Wikipedia talk:Flagged revisions

List of Sub Pages

/Trial

/Trial/Votes

/Trial/Proposed trials

Blogs

A list of blog posts on the issue. Sdedeo (tips) 22:37, 9 September 2007 (UTC)

Framework for evaluation

I would like to see some kind of consensus on what the requirements are for a flagged revision system. It seems to me that there's considerable disagreement about what the purpose behind them is. If there was a framework, we could evaluate the alternatives in light of these priorities, and make an informed decision by evaluating each alternative according to our goals. That said, here is my understanding of the situation, and some of the important metrics the community may be applying:

Problem: Quality has become a desired attribute in the community to prevent random (or intentional) drift in a backward direction. We wish to do this while maintaining the spirit of empowerment and democracy that has been Wikipedia's identity; this is a root principle. If we choose to compromise it to add a new attribute (quality) that wasn't present before, we need to understand the impact of this.

So, in my evaluation, we need to understand:

• Does the proposal maintain the spirit of Wikipedia?
  • Does the proposal result in significant collateral damage?
  • Does the proposal allow democratic access to the mechanisms of control?
    • Does the mechanism assume good faith, to leverage our users good will?
    • Does the mechanism leverage the effect of time to dampen conflict?
  • Does the mechanism promote progress?
  • Does the mechanism promote consensus?
• Does the proposal improve quality?
  • Can we afford to wait for a better technical solution in the near future?

Please note that I think most proposals I've seen *will* promote the quality of Wikipedia. I believe the key distinguishing features will be the costs imposed on the collective goodwill. Delegation to mechanism by default has the effect of empowering the mechanics. I believe we should do everything possible to make the mechanisms as transparent as possible to counteract this effect. Honestly, I haven't seen that done in this discussion. 70.251.156.202 (talk) 19:00, 1 February 2009 (UTC)

I've put my finger on why I'm concerned about the flagged revisions mechanism. It's because it is fundamentally asymmetric. It creates a mechanism for separating two classes of users in a non-democratic way. This should be avoided and/or mitigated at all costs, if you value the democratic process. Wikipedia is founded on a symmetric editing process; any editor can edit/revert any other. I think we need to maintain this fundamental symmetry in any new mechanism, or risk it being corrupted by those in the position of power. 70.251.156.202 (talk) 19:22, 1 February 2009 (UTC)

Idea that might help

I'm really late to the party and this is probably the wrong forum, but could someone explain why we couldn't work in collaboration with search engines to help solve the problem FLR is entangled with? For example, we could arrange that search engines like Google cache the last edit that has remained for more than ten minutes, rather than the last edit. Seems reasonable to me; it'd at least prevent the Obama incident from recurring. —Anonymous Dissident^Talk 13:36, 6 March 2009 (UTC)

If the last edit is a correction on some errur in the article it would show the wrong version, most people improve an article rather than vandalizing it. Mion (talk) 18:14, 6 March 2009 (UTC)

Let's see what we can get

... and Keep It Simple Stupid.

The discussion about flagged revisions has largely died down, again. I say we restart it, and pull it through this time. Jimbo says he is working on something, but it has been a while. Also, while Jimbo may be the god-king of Wikipedia, he is not omniscient, and is unlikely to come up with a better proposal than us.

I am convinced that it is possible to get consensus if we present a clear proposal. However, this proposal must be written with the goal of reaching consensus, not going as far as possible.

After reading the various proposals, I want to start with this. Please answer the questions given.

Administrators can enable flagging on:

• Any article, under similar circumstances as semi-protection.
• Any biographies of living persons (BLP), as long as the flagging backlog is short. Priority to articles that seem to have few to no editors watching.

Any user can edit. It is not shown to anonymous readers until flagged by a reviewer Reviewers:

• Shall the "reviewer" right be given automatically?

• After how long? 100 edits + 3 months?

• Admins can give/take away the reviewer right
• Shall reviewer be tied to rollback?
• ..or shall we make all current rollbackers into reviwers, but keep the rights separate

Shall edits by reviewers be flagged automatically?

• Edits by admins?
• Shall edits by bots be flagged automatically if the previous revision is flagged?

