section because he opposes the process - that's why I undid him. I would never have touched the logs however. People should respect the decision here, and if it's granted or declined, but they disagree, take it to an appropriate venue. We wouldn't find the 'crats blind reverting each other.  Ry an P os tl et hw ai te  16:43, 10 January 2008 (UTC)


 * <Vast edit conflicts, including the edit I am replying to being edited>Um, so when an admin decides not to delete a page and another does, that too is not wheel warring? To sound like Cassandra, I warned people this problem would occur and that it needed further discussion. Reversing an admin action is wheel warring. If that action is closing a request or debate, reversing that decision is still wheel warring, because only admins can perform the particular action. Jeez, wiki-lawyering is just not the right answer to a pertinent point. Hiding T 16:42, 10 January 2008 (UTC)


 * (ec) But it would be wheel warring if a third admin decided that the decision to grant rollback was in error, and undid it&mdash;is that the sense here? If so, isn't that the sort of asymmetry that we weren't supposed to have in this process, and doesn't that make rollback a 'big deal'?
 * Is the process broken if 'keep asking until one admin can be persuaded to say yes' is a workable strategy? TenOfAllTrades(talk) 16:43, 10 January 2008 (UTC)
 * Ten, people ask the other parent in regards to all admin tools, and they get warned or blocked for it. The log will also show any granting or revocation. Even if nobody noticed a user canvasing for rollback, it would be shown on the granting screen if it has been taken away in the past. 1 != 2 16:49, 10 January 2008 (UTC)


 * As crazy as this sounds, no, that's not particularly broken. It's OK to hand this tool out liberally, since we have such an easy way of removing it whenever there's a problem.  Friday (talk) 16:47, 10 January 2008 (UTC)


 * Deletion has specific criteria before it is done, if an admin decided not to delete and another does that is not wheel warring at all. It may be an out of process deletion, it really depends. Choosing not to use an admin tool is not an admin action. I agree with 10 that if someone actually reverts the granting of rollback without good reason or discussion then it could be seen as wheel warring. 1 != 2 16:45, 10 January 2008 (UTC)


 * It seems that people have gotten hung up on my poor choice of terminology. Wherever you see 'wheel war' in my comments, please read it as 'Bad Thing'.  Is it agreed, then, that as long as one admin believes an editor ought to have rollback, no other admin's judgement (before or after) should count? TenOfAllTrades(talk) 16:52, 10 January 2008 (UTC)


 * Treat it like any other choice we make via the wiki process. One can be bold, and the boldness stands until such a time as someone else is bold enough to undo it.  When there's disagreement, we discuss.  Friday (talk) 16:55, 10 January 2008 (UTC)


 * In this case, an admin made a choice, and another decided to erase the first admin's comments and overrule without discussion&mdash;the model that you propose wasn't followed. In any case, I'm a bit twitchy about applying WP:BRD to admin actions instead of edits, which seems to be what you're proposing.  TenOfAllTrades(talk) 17:02, 10 January 2008 (UTC)


 * Yes, that's what I'm proposing. I've been doing it for 2+ years and haven't noticed any problems with it.  Friday (talk) 17:05, 10 January 2008 (UTC)

It's not going to be a wheel war, because I'm not going to take any action. I'm washing my hands of that. I made a decision. Doc decided it was the wrong one. End of story. Pedro : Chat  16:48, 10 January 2008 (UTC)


 * I'm more concerned about the general case. Not everyone is going to be as reasonable (or give up in frustration...?) as easily as you, Pedro.  We don't need the wikidrama, so there need to be some ground rules for these cases. TenOfAllTrades(talk) 16:59, 10 January 2008 (UTC)


 * I would just use regular dispute resolution procedures, or if I felt it was in need of other admins attention I would post on WP:AN. But first I would try to convince the person not to cause a problem. 1 != 2 17:09, 10 January 2008 (UTC)

Simple standards / process suggestion (Idiot proof)
Why not a very simple standard and process of:


 * 1) You ask for rollback and state a demonstrated need for it.
 * 2) Your contributions support that need. If you say "vandal fighting", which is the main reason for this, but have only like 1% of your edits RC patrol, vandal fighting, and AIV reports, that doesn't square up.
 * 3) Your request stays on the request page for at least 12 to 24 hours (lets pick one).
 * 4) You have 1000 edits and 4 months of service.
 * 5) No one objects and has that objection supported by others for 12 to 24 hours, you get the bit. - Key point.
 * 6) You can only apply once every 31 days.

Lawrence Cohen 16:38, 10 January 2008 (UTC)


 * Too many rules. Geez, I will need a checklist to make a decision. And that "No-one objects" thing seems odd, people always object to anything they are allowed to, that does not mean it should not be done.


 * You talk of idiot proofing, but I think our admins will work better in an environment designed for intelligent people than one meant for idiots. We are chosen for are ability to use discretion, I would rather address the behavior admins who abuse their discretion than take it away from all admins. 1 != 2 16:40, 10 January 2008 (UTC)
 * This seems like way too much process for something that is so easily removable. Any admin can remove the bit at any time. If anything, brand new (<10 edits, <4 days) editors should wait. Nakon  16:43, 10 January 2008 (UTC)


 * Idiot proof in that it's quite simple.


 * Step 1: ask for the rollback bit, tell us why you want it.
 * Step 2: Admins check you out for 24 hours, to give all time zones a chance to weigh in. Does your request match what you do?
 * Step 3: Do you have 1000 edits and have been here 4 months? Yes, and no one develops a consensus here to object? Congratulations, rollbacker.
 * Step 4: Archive.
 * Rules: 1000 edits plus 4 months of tenure. You can only apply once a month.


 * That's it. The steps aren't rules, they're a simple suggestion how to do this. Admins obviously are empowered to make some decisions, but anyone can object here--it's not an admins' only choice on who gets this. And since everyone seems to be fighting over how to give this out, why not just strip it down? People with experience ask and justify their reason: no one complains? Done. But it was just a suggestion. Lawrence Cohen  16:48, 10 January 2008 (UTC)


 * This sounds like a fine idea for an essay giving advice to admins, but I don't see the point in enforcing hard rules like X number of edits or Y number of months. An admin needs only enough edits to make their decision, which will vary from editor and admin. I also don't see the point in waiting 24 hours, if a consensus forms after the fact that the user should not have rollback it can be revoked. 1 != 2 16:50, 10 January 2008 (UTC)
 * If there were an approved, consensus standard of minimal thresholds for this (and basic ones, at that) then no one can complain when they get turned down, and admins shouldn't of course have a free hand to do anything they want with something like user rights. The idea is that some basic community standard exist of who gets it. Lawrence Cohen  16:53, 10 January 2008 (UTC)
 * People will complain regardless.  Nakon  16:57, 10 January 2008 (UTC)
 * Why be defeatist? :) Lawrence Cohen  16:58, 10 January 2008 (UTC)
 * The problem with coming up with specific rules comes when someone is turned down for something not on the list. Plus, it doesn't take 4 months to get used to reverting vandalism. Mr.  Z- man  00:05, 11 January 2008 (UTC)
 * I don't desperately require a certain number of edits nor a specific number of months of experience. Recall that three years ago we regularly gave the whole-hog admin bit to people with three months and a thousand edits.  What I would like to see is a 24-hour discussion period.  If nobody raises any complaints, we're good to go.  Honestly, is waiting a day really that bad?
 * I'd also suggest strongly that editors requesting rollback show good use of edit summaries. Any edit that's undone using the default summary or Twinkled without any added summary absolutely must be vandalism, because we know that those are the same edits that are going to get rolled back.  Reverts without edit summary – particularly those that aren't vandalism – ought to be bright red flags. TenOfAllTrades(talk) 17:08, 10 January 2008 (UTC)


 * People will still complain.1 != 2 17:10, 10 January 2008 (UTC)


 * People will always complain. I'm asking if most people think those might be reasonable requirements. TenOfAllTrades(talk) 17:13, 10 January 2008 (UTC)
 * People can complain to their hearts content, but that won't stop others doing what needs to be done. Hollow complaints have no value. Lawrence Cohen  17:22, 10 January 2008 (UTC)
 * These are good suggestions that admins can use for guidance. I don't think they are good requirements considering how easily the tool is removed and how minor the potential for damage from this tool. 1 != 2 17:15, 10 January 2008 (UTC)
 * And what harm comes from making people wait, or establishing a simple baseline threshold for who can get it? If the community decides to support standards like this for admins and applicants to go by, why is this bad? The community can reign in things like this, regulate it, or do anything else it feels it needs to do with consensus. Lawrence Cohen  17:24, 10 January 2008 (UTC)


 * Implementing a 24-hour waiting period (or even a 12-hour one) is something that needs to be done through policy, not just taken as 'guidance'. Are there any objections? TenOfAllTrades(talk) 17:20, 10 January 2008 (UTC)
 * Very good idea. Lawrence Cohen  17:22, 10 January 2008 (UTC)
 * I think that is far too long. If every rollback request sits for a day while numerous people comment on it, it becomes a big deal and an admin summarily removing rollback is now "overriding consensus." Mr.  Z- man  00:05, 11 January 2008 (UTC)

Not RFA
It's an undo button that works three times faster. Everyone calm down. It's gunna be okay.  Lara  ❤  Love  16:54, 10 January 2008 (UTC)


 * Which seems to say all that needs to be said. DuncanHill (talk) 16:58, 10 January 2008 (UTC)
 * Lol @ everyone getting worked up over nothing.  Majorly  (talk) 17:08, 10 January 2008 (UTC)
 * Who thought it was an RFA? :) Rudget . 18:08, 10 January 2008 (UTC)

We don't need to leave requests up for 24 hours to allow all timezones to weigh in. It doesn't take 24 hours and a boatload of admins to determine if someone is responsible enough to revert in one edit as opposed to three. This isn't RFA. It's a speedy revert button. *GASP*. The crystal-balling is horrible over this.  Lara  ❤  Love  19:30, 10 January 2008 (UTC)

Is Requests for Rollback broken?
After the perennial "Is Requests for Adminship broken?" threads, we now have the very first "Is Requests for Rollback broken?" thread! Actually, the question might better be "Has Requests for Rollback been fully constructed yet?" Carcharoth (talk) 17:07, 10 January 2008 (UTC)


 * Nothing in Wikipedia is fully constructed yet. I say we wait a week and see how things are going and let the policy develop based on experience. 1 != 2 17:17, 10 January 2008 (UTC)
 * I agree: give this time, and it'll be sorted out. Don't worry. :) Acalamari 19:33, 10 January 2008 (UTC)
 * Looking at the rights log, I see it's mainly admins who grant (or revoke?) this status. Is it that, or are they part of some othe group I'm unaware of? Rudget . 19:35, 10 January 2008 (UTC)
 * No, anyone with sysop rights can do it. Mr.  Z- man  00:07, 11 January 2008 (UTC)

Removing rollback from admin tools
Is it possible to get rollback removed from the admin tools? I'm very unlikely to ever use it, and I'm always worried I'll accidentally click a rollback link. Alternatively, do scripts exist that will hide the rollback links wherever they appear? Carcharoth (talk) 17:16, 10 January 2008 (UTC)


 * No, the admin group and the rollbacker group has the rollback right. If you are in one group you have the right. The only way to take rollback from an admin is to take away the admin bit. As for a script, I think it could be done in javascript without too much difficulty. 1 != 2 17:18, 10 January 2008 (UTC)
 * in your monobook css should fulfill the latter request, at least. – Steel 17:20, 10 January 2008 (UTC)
 * Right. I've heard bad things about changing the monobook css. Would you (or anyone) be able to talk be through it on my talk page later? I'm not going to be able to reply straightaway, but any advice there would be appreciated. Carcharoth (talk) 17:30, 10 January 2008 (UTC)


 * Fiddling with your monobook.js can lead to being unable to edit your monobook.js or do things you did not want to do(if the code does that), but monobook.css only changes the styles of how things appear. CSS cannot make you do edits or anything like that. Just adding the line should be plenty safe. JS is a programming language that can make logical decisions, CSS is just a lookup table your browser uses to decide how to render things. 1 != 2 17:32, 10 January 2008 (UTC)
 * Ah, right. Thanks. That sounds a lot more like what I was thinking of. Is this all explained somewhere? I see I've never turned on Navigation popups, so would anyone recommend that? Does the basic installation of that mess anything up? Carcharoth (talk) 17:34, 10 January 2008 (UTC)
 * If you do accidentally click the rollback, just re-edit the page back to what it was and apologise in the edit summary. That's what most of us admin's do.  If you meant to make the edit, then remake the edit properly with the proper edit summary. Hiding T 22:36, 10 January 2008 (UTC)
 * Thanks, but I prefer not to have little time-bombs lying around that I can accidentally click on. I like all the clicks I make on Wikipedia to take me to a screen where I can review my edit before committing changes. That's why I don't think I've ever used rollback and don't intend to, even though it is part of the admin tools I have. Undo is perfectly fine for me. Carcharoth (talk) 00:02, 11 January 2008 (UTC)

Disputed policy?
This is not a policy page, it is a noticeboard for people to get attention. The process described is for the management of the noticeboard, As far as I can tell there is yet to be a codified policy. I don't see the point in putting a policy disputed tag on something that is not a policy page. 1 != 2 18:06, 10 January 2008 (UTC)


 * Maybe it should be changed to a "disputed process" template, then? Whatever is happening here, it's quite clearly disputed, and that should be mentioned on the page. --Conti|✉ 18:18, 10 January 2008 (UTC)

Archive
Well, it seems one or two people are refusing to accept the tagging and archiving mechanism. Can we make it clear that this is what we want? I'll put it in the header so people are aware exactly what to do should we come to a consensus here. To make it clear, I suggest we tag requests as done or Not done and then requests are archived by a bot. Now, for the reasons that we need a bot, I quote myself earlier; So we've got a clear record of who has been granted the tool and who hasn't. It'll make it far easier to evaluate the tool and see if there are any discrepancies in who grants it and for what reasons. It gives the page some evaluation and can also help to look at trends for who requests the tool and when.  Ry an P os tl et hw ai te  18:30, 10 January 2008 (UTC)


 * I object. This is simply instruction creep. The logs give a record of who has the tool. Tell me why we need more paper work? Why is it useful? Besides which, the tool can be granted without reference to this page. Or is that to be banned by more instructions? I predicted that this would lead to instruction creep and a whole load of processes, and the proposers kept telling me I was wrong. It seems I was right - process is exactly what they want, down to me being slapped for using the wrong tick.--Docg 18:34, 10 January 2008 (UTC)


 * The current archive isn't particularly useful for that. There's not even a one-click way to see who has granted rollback to a user. Special:Log is currently superior, listing also requests that were granted without the use of this page. Kusma (talk) 18:35, 10 January 2008 (UTC)


 * Precisely, it is comprehensive and there is no paperwork.--Docg 18:38, 10 January 2008 (UTC)


 * Yeah, I'm strongly on the side of using the logs and not trying to maintain a separate list also. Unless there's some clearly compelling reason for it, which I do no yet see.  Friday (talk) 18:40, 10 January 2008 (UTC)


 * An archiving similar system similar tp RFPP would probably work better, where requests a left up for a period of time after they've been dealt with to give users chance to pop back then they could be archived periodically (once/twice a day) without hassle.  Ry an P os tl et hw ai te  18:44, 10 January 2008 (UTC)


 * Why do we need an archive? Wikipedia never invents paperwork without damn good reasons. You have been criticising me for not conforming to the process, but you have no agreed process and no argument for one. Please explain why we need to archive anything?--Docg 18:49, 10 January 2008 (UTC)


 * An archive is good, especially for denied requests. If you don't like the archive system, go and propose that changing usernames aren't archived either. This isn't the only thing.  Majorly  (talk) 18:47, 10 January 2008 (UTC)


 * Why is it good? And no, this has nothing to do with usernames, we don't archive blocks or protections, we have a log. This is instruction creep gone wild.--Docg 18:49, 10 January 2008 (UTC)
 * When I say RFPP, we allow the bot to organise a page. We don't need an archive per se, just something to keep it in order.  Ry an P os tl et hw ai te  18:51, 10 January 2008 (UTC)


 * Doc, we know you don't like it, so I suggest you stop coming to this page so it doesn't upset you anymore. Thanks,  Majorly  (talk) 18:51, 10 January 2008 (UTC)


 * That's civil. As stated from the beginning this tool is "no big deal" so I see NO reason to invent more process. I agree with Doc... and this is exactly why I opposed the proposal as well. Ideally this page disappears and every account eventually gets rollback, but until then, I'm in support of anything that will reduce this bureaucratic nonsense. Why archive denied requests? This is a 5 minute job right? Seems it would take just as long to dig through an archive to see if a request was denied as it would to recheck their logs.  Justin  chat 18:53, 10 January 2008 (UTC)


 * Doc clearly hates the idea of giving rollback (despite the fact he's given it to several users), and now he's simply complaining about each and every thing he can possibly think of, and frankly I'm tired of it. If you don't like it don't come here.  Majorly  (talk) 18:57, 10 January 2008 (UTC)
 * Not helpful. Friday (talk) 18:58, 10 January 2008 (UTC)
 * My problem with the archive is that the editors granted rollback are merely listed, while those not granted rollback have their whole discussions archived. I tend to think that all discussions should be archived, not just names.--CastAStone//(talk) 18:52, 10 January 2008 (UTC)