Shall we automatically flag revisions after a time limit, say 1 week?

Do we need a trial?

• No, start with a few articles and roll it out gradually?
• Yes, a time limited trial?
• Yes, a trial period where we try flagging, but readers still see the newest revision, flagged or not?

I intentionally kept this simple with as few new user groups and protection levels as possible. Perhaps we want a new group to help admins enable flagging on articles, and perhaps we want a kind of full-flag-protection where only admins can flag revisions. However, let's start simple and add that later if needed.

What we need to do now is discuss and make a consensus proposal. I suggest spending two weeks doing this. After that we open an RFC to see if there is consensus. This would take about a month, and if there is consensus, then we can start turning this on towards the end of April. Optimistic? Yes, but possible. --Apoc2400 (talk) 15:20, 11 March 2009 (UTC)

I made pretty much the same suggestion here. — Blue-Haired Lawyer 15:29, 11 March 2009 (UTC)

Yes I saw it. The difference is that I prefer to do it in the other order: Instead of starting with a straw poll, start with discussion and create a clear proposal together. Then hold an RFC/straw poll on a well-defined proposal. Do you want to help out? --Apoc2400 (talk) 15:34, 11 March 2009 (UTC)

I have several responses and concerns:

• Flag revs should be BLP and BLP related. Any time an article can do real damage to real people - even in theory, it deserves heightened protection, at any administrator's discretion, in conflict, default to on. Issues outside of BLP that cause real damage can also be flag reved but should go through a consensus process (AfFR or RfC), not admin discretion.
• If flagged revs are BLP focused, then the reviewer permission should be at a higher level of scrutiny, that is, a sort of big deal (between rollback, but the permission should not be tied to any other permission - BLP judgment is different from DR judgment and consensus judgment. Most admins should be able to get the sighting right, but who knows, maybe some don't. Most experienced users should be able to gain the sighting right, but defaulting to off in the event of disagreement.
  • I do not mean here any disagreement, but significant and reasoned disagreement. This is a judgment call, one that should be given to all administrators. I'd normally say 'crats, but we don't have enough. Another possibility is any reviewer can grant (but not remove) the reviewer flag. A noticeboard may be in order, but those are details we can work out later
• Sighting is not automatic, unless there is a reversion to a previously sighted version (if this can be implemented). Reviewers may sight their own revisions, a power that can be revoked quickly if abused.
• Reviewers are obligated to sight revisions that they disagree with but that do not violate BLP. Normal editorial practices must persist.
• 2 week trial period, targeting a cross section of BLPs and a couple non-BLP articles of various visibility OR 2 week trial of what I've described above.
• I have concern with the idea of giving priority to low activity articles is that those articles are the ones which are in the greatest need of improvement, and initial improvement (fact and content generation) often comes from IP and redlink users. Leaving an article barely readable isn't much more helpful than protecting it from drive by vandalism.

Lets get this show on the road.--Tznkai (talk) 16:55, 11 March 2009 (UTC)

Reviewers are obligated to sight revisions that they disagree with but that do not violate BLP. Normal editorial practices must persist. -- but if they disagreed, 'normal editing' would mean they'd revert it, would they not? Why should an edit be sighted and reverted, instead of just rejected, especially if the later takes less time? ♫ Melodia Chaconne ♫ (talk) 17:12, 11 March 2009 (UTC)

I understand you want to extend FP to many articles and have very high requirements on reviewers, but I think it is necessary to do just the opposite if we want consensus on this. Ideally only the most trusted users should review BLPs, but there is no way that the hundred or so that are really active on BLP issues will be able to watch over half a million BLPs + whatever other articles. Too strict requirements for becoming a reviewer also gives a strong impression of a cabal. With a lower requirements there will be bad sightings, but some improvement on BLP issues is better than nothing. Anyone who persistently sights libelous edits will have the sighting permission removed. (Addendum: Compare this proposal to semi-protection of all BLPs which is seen as quite extreme, but would actually make a rather low barrier to libelous edits)

I suggest automatic sighting after one week as a safety valve. If an edit is neither sighted or reverted in reasonable time, then the flagging system is failing. Hopefully it will never get to that.

As long as edits are sighted quickly it would not really stifle the improvement of new articles much at all. High-profile articles like Barack Obama on the other hand is likely to have a lot of edits and would swamp the reviewers when the edits are better dealt with by the regulars of that article.

I strongly agree with the spirit of Reviewers are obligated to sight ... Normal editorial practices must persist. One of Wikipedias greatest strengths is that every edit is accepted unless reverted. If somebody disagrees with an edit a little bit but not enough to revert, then it sticks. However, I don't see how it is practically possible. How would we prove that sighter A did not sight an edit because he disagreed with it, and that he didn't just miss it. --Apoc2400 (talk) 17:50, 11 March 2009 (UTC)

I'm less concerned with creating an enforceable mechanism than creating a cultural standard - a norm. The idea is that reviewers will understand that sighting is not a tool for editorial disputes - likewise, I'm very much ok with the idea that the reviewer flag is easy on, but also easy off: users are assumed to be trusted, but we're also willing to quickly remove the permission if there are problems. Is there a way where auto-sighted articles will show up on a separate list so volunteers can review them? Kind of like recent changes.

I'm still very concerned by the new and underdeveloped articles. Click random article a few dozen times, and you'll run into stubs that have been around and untouched since 2006, and I think Flag revving them will make that problem worse. I myself started out as an IP editor - and I imagine if I couldn't "see" my revisions, I may never have started getting into it, and thus never bothered registering an account 3 months later. With BLPs, that may have to be the compromise we make, but I hope someone can think of something clever instead.--Tznkai (talk) 21:02, 11 March 2009 (UTC)

We don't have consensus to use flaggedrevs arbitrarily on blps, and we won't have it. Let's face it, the blp survey, the trial discussions, all show it. Another discussion in this sense will end in no consensus and the community will be a little more tired of FlaggedRevs. We need to break the issue in two: one passive flag used for all blps, one active flag for certain blps, with a use defined by an appropriate protection policy. Cenarium (talk) 18:00, 11 March 2009 (UTC)

I do like your idea. The number of BLPs with active flagging can surely be increased over time if the system works well and edits are flagged in a timely manner. Tye last flagged protection proposal got shot down mostly by a bunch of pro-FR people who saw it as an attempt to water down FR. Today they probably realize it's either something like that or the status quo. --Apoc2400 (talk) 18:09, 11 March 2009 (UTC)

I think the problem is as follows: we need to improve our monitoring of blps, but we do not have consensus to use 'strict' FlaggedRevs arbitrarily for blps. We need to propose something that has a chance of being adopted. So, I think we could start by proposing the use of a passive flag for all blps, and use an active flag on a case-by-case basis, with use defined by the protection policy. It's what I propose in Wikipedia:Flag protection and patrolled revisions. See also this essay. Cenarium (talk) 17:43, 11 March 2009 (UTC)

Could you explain the difference between passive and active flagging to me? I'm not sure I know what you mean.--Tznkai (talk) 20:47, 11 March 2009 (UTC)

I think passive means after a certain time it gets flagged anyway. If so, I'm not keen on that idea. We should keep our backlogs down but every flagging should be on purpose. ++Lar: t/c 20:56, 11 March 2009 (UTC)

I think what Cenarium means with passive flagging is that flagging is enabled, but readers see the latest version, flagged or not. This does not have any direct effect on content, but gives us way to make sure that every edit to BLPs is checked.

Back to time limit, I think the best solution would be that if a revision is not flagged or reverted within one week, then flagging is temporarily disabled on that article. Readers will see the latest version until someone gets around to flag or revert. However, I don't know how difficult this would be to archive in software or with a bot. --Apoc2400 (talk) 22:22, 11 March 2009 (UTC)

My view on this is that I favor any partial measure that is likely to get implemented and that is not going to stand in the way of a better measure. That said I do not favor giving flagger or reviewer out under any process that does not involve review by humans. Mere edit count or tenure are insufficient. ++Lar: t/c 20:56, 11 March 2009 (UTC)