The user rights log is plenty. We don't need to be producing dozens and dozens of new non-encyclopedia pages. Mahalo. --Ali&#39;i 18:53, 10 January 2008 (UTC)
 * Then go and propose all the user rename archives are deleted too, if it bothers you so much. Many thanks,  Majorly  (talk) 18:54, 10 January 2008 (UTC)
 * User renames have nothing to do with rollback. Perhaps instead of making falacious arguments, you could actually propose why an archive is important, besides "this other unrelated procedure has an archive".  Justin  chat 18:57, 10 January 2008 (UTC)
 * Why do they get archives and this one doesn't? There's no difference at all.  Majorly  (talk) 18:58, 10 January 2008 (UTC)
 * Each procedure must show their own merits for having X process. We don't invent process because another process like it exists.  Justin  chat 19:00, 10 January 2008 (UTC)
 * One for denied requests will show why the user was denied. When they come back after a few months, they can easily reference their request, and show they've improved.  Majorly  (talk) 19:04, 10 January 2008 (UTC)
 * The log allowed for a reason to be given also. Friday (talk) 19:05, 10 January 2008 (UTC)
 * There is no log for _not_ giving someone a permission change. —Random832 19:13, 10 January 2008 (UTC)


 * Yes, there is. That's what I said right below.  See here.  Friday (talk) 19:16, 10 January 2008 (UTC)

For what it's worth, a null permissions change can be logged to denote someone being turned down for the flag. Friday (talk) 18:55, 10 January 2008 (UTC)

How about simply using edit summaries? When granting or refusing a request, say so in the edit summary, and the the page history is all the archive we need. --Tango (talk) 19:06, 10 January 2008 (UTC)
 * Less good than the log. The page history is already quite large compared to how much will ever be in the log for a single user.  Friday (talk) 19:07, 10 January 2008 (UTC)
 * That's how we "archive" WP:AIV. Does anyone want to propose an archive of denied requests for that page? Kusma (talk) 19:08, 10 January 2008 (UTC)
 * I don't think recording a refusal in the log is a good move. It will surely be held against the editor, for example at RfA, and it will not be useful for much else. Perhaps a rejection would be better on their talk page in addition to this page. I like the idea of RFPP style archiving, where requests could be kept on the page for about a day and then removed completely without archiving. It allows the requester (and others) to catch up on any discussion without having to browse all through the history. -- zzuuzz (talk) 19:24, 10 January 2008 (UTC)
 * You're right, AIV is a good example where archiving isn't appropriate. With many of the users on Wikipedia having a desire to use rollback, there is a good chance that this page will be outrageously backlogged within no time at all. Would the bot (as Ryan mentions) be a good idea? Although, I can forsee, the bot, needing to archive on at least a two-hourly basis. So what I'm trying to say is, although archiving is a good idea (especially with a bot on hand), I just don't think it's appropriate for the page. Any other thoughts? Rudget . 19:33, 10 January 2008 (UTC)

I agree with Doc, Kusma, Friday et al. in that there is no need to log a refusal, nor is there a need to keep a record; the user rights log is sufficient. As long as this functionality is available in wiki, I see no issue giving it to users via e-mail requests, or even bestowing it unasked for on trustworthy users. There is as much "consensus" for how it should be given as there was for if it should be given. -- Avi (talk) 20:01, 10 January 2008 (UTC)
 * Even though off-topic. Nice grin. ;) Rudget . 20:09, 10 January 2008 (UTC)


 * I agree with Doc., The proposal that allegedly got consensus did not require an RFR page, and proposed that admin's just hand them out as they saw fit. Wikipedia is not a bureaucracy. You have an edit history and a log. That should suffice. Hiding T 22:41, 10 January 2008 (UTC)

A more different archive
I cleaned up this page's archives, moving them sequential numbered pages and adding navigation, above. If you were looking for the dated archive pages, that's where they went. Best, UltraExactZZ Claims~ Evidence 19:13, 10 January 2008 (UTC)
 * Did you talk to ST47? That might confuse the bot. Mr.  Z- man  00:08, 11 January 2008 (UTC)
 * Fumble. I thought the bot was archiving the requests on the main page, not anything on the talk page. I'll message him. UltraExactZZ Claims~ Evidence 01:05, 11 January 2008 (UTC)
 * Oh, the talk page, never mind, I thought you were referring to the request archive. Mr.  Z- man  01:42, 11 January 2008 (UTC)

Required?
Does one wishing to obtain rollback have to request here? Or can any admin at any time give it out? I ask because I wanted to give it to my old adoptee, but he has twinkle and probably wouldn't request it, but I want him to have the option available. He has a good reverting history, and a good vandalism fighting record. More importantly, how would I give it to him? Are admins able to take it away as well? J- ſtan ContribsUser page 21:02, 10 January 2008 (UTC)
 * I should think if with good faith (and without prior history of content disputes) the user should be given the rollback permission. And per the question about revoking it aswell, yes that is possible. Rudget . 21:05, 10 January 2008 (UTC)
 * Any admin can give it out on request.--Docg 21:06, 10 January 2008 (UTC)
 * (ecx2)In order - Ideally here, but you can just do it. Yes any admin can do it. You can give it to him via Special:Userrights. And yes, you can take it back again. Pedro : Chat  21:06, 10 January 2008 (UTC)
 * Cool, thanks. Yeah, I know my old adoptee's edit history well, and I can't think of a serious dispute involving editwarring. J- ſtan ContribsUser page 21:07, 10 January 2008 (UTC)


 * I'm going to make it my policy to ask any sensible editors I find or know of whether they want rollback. If they do, I'll give it them, no jumping through hoops. Surely we could save a lot of bureaucratic nitwittery if rollback was handed out to everyone vaguely clueful ASAP. Angus McLellan (Talk) 21:28, 10 January 2008 (UTC)
 * Yeah. Sensible editors should just get it. J- ſtan ContribsUser page 21:32, 10 January 2008 (UTC)
 * I've only answered requests at RfR for now, but if I'm familar with a user, and know of their history, I don't mind giving them rollback if they ask for it. If I don't know them, I'll send them here or research them myself. However, if it becomes policy that rollback can only be granted here, then I'll follow that. Acalamari 21:36, 10 January 2008 (UTC)
 * See also: earlier discussion on this at . – Steel 21:58, 10 January 2008 (UTC)
 * I like the idea of making this function like requesting protection. At least at first, we probably shouldn't require a request here, as there are so many trusted users. Maybe in the future. J- ſtan ContribsUser page 22:22, 10 January 2008 (UTC)

Why is the page s-protected?
If any user is permitted to comment on a request (including IP users), I don't know that s-protection is appropriate. If the reason for s-protection is short-term concern about trolling/vandalism, I can understand that, but we probably should add a note to that effect saying that if an IP user would like to comment, he/she may do so on the talk page. --B (talk) 22:03, 10 January 2008 (UTC)
 * I move-protected the page, as there's no reason to move it, but Nakon's reasoning for semi-protection was because IPs can't receive rollback, and therefore, won't need to edit the page. I personally think that IPs should edit the page in case they have information to offer on a user. Acalamari 22:07, 10 January 2008 (UTC)
 * I sprotected the page as I noticed an IP impersonating a user and by the fact that IPs can not be given +rollbacker (or any other bit through Special:Userrights). Nakon  22:49, 10 January 2008 (UTC)
 * Impersonation is a concern, but there is the page history to look at. Acalamari 22:52, 10 January 2008 (UTC)

Consensus?
Where exactly was determined consensus for approval of this proposal on rollback given on individual basis? Last time I saw it there wasn't the slightest consensus for approval. Hús ö  nd  22:20, 10 January 2008 (UTC)
 * This is a wiki, we just did it. If you don't think we filled in the right forms along the way, well, never mind – Gurch 23:09, 10 January 2008 (UTC)

Need a better process
I think we may need a more formal requests for rollback process. User:Platypus222 made a request there today. I briefly reviewed his recent contribs, and found a couple of edit summaries that I wasn't really happy with, e.g.  - but as I was posting them to the page for other admins to consider in deciding whether to grant the request, I was edit-conflicted by Alison granting the request and by the bot removing the successful request from the page. I know my concerns weren't necessarily serious enough to refuse rollback, but I think we should have a brief discussion, giving users time to raise relevant evidence, before actually deciding whether to grant the rollback status - maybe a bit like an XfD. WaltonOne 22:27, 10 January 2008 (UTC)


 * As has been said before - it's just a faster undo button. DuncanHill (talk) 22:30, 10 January 2008 (UTC)
 * There is normally a discussion if someone spots something, but if you notice something after an admin has granted rollback, just take it up with the admin, like I did with Doc glasgow earlier. Acalamari 22:35, 10 January 2008 (UTC)
 * It isn't really important enough to do that in this case, but I still think we need to take more time over each request. WaltonOne 22:36, 10 January 2008 (UTC)
 * I do agree that we should look carefully at a user's reverts, and the quick removal of rollback isn't an excuse to sloppily hand it out. Acalamari 22:44, 10 January 2008 (UTC)
 * I agree with Duncan and Acalamari. The idea of having a more formal process was already thrown out when this whole proposal was first discussed by the community. It's a very minor tool and it's easy to take away, so there doesn't need to be any delay while we discuss it. I check the user's block log, talk page, and recent edits. That tells me whether they can be trusted with it and whether they need it, and that's all I need to know. Edit summaries don't matter. Quality of contributions don't matter. If it's abused, it can be taken away just as easily. Kafziel Talk 22:38, 10 January 2008 (UTC)
 * Yes, this isn't like adminship where, if someone starts abusing it after promotion, we have to go through ANI, RFC, and ArbCom before its finally removed. With rollback, if someone abuses it, they get it taken away as quickly as it was granted. Acalamari 22:42, 10 January 2008 (UTC)
 * I don't even think we need this process. We could have the category for admins who are willing to give it out, and a category for those wishing to receive it. This current process seems to add too much bureaucracy. J- ſtan ContribsUser page 22:53, 10 January 2008 (UTC)
 * No, I do believe we need this page. Not everyone will want to ask an admin directly, and a page like this is fine. Also, on this page, it's possible to get an immediate second opinion if one admin spots something that another admin didn't. Acalamari 22:57, 10 January 2008 (UTC)
 * Also, I just noticed a good reason why this page is a good idea: . Acalamari 22:58, 10 January 2008 (UTC)
 * LOL, that instance involved me! And I should have made clear that asking directly isn't required, if they're in the category, they're eligible. But I see what you're saying. J- ſtan ContribsUser page 23:00, 10 January 2008 (UTC)
 * Yes, well, with this event, that's a very good reason for this page to exist. :) More pairs of eyes! Acalamari 23:12, 10 January 2008 (UTC)

Responding to Duncan Hill "As has been said before - it's just a faster undo button" - rollback is not just a faster undo button. Unless you are using a script to modify the URL, then rollback doesn't allow editing of the edit summary, whereas with undo you can edit the edit summary. That's one of the big differences. Carcharoth (talk) 23:58, 10 January 2008 (UTC)
 * What are the other big differences? DuncanHill (talk) 01:16, 11 January 2008 (UTC)
 * Doesn't provide a diff showing what you are reverting. Try it out at New admin school/Rollback and see the difference. Carcharoth (talk) 01:36, 11 January 2008 (UTC)
 * I was planning on trying it on some vandalism - but there seems to be less than usual affecting articles on my watchlist. DuncanHill (talk) 01:48, 11 January 2008 (UTC)

Removing rollback after it is granted
Other than civility (which is important, of course) is there any reason why Admin B should not remove a rollback granted by Admin A. More importantly, is this an "administrator action" that could be part of a wheel war? -- Avi (talk) 22:59, 10 January 2008 (UTC)
 * I don't see how civility is anything to do with this. If there's edit warring with it, then it should be removed, most probably without discussion (obviously it can be reported to AN/I if needed). A wheel war would be when admins grant and revoke the tool with no extra circumstances involved.  Ry an P os tl et hw ai te  23:02, 10 January 2008 (UTC)
 * There are lots of reasons other than civility, obviously. Vandalism, for a start, and if the user is blocked for any reason an administrator may well decide to remove rollback at the same time as their ability to edit. Obviously it's an administrator action; if there is a problem with that it is a problem with the conduct of individual administrators and not the process, and should be dealt with in such a manner – Gurch 23:04, 10 January 2008 (UTC)
 * I took Avraham's wording to mean civility in terms of politeness between administrators (more like Wikiquette), not civility as in a reason to remove the rollback feature. If the question is whether admin B can take back what admin A has granted, as far as I'm concerned, anyone is free to rescind rollback from anyone I've given it to. I certainly wouldn't expect them to notify me. Honestly, I'll probably do hundreds of these things, and I won't remember any of them. Kafziel Talk 23:07, 10 January 2008 (UTC)
 * Kafziel, you divined my intention of the use of the word "civility" perfectly [[image:face-smile.svg|25px]]. -- Avi (talk) 00:04, 11 January 2008 (UTC)
 * As far as I am concerned, any administrator may reverse an action of mine, I have no problems with that: I will not wheel-war. If I grant someone rollback, and another admin feels that that granting may not be a good idea, I won't object to the action being undone. The only thing I would like is a courtesy note. Acalamari 23:28, 10 January 2008 (UTC)

It is probably best for administrators who are handing out lots of rollbacks to have a note at the top of their talk page saying "feel free to revert any rollbacks I've granted without notifying me" (or something similar), or to note this in the log (the best method). Anything beyond that (ie. a back-and-forth grant-revoke-grant-revoke cycle) should be avoided without further discussion. Carcharoth (talk) 00:10, 11 January 2008 (UTC)

Overzealous bot
Can we please slow the bot down a bit? It archives too quickly, and some requests might need a couple more eyes. J- ſtan ContribsUser page 23:03, 10 January 2008 (UTC)
 * Agreed: I just restored a disputed request. It would be nice if the bot was a bit slower. Acalamari 23:05, 10 January 2008 (UTC)
 * I asked that up there ^ :/ Viridae Talk 23:06, 10 January 2008 (UTC)
 * I'm not sure how many times the bot operator needs to be asked before they'll do it, but I asked them again anyway – Gurch 23:08, 10 January 2008 (UTC)

I suggest that this be taken to AN or ANI if this is not resolved shortly. --Rschen7754 (T C) 00:07, 11 January 2008 (UTC)

Accepted rollback requests
I notice we archive the denial requests, but not the accepted ones. They should be done, especially in those that more than one administrator discussed before granting without having to dig through the page history. -- ReyBrujo (talk) 23:24, 10 January 2008 (UTC)
 * I don't know if that is necessary, those with rollback granted show up in the logs. —  Save_Us _ 229  00:21, 11 January 2008 (UTC)
 * Yes, but it only shows who granted the rollback, not whether there was a long discussion to get him the bit or it was a straight issue. -- ReyBrujo (talk) 03:11, 11 January 2008 (UTC)
 * Now I understand. Well, it can be done, just I don't know if we are going to be willing to make that many archives. Over the last 3 days or so we have made 300+ rollbackers (or about 100 requests per day). I think going through Requests for rollback/Approved and seeing what date they were approved and going through the history to see the archived bit is enough, that's just my two cents. 100 per day could get tedious if we really wanted to do that tasks. Ask the owner of User:RFRBot what he thinks. — Save_Us _ 229  04:04, 11 January 2008 (UTC)

A few questions

 * 1) Why doesn't the Template:Rfr have any usage documentation like the other policy related ones?
 * 2) Also I created a redirect Template:RequestsForRollback which I thought would be helpful, protect maybe?  VivioFa teFan   (Talk, Sandbox) 00:37, 11 January 2008 (UTC)

Rollback Question
This comment was part of one of the Requests for Rollback:

"Rollback will save a lot of time over the manual methods, which makes warning for every vandal revert extremly tedious."

I get the impression from this comment that Rollback posts a warning to the vandal, as well as doing the revert. Is this the case?

Wanderer57 (talk) 02:32, 11 January 2008 (UTC)
 * No. Rollback only reverts the edit. Warning is done separately. Sean William @ 02:33, 11 January 2008 (UTC)
 * No, it doesn't. Nakon  02:33, 11 January 2008 (UTC)
 * Rollback does the following:
 * appear only next to the top edit
 * revert all top consequent edits made by last editor
 * work immediately, without intermediate confirmation diff page
 * add automatic edit summary "Reverted edits by Example (talk) to last version by Example2", marking edit as minor
 * Warnings need to be handed out by the reverter. — Save_Us _ 229  02:35, 11 January 2008 (UTC)


 * Thanks muchly for the answers.  Wanderer57 (talk) 02:51, 11 January 2008 (UTC)
 * Though it does give a link to the talk page of the user who was rolled back on the confirmation page. Mr.  Z- man  03:19, 11 January 2008 (UTC)
 * To add to that, if you get a tool like WP:TWINKLE (which anyone can use), it has vandal-warning functionality. The only downside (or upside?) is that you have to use Firefox in order to use it. --B (talk) 05:43, 11 January 2008 (UTC)
 * <ad>amelvand works in IE</ad> :P Grace notes T § 06:10, 11 January 2008 (UTC)

Arbitration
Following Jimbo's suggestion, I have asked for arbitration of the matter of consensus at WP:RfArb —Preceding unsigned comment added by Doc glasgow (talk • contribs) 00:55, 11 January 2008 (UTC)
 * Thanks for filing the case, Doc. You may get a lot of flack for it, but you have my support. I'll try and add a statement at some point. You may be interested to know that the number of rollbackers is now at 350+. I was going to start a sweepstake as to when the number of rollbackers would exceed the number of admins, but I think that would be distracting at this point in time. :-) To reduce drama, it might be a good idea to make clear (if you agree) that reversing what has happened is not the intent here, but merely asking what should have been done, and what should be done in future. My feeling is that the breakdown in communication occurred when Ryan posted in the bugzilla thread that he personally saw consensus and asked the developers to have a look and judge for themselves. What should have happened instead was for uninvolved en-wiki bureaucrats to be asked judge the consensus, and for the result of that judgement to be posted at the bugzilla thread. Ideally, the bureaucrats would have been asked to 'clerk' the poll beforehand, so they could remain uninvolved if needed. Actually, I think I've just written my ArbCom statement! :-) Carcharoth (talk) 01:14, 11 January 2008 (UTC)
 * My preference would now to be to move to an autogrant system of some type. That removes this bastard process without removing anyone's rollback. However, since rollback is no big deal, I see no issue with removing it. As long as this process of admin rights is terminated soon.--Docg 11:06, 11 January 2008 (UTC)