Passive flagging is the wet-dream of vandals, because once the default time has elapsed, nobody thinks it might be vandalised, because we have flagged revisions, don't we? :-) -- Kim van der Linde ^{at venus} 21:03, 11 March 2009 (UTC)

See above. Also: It only happens if an edit goes un-flagged but also not reverted for a week. --Apoc2400 (talk) 22:22, 11 March 2009 (UTC)

To do so is making it relative easy to vandalise. There is no objection that some content is not visible, there is objection to have bad content visible. -- Kim van der Linde ^{at venus} 22:48, 11 March 2009 (UTC) reply

Can you explain why getting review must be manual? It does create a very strong impression of a cabal. Reading through oppose comments to flagged revisions in the recent polls, I see a lot about FR giving too much power to certain groups. I actually think that even giving review right to IPs would be a vast improvement, as long as you can not review your own edit. --Apoc2400 (talk) 22:22, 11 March 2009 (UTC)

Let those with fears for the cabal deal with the shit then. Are they willing to do that? Can they do that. Let see how long Jimbo is willing to deal with piles of vandalism and corresponding news items because some people are afraid for the cabal. Really, opposing is easy, because it does not oblige you to anything. -- Kim van der Linde ^{at venus} 22:48, 11 March 2009 (UTC)

Make the "cabal" huge enough that it doesn't matter. I reiterate my support for easy on easy off.--Tznkai (talk) 00:12, 12 March 2009 (UTC)

(Undent) We're talking a lot about backlogs, without any understanding of how bad they are, because we have no data. How about turning on flagrevs for two months on a set number of articles (say, 3000 across a cross section of BLPs and other articles of varying profiles), getting a group of sighters, and seeing how fast the backlog grows?--Tznkai (talk) 08:45, 12 March 2009 (UTC)

Like some variant of WP:Flagged protection and BLPs? :) I keep being told that noone wants BLPs to be flagged any more strictly than other articles... Fritzpoll (talk) 10:42, 12 March 2009 (UTC)

Just imagine, flagged revisions will make it easier to deal with vandalism, and people are afraid that there will be a backlog in flagged revisions? That implies that the backlog in vandalism removal is even larger. And people are opposed to flagged revisions. I think the lost the overview of the complete picture. A test is not going to work to get an idea how much backlog will arise, because articles in such a test will have more people on it than the average article on wikipedia once it is implemented. -- Kim van der Linde ^{at venus} 13:26, 12 March 2009 (UTC)

I don't want to protect BLPs in any way more than others. I'd like the world to collectively get over it, read Wikipedia (and for that matter, mass news media) with a critical mind, check their own facts. I'd like a million strong army of volunteers well read on BLP policy going every revision with a finetoothed comb. That isn't going to happen, so flag revs will have to do. I'm honestly not concerned about simple vandalism, I'm concerned about "real" harm that can and I'm led to believe, does happen, to real people, because of inaccuracies on Wikipedia. Likewise, I'd like for Wikipedia to stop being used a tool for ne'er-do-wells and the malicious to do social violence unto others. Flagged revisions is a way to do that. I realize that any sort of trial cannot accurately replicate 2.7 million articles under flagged revisions, but any sort of pilot program and the subsequent experience and data is better than arguing on principle and ignorance.--Tznkai (talk) 13:52, 12 March 2009 (UTC)

Let's trying to compromise and find something that has a chance of being adopted here. If people refuses to do that, we'll necessarily fail, again. We know that the concept of using flagged revisions actively for arbitrary blps has no chance of gaining consensus: the blp survey, the trial discussions, etc, it's clear that proposing this will result in another no consensus and we will be in the same situation, with the status quo. What I propose is this:

• Using a passive flag for all blps (that is, it just act as checkpoints, doesn't act on edits but it sill allow to monitor blps)
• Make it "active" on a per-page basis, when ? being determined by a policy (to be merged in the protection policy, and initially with the same requirements).