Stop the process
This is an astonishingly bad idea being made even worse through poor execution. We need to put this process on hold immediately until we can sort out all the many questions. Rossami (talk) 01:20, 11 January 2008 (UTC)
 * Please point out one problem that would be solved by stopping this. John Reaves 01:21, 11 January 2008 (UTC)
 * I think stopping the process would do more harm than good at this point. We're not having any major catastrophes thus far, administrators can grant and remove rollback bits and the entire feature can be removed in a click or two by developers. Discussing it in mid-progress does no harm. — Save_Us _ 229  01:25, 11 January 2008 (UTC)
 * Please point what harm would be caused by stopping this at this point. Jons63 (talk) 01:29, 11 January 2008 (UTC)
 * I would like to know too. I think it is going fairly well, with a few minor problems, so far.   jj137  ♠ 01:31, 11 January 2008 (UTC)
 * For starters, the rollback is being given out only a minute or 2 after the request was made. That isn't enough to give any consideration to the users contribs, talk page etc. RxS (talk) 01:33, 11 January 2008 (UTC)
 * That plenty of time, and also there's no point in checking users you are already familar with,  Majorly  (talk) 01:36, 11 January 2008 (UTC)
 * Really, you can tell if an editor of 2 years (in one example) has been involved in edit wars in 2 minutes? Ok....RxS (talk) 01:42, 11 January 2008 (UTC)
 * I can't imagine any user who has never been in an edit war, ever. I think most admins have too, so whether they have or notis irrelevant - unless, of course it's recent. Which users do you have issue with?  Majorly  (talk) 01:45, 11 January 2008 (UTC)
 * You'll note that I said edit wars...2 minutes isn't enough to determine if an editor has a history of edit warring. RxS (talk) 02:31, 11 January 2008 (UTC)
 * 2 minutes is plenty of time to see if an editor has a recent history of edit wars, and that's all that matters. You don't need to go back through 2 years' worth of history; we don't even do that on RfAs. The last couple of months are all that matter. Kafziel Talk 02:41, 11 January 2008 (UTC)
 * Probably a lot of bad feeling on the part of those who see others with a tool they don't have (I know they can use scripts, but that's human nature). Ripping the tool away from 300+ people who have it would also be a bad idea. I think the process can be sorted out on the fly, so let's concentrate on doing that (sorting it out). Carcharoth (talk) 01:33, 11 January 2008 (UTC)
 * It really only takes me maybe three minutes at most to go through past 1000-2000 contribs (the edit summaries say a lot), check the block log, and usually the user page, talk page, and/or edit count. Plus, if I am already familiar with the user, there's no point waiting.   jj137  ♠ 01:40, 11 January 2008 (UTC)
 * I think Carcharoth was about to say what I was going to Jons63. If your going to have 300+ users removed of a tool, there should be either a problem with the tool a serious misuse of it, neither of which are occurring that I know of. Are you willing to risk the assumption of bad-faith being given to that many editors who were just trusted with it? The tool can be removed from someone as easy as it can be granted, I don't see the emergency in disabling it. — Save_Us _ 229  01:46, 11 January 2008 (UTC)
 * Reviewing contribs is looking for misuse of a rollback script or edit summaries. Edit warring and incivility and other things that would tank and RfA are not relevant here. John Reaves 02:03, 11 January 2008 (UTC)
 * It gets taken away from me, I go back to how I use to fight vandalism and move on. No problem, no harm.  Jons63 (talk) 02:04, 11 January 2008 (UTC)
 * I do take random diffs when reviewing contribs.   jj137  ♠ 02:06, 11 January 2008 (UTC)

Jimbo says: "I recommend that people basically do nothing at all here, i.e. please don't go awarding this ability to lots of people in an effort to create "facts on the ground" about how it is used. :-)" &mdash;Ashley Y 02:19, 11 January 2008 (UTC)
 * Jimbo Wales doesn't reflect the entire communities perspective as a whole. Granted he has more say than probably anyone, but if you read furthur in his post rather than pulling one excerpt he goes on to say that "I can not guarantee that the Foundation will agree, as I am only one of 7 board members, and not involved in management at all, but I consider it highly unlikely that they would disagree with a formal decision of the community." If the community as a general whole feels the same way, Jimbo does not have a veto to disregard us whatsoever. — Save_Us _ 229  02:24, 11 January 2008 (UTC)


 * He's saying that the Foundation has the last word on whether this gets implemented in the end, not the community. In the mean time, we should probably follow his advice for a temporary moratorium while ArbCom reviews the matter. &mdash;Ashley Y 02:28, 11 January 2008 (UTC)
 * Hence why I pulled out the bit about I consider it highly unlikely that they would disagree with a formal decision of the community, like Jimbo, I feel the Foundation wouldn't force an issue if the community felt otherwise. — Save_Us _ 229  02:32, 11 January 2008 (UTC)
 * I'd like to know if there more users abusing this feature than improving the encyclopedia. Lord Sesshomaru (talk • edits) 04:56, 11 January 2008 (UTC)
 * That was exactly what Jimbo's point was. You can't do something out of consensus, implement it at breakneck speed, and then say "See it works fine!". Setting that kind of precedent is appalling. I may have to say this 1000000 times... the issue is not the tool, it's how the tool was implemented.  Justin  chat 05:05, 11 January 2008 (UTC)
 * Most of the edits made using this feature are improving the encyclopedia, so you may put your concerns to rest. Grace notes T § 06:18, 11 January 2008 (UTC)
 * If it does work fine, and as Gracenotes says is improving (or rather stopping damage to) the encyclopedia, what is the problem? The precedent is already set. Mr.  Z- man  06:31, 11 January 2008 (UTC)


 * Considering that our edit warring policy is firmly grounded in the principle that you can do things the wrong way even if you are right, it's kinda scary to see people going "well, yes, we did this without consensus, but that doesn't matter because now it's implemented!" -Amarkov moo! 06:34, 11 January 2008 (UTC)
 * Nothing like advocating policy by force.  Justin  chat 07:09, 11 January 2008 (UTC)

It's just rollback. Why is everyone seeing ghosts here? We don't need a big complicated process and we sure as hell don't need ArbCom or even Jimbo. This is what I both love and truly hate about Wikipedia: Every imaginable part of the site and project is debated to death, usually two or three times. As for the vote that has been initiated - voting is evil. Now let's stop playing debate club and junior UN and get back to writing the encyclopedia. That is why we are here, right? EconomicsGuy (talk) 07:17, 11 January 2008 (UTC)

For those who don't know
Requests for rollback/Vote. I guess this is our alternative to those who think that consensus was not carried out and/or they see a problem. — Save_Us _ 229  06:59, 11 January 2008 (UTC)

Overkill
This is an entirely unnecessary tool. Rules and guidelines already exist for vandalism and reverting, deletions, etc. I see this being used as a tool by disgruntled dogmatic and dictatorial editors to eliminate instantly things that they don't like. It is a form of enforcing personal opinions by yet another of the countless overlapping Wikipedia guidelines. Regards, David Lauder (talk) 10:47, 11 January 2008 (UTC)
 * See below. Using rollback in content disputes or when a rationale would be required such as non-obvious or borderline vandalism has always been disallowed. You're not really supposed to use popups, undo or TW without a specific rationale in those situations either. Nothing has changed - we just got a faster undo button next to the latest edit on the article. EconomicsGuy (talk) 11:06, 11 January 2008 (UTC)

Reminder

 * Help:Reverting < --- Page you should have read before making a comment on this page

People are expected to follow that whether they are reverting via rollback or via other means such as scripts or the history link. There is nothing controversial, disruptive, or in any way problematic if people observe that linked text. If they don't WP:3RR applies. WP:3RR is not a license for three reverts.

Suggestion: Have two rollback's one for removing nonsense (vandalism, spam, etc) and second for everything else which leaves a note to check the talk page.

-- Cat chi? 15:17, 11 January 2008 (UTC)
 * Only if you then go and leave a note on the talk page. Which means the time saved using rollback instead of a manual edit is an insignificant saving. Don't try and extend rollback like this. Normal editing is fine for non-vandalism. Carcharoth (talk) 15:27, 11 January 2008 (UTC)
 * Right. It is something I would never do. Was just a suggestion... Perhaps a compromise. -- Cat chi? 17:12, 11 January 2008 (UTC)


 * Simply append "&summary=" to the existing rollback link url to do this. you could add a script that does that. —Random832 17:45, 11 January 2008 (UTC)
 * Also note that the rollback URL includes the name of the user being rollbacked, so it is possible to include the name of the said user in the summary without too much lookup effort. Grace notes T § 17:47, 11 January 2008 (UTC)

Users with Rollback Category ?
Would it be a good idea to create this category and add users to it as part of the process of giving the privledge to them? Russeasby (talk) 17:47, 11 January 2008 (UTC)
 * No need. See here. Part of Special:Listusers. Carcharoth (talk) 17:54, 11 January 2008 (UTC)
 * Actually, it would seem that someone beat Russeasby to the category idea Category:Wikipedia users with the ability to rollback, although not an automatic thing like he was saying.  MBisanz  talk 01:41, 12 January 2008 (UTC)

Wow
I've only begun testing it for real some time ago on vandalism (after some month of not opening vandal fighter due to a diminishing lack of interest) and it's really cool, honestly... my interest in fighting vandalism has been renewed... I'm not sure why all this fuss but I now think that rollback can actually help WP not harm it since non-admins would be more willing to combat it (since javascript is MUCH slower). -- Menti  fisto  18:05, 11 January 2008 (UTC)
 * Rollback saves a mouse click or two, and the increased speed of execution is perceptible. These small gains have affected my vandal fighting behavior on WP, encouraging me to expand my watchlist specifically to fight vandalism on less watched scientific subjects.  The lack of rollback (for autoconfirmed members especially) is an artificial, and unproductive, throttle on our trusted user base. With it we can afford to relax our attitudes a notch, and take the time to be more magnanimous to new and anon editors. -- Paleorthid (talk) 19:15, 11 January 2008 (UTC)


 * Agree wholeheartedly with above comments, it's very useful. I've been reverting some odd, often overlooked vandalism in the image namespace (on image description pages) by IPs. It's a whole lot quicker with rollback. • Anakin (contribs • complaints) 22:40, 11 January 2008 (UTC)
 * This sort of thing for example: []. It happens all the time. Rollback makes reverting it quick as pie. Want to help? [ Click here!] • Anakin (contribs • complaints) 22:54, 11 January 2008 (UTC)

A suggestion
Please stop advertising all around Wikipedia (such as the Watchlist, or WP:CENT). Let this be for a moment, see what happens without advertising it, then allow people to decide whether it's good or bad. Rule of thumb: if no one notices any bad effects, it's probably good.

Without reading any of the discussion above, and violating assume good faith in a bad way, I have a strong feeling that all the fuss about this is an attempt to make something uncontroversial a big issue by a minority as a way of filibuster. Keep in mind that I'm not accusing any specific side here, as I didn't read the discussion, it may apply to any of the two. It's just a pattern that recurs very often, on many of the lamest edit wars. User:Krator (t c) 19:42, 11 January 2008 (UTC)

And there it goes...standards creep.
An editor in good standing has been refused rollback. He wants to fight vandalism, even adding twinkle to his monobook. But, despite being here for 6+ months, he has just 45 edits. So, he was refused. There's no reason to suspect this editor is anything other than a good editor. But, he's not good enough to have rollback. He was specifically denied because he had only 45 edits. And so it begins. Standards will continue to rise. --Hammersoft (talk) 20:12, 11 January 2008 (UTC)
 * Er, hang on - I didn't refuse it, just gave my opinion. If someone else wants to grant it, I won't complain.   This is the problem with introducing this system - no-one's really sure where they stand.  BLACK KITE  20:15, 11 January 2008 (UTC)
 * When you put not done, that's refusing it. If you want to leave an opinion, do so but don't use the not done tag.  Nakon  20:17, 11 January 2008 (UTC)
 * Ah, OK, my bad. I'd stick to my opinion though - only 16 article-space edits?  I don't see the need, or the experience.  BLACK KITE  20:19, 11 January 2008 (UTC)
 * Wants to fight vandalism. Userbox and twinkle install shows good intent. No reason to deny. --Hammersoft (talk) 20:20, 11 January 2008 (UTC)
 * Personally, I'd like to see a little experience actually using Twinkle first, but it is no big deal. Mr.  Z- man  20:21, 11 January 2008 (UTC)
 * As I said, I won't complain if someone wants to grant it.  BLACK KITE  20:22, 11 January 2008 (UTC)
 * I gave it to him. He seems to show a desire to help out the community through vandal-fighting. I can't say that he definitely won't abuse it, but for now, I see no reason not to grant it. J- ſtan ContribsUser page 20:22, 11 January 2008 (UTC)
 * (edit cornflakes) Have you seen this comment? I don't think I can support Bpeps having the rollback just yet. --Merovingian (T, C) 20:23, 11 January 2008 (UTC)
 * Oh yes, that comment right there should do it. Take it away immediately. <strong style="color:#fff;background:#DC143C;border:1px solid #000;font-family:Verdana ">ALLSTAR <strong style="color:#FFF;background:#0F4D92;border:1px solid #000"> echo 20:24, 11 January 2008 (UTC)
 * Done, but it's not like that's the only one. --Merovingian (T, C) 20:25, 11 January 2008 (UTC)
 * (e.c.)Well, that's why people have been arguing for more time limits before approving or more commentary by more than one admin. It took you longer apparently to find that than it did for another admin to approve the user. <strong style="color:#fff;background:#DC143C;border:1px solid #000;font-family:Verdana ">ALLSTAR <strong style="color:#FFF;background:#0F4D92;border:1px solid #000"> echo 20:28, 11 January 2008 (UTC)
 * It didn't take me that long; Bpeps has <100 edits. Also, the edit summary with "STALK" in it was hard to miss.  --Merovingian (T, C) 20:30, 11 January 2008 (UTC)
 * I suggest informing the user of this change. I admit I missed that edit. Acalamari 20:27, 11 January 2008 (UTC)
 * I mentioned it on the main project page. --Merovingian (T, C) 20:28, 11 January 2008 (UTC)
 * hmmm, though I admit I missed that edit too.  BLACK KITE  20:30, 11 January 2008 (UTC)
 * This was a definite "no" case. Most requests are definite "yes", and those that need more investigation are naturally not as clear-cut.  --Merovingian (T, C) 20:32, 11 January 2008 (UTC)
 * (e/c)This was an example of why a thorough review of contribs is necessary (and why AGF alone is not enough for this). However, it is also a good example of why making this not a huge deal is necessary. We didn't need to have an extended discussion before we could remove it. Mr.  Z- man  20:36, 11 January 2008 (UTC)
 * This is also another example of why this page is necessary rather than just admins granting rollback out to any user: if one admin or a non-admin spots something a granting admin missed, it can be discussed and brought up here, and the problem is sorted out as quickly as possible. Acalamari 21:17, 11 January 2008 (UTC)

Need to move him from Requests for rollback/Approved to Requests for rollback/Denied as well. <strong style="color:#fff;background:#DC143C;border:1px solid #000;font-family:Verdana ">ALLSTAR <strong style="color:#FFF;background:#0F4D92;border:1px solid #000"> echo 20:33, 11 January 2008 (UTC)
 * Done. --Merovingian (T, C) 20:38, 11 January 2008 (UTC)

Rollback logging per user
You know, something occurred to me just now, with my first use of the rollback feature. What if I start a log!? So viola, I have. I did this to display that the function CAN be used responsibly by regular editors, and that we can also exercise enough self-discipline as to keep track of each time we do this. My log is at: User:Edit Centric/Rollback Log. Edit Centric (talk) 22:16, 11 January 2008 (UTC)
 * Doesn't keeping a log mean an additional edit every time you use it? Doesn't that kind of defeat the purpose of having rollback, that of making it easier to undo edits??? --Hammersoft (talk) 22:18, 11 January 2008 (UTC)
 * (ec)I've learnt to not to things like this, unfortunately. Could this be doing something it shouldn't? Congratulations on gaining it by the way. :) Best regards, Rudget . 22:20, 11 January 2008 (UTC)


 * It's not hard to see uses of rollback in someone's contribs. That's what they're for. Why maintain a log? • Anakin (contribs • complaints) 22:36, 11 January 2008 (UTC)


 * No, what this does is keep a personal record of every use of rollback. That way, if a situation arises in the future where I have to account for my actions on a ceratin article, I have a point of reference. Also, since rollback does not allow me to enter verbose resoning for the revert, I can do that on my log, again to keep track. It's a matter of self-discipline, and common sense "CYA" (If you don't know what CYA stands for, ask someone...) Edit Centric (talk) 22:40, 11 January 2008 (UTC)
 * It's entirely up to you whether you keep a personal log of rollbacks or not: we can't enforce it, as that would be too bureaucratic, and telling people what to do, but I have no problem with people keeping their own logs if they wish to. Acalamari 22:44, 11 January 2008 (UTC)
 * Well I for one will applaud you, Edit Centric, for doing this. It shows real respect for the permission and overall responsibility. --Merovingian (T, C) 22:45, 11 January 2008 (UTC)

Wayyyy too much work. If you start reverting a lot of vandalism at once (try "newbies' contribs") you'll get sick of this real quick. If you're using rollback properly then the reasons should be mostly all the same (after all, vandalism is fairly narrowly defined) &mdash; TheBilly(Talk) 00:02, 12 January 2008 (UTC)
 * You could probably write a script that logs every rollback you make automatically, but with the extra time for the script to run you would lose any time benefit of rollback. Mr.  Z- man  00:35, 12 January 2008 (UTC)
 * I'm pretty sure there's a script created for this at User:Splarka/rightsfilter.js. Haven't tested it yet, though.   jj137  ♠ 00:39, 12 January 2008 (UTC)


 * A noble move, but if we all did this I guarantee some people would be less than honest about it... Master of Puppets   Call me MoP! ☺  00:49, 12 January 2008 (UTC)


 * As Qxz I always used to log vandalism reversions; never had rollback, but the principle is the same. That was handled automatically by the software at the end of each session so didn't take up any extra time... in the end I gave up, though, partly because the log got quite long very quickly and had to be archived all the time, and nobody ever actually read it. Logging every rollback individually would of course be stupid because it would double the number of edits you'd have to make and so lose the advantage of rollback in the first place – Gurch 01:54, 12 January 2008 (UTC)

Sweepstake
At the moment we have 1463 users with the sysop bit (administrators), and 413 users with the rollback feature enabled. There is some slight overlap, but given the drama lately, does anyone want a light-hearted distraction in the form of a sweepstake on when the numbers of those with rollback will overtake (as it should) those with the admin bit? I predict the post-flood rate per day will be about 30, and that rate will be reached from around 450, so that is about 1000/30 days from tomorrow, which should be about 34 days from now, which would be about 15th February. That's probably an underestimate, so I'll pick 14th February at 05:00 UTC for the crossover point (in reality it will almost certainly be sooner). Sweepstake should close in a few days time. A signed print of the rollcat picture will be the prize up for grabs. Well, when I say signed, it depends on the cat and Gurch. See Requests for rollback/Sweepstake. Carcharoth (talk) 01:52, 12 January 2008 (UTC)
 * Funny, I was also wondering when people with the rollback feature would overtake those with the sysop abilties. Acalamari 02:37, 12 January 2008 (UTC)

Since the notice isn't plastered at the top of every page anymore, I don't see this happening as fast as most people do. But as much as things keep getting changed in this circus, there's the possibility that it could happen in 2 weeks if the notices are put back up or never if this thing is shut down altogether. <strong style="color:#fff;background:#DC143C;border:1px solid #000;font-family:Verdana ">ALLSTAR <strong style="color:#FFF;background:#0F4D92;border:1px solid #000"> echo 02:56, 12 January 2008 (UTC)
 * On the other hand, after having just reviewed Requests for rollback/Denied and seeing only 1 person has been denied today, it could happen overnight. ;] <strong style="color:#fff;background:#DC143C;border:1px solid #000;font-family:Verdana ">ALLSTAR <strong style="color:#FFF;background:#0F4D92;border:1px solid #000"> echo 02:59, 12 January 2008 (UTC)

What we really need...
...is a "rollback" button that, when clicked, rolls back everything that's ever been said on the subject of rollback. Just thought I'd share that with you – Gurch 01:58, 12 January 2008 (UTC)
 * All in favor of quitting the sarcasm on rollback: Aye! Seriously, enough. It's working well for now. We know it's not a popular system, but it is a system. There is reasonable consensus to keep this process. J- ſtan ContribsUser page 02:19, 12 January 2008 (UTC)

Request for review of removal
I have removed 's access to the rollback tool temporarily, while waiting for the outcome of a review here. The log of the action is available here.

See my comments here. Not only was this a poor rollback, but per common sense and our biographies of livings persons policy (as quoted on the page), this was an extremely dangerous, problematic and harmful revert which flagrantly misused the rollback tool and possibly caused further distress to what is presumably the subject of the article.

Given the severity of the circumstances, I felt compelled to remove first and discuss later. We cannot risk a repeat of this behaviour at all. I ask for a discussion as to whether to regrant acess, either immediately, automatically in a certain period of time down the track (pending evidence of Save Us 299's ability to use reversion tools appropriately), or not at all (will require reapplication through the process). Thanks, Daniel (talk) 10:50, 11 January 2008 (UTC)
 * You should probably crosspost this to ANI, because it's a bit hidden here. -- lucasbfr  talk 10:59, 11 January 2008 (UTC)
 * Good point, done. Daniel (talk) 11:07, 11 January 2008 (UTC)
 * This is in no way more severe than the same behaviour using manual revert and an edit summary of "rvv". It happens all the time that people see a section blanking and revert it without thinking. The main problem was that he actually gave a vandalism warning for the section blanking instead of talking to the IP, which is totally independent of how he performed the revert. Kusma (talk) 11:04, 11 January 2008 (UTC)
 * If I saw a user revert an edit like this as "rvv" I'd similarily remove the rollback tool as it shows they can't evaluate what vandalism is and isn't, and hence can't be trusted to use the rollback tool within the bounds set out. If the revert was less dangerous to Wikipedia, I would leave a warning, but given the nature of the revert I thought and emergency action and referral here for discussion was the safest way forward. Daniel (talk) 11:07, 11 January 2008 (UTC)

Wow, I can't even being to tell you how upset I am right now about Daniel's 'shoot first and ask questions later' approach to this, and without warning. I can't believe that no one thought of the fact that I thought the user was inserting that into the article, and not removing it. It is a clear BLP violation, I know that. I must have been going through the diffs a little too fast and thought that User:66.253.251.184 (who was inserting it) was the one removing it and that User:76.87.252.84 (who was removing it) was the one inserting it. There is no way in hell I would make a rollback to reinsert content like that intentionally. I find it hurtful that rollback was removed in such a rude and accusational behavior, and without given the chance for me to respond. The revert and warning I made were absolutely wrong and was meant for this and not this. And I honestly cannot believe that you find my judgement impaired because I didn't read it correctly once and you felt the need to remove it without letting me comment first because you thought I was a "threat" for making this mistake. Not only did you accuse me of doing it to harm the project and to distress someone but you said I did it on purpose. I would like to know where you get the balls to play the part of, an albiet unfair, judge, jury and executioner. — Save_Us _ 229  11:24, 11 January 2008 (UTC)
 * "I ask for a discussion as to whether to regrant acess, either immediately, automatically in a certain period of time down the track (pending evidence of Save Us 299's ability to use reversion tools appropriately), or not at all (will require reapplication through the process)" — I have and had no intention of playing "executioner" in the permanent sense you imply (or playing prosecutor, either, in the discussion which will hopefully ensue), and deliberately initiated a discussion explaining the situation and alternatives. I personally cannot see how the negatives of you temporarily not having access to rollback (which will be the case—"temporary"—if consensus decides it so, and which I will happily honour — I take no sides in this, but merely wanted to see discussion and felt the circumstances merited a temporary action, as I noted above) outbalances the risks when reverting biographies, whether accidentally or not, is involved. You also accuse me of suggesting you "[were] a "threat"", despite the fact I never said as such (especially using that word, despite your quotations). Attacking me does nothing for forming consensus going forward in this discussion, which was the idea of it. Daniel (talk) 11:37, 11 January 2008 (UTC)
 * It's a tough one, and we all make mistakes. However, with the grant of a powerful tool comes a duty to use it with great care. Given that this whole area is stirring up a hornet's nest of controvery, I believe that (for now) caution must be used to ensure that the tool doesn't go to people that might cause harm with it.
 * I see no reason to suspect that you misused the tool with any malicious intent, but I do think that you were insufficiently careful and/or trigger-happy with it. I would suggest that, rather than getting upset, you discuss this with Daniel, explain your error, undertake to be more careful, and ask for the tool back. Mayalld (talk) 11:41, 11 January 2008 (UTC)
 * I must say that this isn't the reaction I hoped for or intended to occur, although I hardly consider it my own doing as I believe my rationale was sound (although this naturally has some bias to it, potentially). In addition to discussing the apparently-now-moot issue of specifics with relation to Save Us 229, I also welcome comments about my own course of procedure and any suggestions about improving it into the future (to also improve Wikipedia), if you feel I erred. Daniel (talk) 11:48, 11 January 2008 (UTC)
 * My take is that Save_Us_229 was over zealous in using the tool that way. However I would suspect that if he had not had roll back the same error would still have occured, but by way of undo, Twinkle or whatever. I think Daniel was sort of right in a precautionary removal of the tool pending discussion. However I am deeply upset that this may cause us to loose a fine and valued editor. However that is not Daniel's doing. I would, perhaps, prefer a "three strikes and your out" (or two or one) approach however. The same as we don't block without sufficent warning, perhaps a reproach on the users talk page first would have been the better take here, with removal if further errors occured or were subsequently uncovered. Pedro : Chat  11:56, 11 January 2008 (UTC)
 * Without commenting on the merits of your proposal, I wish to emphasise that had this been basically any other edit (with the exception of maybe an official, Arbitration Committee-related, or OTRS action or otherwise a mass-revert) I would not have taken the precuationary measure. However, given it involved BLP, I felt that it was the best course of action. Whether or not others agree with this is what I hope to find out, so we can all learn and move forward. Thanks, Daniel (talk) 11:59, 11 January 2008 (UTC)
 * I agree that because of the WP:BLP issue this makes this instance a little harder. But I think we all agree that the rollback was in good faith, just in error. A cautionary lesson to us all I think. Pedro : Chat  12:02, 11 January 2008 (UTC)
 * Might I suggest that Daniel's actions were correct on the precautionary grounds that he stated. Indeed that removal of the tool was a proper and neutral act which mitigated against a possible risk. As it now seems accepted that this was an innocent, if careless, mistake, rollback should be restored as a matter of course. The short withdrawl being more than sufficient to ensure that the user is more careful in future. Mayalld (talk) 12:14, 11 January 2008 (UTC)
 * Although I would personally prefer some more input from others on the issue of whether to regrant access (note: not the same as the discussion about the process used, which should not start or stop in relation to the discussion about the specific user), I will happily defer to the judgement of any administrator on the issue if they feel sufficient discussion/explanation has occured to reapply access to rollback. Daniel (talk) 12:23, 11 January 2008 (UTC)
 * Restoration of rollback is in order here. Two principles apply:
 * >Unlike blocking, no one is going to lose a big edit or their ability to continue to work on any articles when rollback rights are removed, and it is not terribly harmful to remove rights when in doubt on precautionary grounds.
 * >Once the rollback action is identified as having been an innocent mistake, rollback should be promptly restored. No need to prolong a drama, time to move on. The rollback rights log of actions should not be equated with a block log. Rollback is just a time saver, some toggling on and off should be no big deal. For the sake of future incidents, lets establish a good precedent here. NoSeptember  12:26, 11 January 2008 (UTC)
 * Fair points on both counts, and I agree strongly with the sentiments presented of it being removed on precautionary grounds and (as any Wikipedian should) minimizing the prologining of drama. I support another administrator restoring access per the comments above with a permalink to this discussion for historical sake (I will refrain from using my administrative tools in this matter in any way). — Preceding unsigned comment added by Daniel (talk • contribs)
 * Based on your statement, rights have been restored. NoSeptember  12:40, 11 January 2008 (UTC)
 * SaveUs made an error. This happens to both regular editors and admins, and may have happened whether he was using the tool or not. SaveUs happens to be an editor who has given an enormous amount of good faith edits to this project, and often occupies himself doing the mundane, boring tasks that other editors can't be bothered with. Therefore, I think a little WP:AGF is in order. Daniel, I'd have preferred to see you discuss this with SaveUs, rather than immediately remove his access to the tool. I also think your message to his talk page was abrupt to the point of rudeness, especially to a contributer who has spent so much time improving the project. Please restore his access to the rollback function. I trust him with tools more than I trust most admins. Jeffpw (talk) 12:39, 11 January 2008 (UTC)
 * I agree. We all make mistakes and as long as this doesn't start to be a pattern, I don't think a one time error should grant the removal of the tool (Arguably, we all once warned/blocked people/deleted articles by mistake and we lived through it). -- lucasbfr  talk 13:13, 11 January 2008 (UTC)

As a note, I support the restoration of the tool, and feel confident that User:Save_Us_229 will have learnt from this mistake. I hope that he will soon be back from is break. Pedro : Chat  12:43, 11 January 2008 (UTC)

The rollback in question was a simple mistake. Mistakes are certainly allowed. However, Save Us's response here leaves me very unimpressed. He seems quite easily upset and emotional - not the kind of calm, rational thinking I expect from a good editor. So I'm in favor of not giving it back. I want all editors (and rollbackers in particular) to be cool and calm. Friday (talk) 15:46, 11 January 2008 (UTC)


 * I agree with Friday, though I defer to the decision to restore the access from those more involved in this matter. I think we should make it clear that mistakenly making poor use of this tool will lead to its removal, deliberately misusing it is not necessary for that to happen. Wikipedia functioned fine without this tool a few days ago, it is not suddenly something absolutely necessary no one can live without. I worry that losing rollback is starting to be seen as akin to a desysoping, rather than as a more minor form of block (as was originally envisaged). WjBscribe 16:15, 11 January 2008 (UTC)
 * I think was treated just like a block, not at all like a desysopping (with its long community/ArbCom argumentation). Blocks are preventative and so was this removal, once it is clear that it was just a mistake and that the user does understand the intended uses of rollback, rollback was restored just as a block would have been lifted in similar circumstances. Not restoring rollback in cases like this would be akin to deciding to keep someone blocked as a punishment, having lost any sense of a preventative block. There will be cases where a user demonstrates an misunderstanding of what rollback is for, that is when we should be hesitant to quickly restore it, not in this case of an innocent mistake. NoSeptember  23:12, 12 January 2008 (UTC)

Formal process needed
Given the volume of people being granted rollback, I think we need a formal "Requests for rollback review" to deal with issues like the above, so that we can discuss and build consensus on whether someone's rollback tool should be removed. As I said, we also need a proper, 3-day discussion process before granting rollback in the first place. WaltonOne 13:48, 11 January 2008 (UTC)


 * Easy come, easy go? Why not (I'm too lazy to look for where, but, I recall this being proposed before) have seperate sections here... Requests for addition, removal, and, maybe an 'extended discussion' section, where more difficult cases can be moved? SQL <sup style="color:#999">Query me!  14:41, 11 January 2008 (UTC)
 * I think we should have a proper, one-week discussion process before we grant users the right to edit articles. Kusma (talk) 14:57, 11 January 2008 (UTC)
 * With the request made in triplicate, and notarized. Mackensen (talk) 15:00, 11 January 2008 (UTC)
 * Ah heck, why not just put it in Gadgets on people's preferences and then we only need the triplicate forms for removing it. Much ado about nothing. This should not be treated like a mini-RfA; it does nothing that other extant and easily available tools already do.  Risker (talk) 15:06, 11 January 2008 (UTC)

We don't have a formal "block review" before blocks are made - blocks are much more serious than the removal of rollback. We trust admins to make blocks based on their own judgments and sometimes these actions are reviewed at WP:ANI. It seems odd to make removing rollback a more significant action (requiring more discussion) than blocking a user... WjBscribe 16:12, 11 January 2008 (UTC)
 * I think a week may be slightly too long, about 2 days? We're not granting them other rights remember. Rudget . 16:31, 11 January 2008 (UTC)
 * The time period isn't really the issue I wanted to raise. If I think you're being disruptive I may block you. I don't have to ask permission and wait 2 days first. I would be required to give a good explanation of that block on your talkpage and on request at WP:ANI but nonetheless as an admin I don't require approval to block. Why then should prior approval from others be need to remove rollback, something which hinders the user's ability to edit Wikipedia far less? WjBscribe 16:44, 11 January 2008 (UTC)
 * You're right. There's no form of a block review, and they are more serious than rollback. ANI is a good alternative for request based discussion instead of here, although if it's for granting couldn't say two administrators just agree to a decision? Rudget . 16:54, 11 January 2008 (UTC)

The lack of a reviewing process for blocks does not mean that no such reviewing process is needed for rollback removal. Rather, a reviewing process is needed for blocks, and I have said so for a long time - hasty blocks from trigger-happy admins have driven many good editors away from the project in the past. I don't understand why a lot of Wikipedians seem to think that the biggest problem we face is too much bureaucracy. Bureaucracy is a necessary evil in order to ensure accountability, and I would rather have more bureaucracy and less discretion on the part of those with positions of trust. WaltonOne 17:07, 11 January 2008 (UTC)
 * If we require a 2 day discussion before granting and a mandatory discussion for every removal, it becomes a big deal, which most people don't seem to want. Much of the opposition on the straw poll was that this process, with no strict requirements and no required discussion was too bureaucratic. Granting it should be no big deal. Do they have decent experience? Yes. Do they get into edit wars? No. Do they improperly revert? No. Then give them rollback. If they misuse or abuse rollback, warn them and/or remove it (depending on the severity). Whether or not there was consensus for us to implement this in the first place, there was definitely no consensus (perhaps consensus against) an RFA-lite system where rollback is exponentially harder to remove than it is to grant. Mr.  Z- man  19:54, 11 January 2008 (UTC)

Are we going to implement a 2-day waiting period for permission to edit your monobook.js, too? If anything, rollback is much less of a security risk than letting you edit your monobook. Heck, a long time ago before I knew it was against the rules, I used my monobook for mass automated userfication of userboxes following a TFD discussion. The capability for mass botlike vandalism is far more potentially harmful than having a rollback button, which your monobook lets you have anyway even without the special permission. --B (talk) 23:38, 11 January 2008 (UTC)
 * I agree with B. Why does this have to become a mini RfA? It is a tool that can be removed at the click of a button. Waiting several days is really unnecessary.   jj137  ♠ 23:41, 11 January 2008 (UTC)