See User:Cenarium/On FlaggedRevs for more on this. We won't receive much of the classic opposition to FlaggedRevs and it'll still be a huge improvement over the present situation. See the proposal. Cenarium (talk) 13:47, 12 March 2009 (UTC)

I'll proceed in three steps:

• bring Wikipedia:Flag protection and patrolled revisions to VPR and various FlaggedRevs/BLP-related forums for a preliminary discussion
• if there's some support for it, make a centralized discussion, work things out, address concerns, etc
• if there is a good support and a reasonable roll out strategy is found, watchlist notice to gather final consensus for implementation

Cenarium (talk) 16:55, 12 March 2009 (UTC)

Hold up. We're definitely not going to get anywhere unless we can actually educate editors on what Flagged Revs in general and I still don't understand what active and passive flagging is, and what the difference. Try explaining it like I'm reading Simple.--Tznkai (talk) 17:28, 12 March 2009 (UTC)

A passive flag will simply allow reviewers to flag a revision, but it won't affect the version viewed by readers. It means that we can monitor the pages, it'll give a list of all pages that have not been flagged, and changes to flagged pages (and it's filterable by category, so by Category:Living people). An active flag will affect the revision viewed by readers by default, it'll be the latest flagged revision.

My proposal is, broadly speaking, to use a passive flag for articles (and in practice especially for blps), and to "active it", on a per-page basis. When the flag should be "activated", should be determined by a specific 'protection' policy. It's also explained in this essay. Cenarium (talk) 17:52, 12 March 2009 (UTC)

I'm for the idea of passive flagging. I don't want the entire encyclopedia to have this, as it will create backlogs and other problems, but this way, it should make everybody happy, for the most part.Deavenger (talk) 20:17, 12 March 2009 (UTC)

I like your proposal Cenarium, but please write a very simple explanation of what passive flagging means. I understand it now, but it seems like everyone gets it wrong the first time. Also, do we really need "Full Flag Protection"? Normal full protection is still available. Also, what requirement for becoming a reviewer do you suggest? --Apoc2400 (talk) 21:04, 12 March 2009 (UTC)

If I understand this correctly, passive flagging would essentially create a list of articles that have not vetted by a reviewer yet, similar to how new page patrol works?--Tznkai (talk) 16:55, 13 March 2009 (UTC)

No forced vandal whacking: Flagged revisions now

I have a radical proposal.

1. Those who object to flagged revisions can do the vandal whacking work.
2. Those that want flagged revisions sit back and let those who object to that do the vandal fighting.

Add this user box to your page:

This user demands flagged revisions NOW and refuses to do vandal fighting.

Vandal fighters, good luck, I hope many will sit back and see you guys do the work. -- Kim van der Linde ^{at venus} 19:51, 11 March 2009 (UTC)

Nice, but not wildly constructive. ╟─Treasury Tag►contribs─╢ 20:21, 11 March 2009 (UTC)

Might make the point, though. Thanks Kim. (you have to ask yourself, why are the poll results skewed, time after time... why do old timers and admins favour flagged revisions more than less seasoned users and non admins... what do they know about it?) ++Lar: t/c 20:35, 11 March 2009 (UTC)

Wow, it is not wildly constructive? No, having to fight vandalism is constructive. I think Jimbo should just implement it, with all with roll-back can flag, and the relevant policy pages will be done within days. The current talk and talk and talk and talk has been wildly constructive I see, we are still whacking vandals and no flags in sight! My guess is that my userbox and actions are MORE constructive than the archives of talk conducted here. Once the vandalism is out of control because many do not fight it anymore, I think many of the nah sayers will be glad to have flaged revisions. Especially when more and more hughly embarrasing vandalism remain for grab for the press. -- Kim van der Linde ^{at venus} 21:00, 11 March 2009 (UTC)

I do take your point, but it's not constructive in the sense that it's not going to move us forward. If some anti-terror law is under debate in Parliament, politicians have long realised that the, "OK, if you won't let us fight terrorism properly then you can all get blown up," argument doesn't work. Making jibes at the opposition isn't helpful IMO. But don't mind me :-) ╟─Treasury Tag►contribs─╢ 21:11, 11 March 2009 (UTC)

Wrong analogy, I think... the better analogy is if some significant fraction of workers at a particular skilled trade go out on strike, (for example air traffic controllers) there's an impact on the operation (for example, flights) that their trade supports. Either the issue causing the strike needs to be resolved satisfactorily (for example by implementing new air traffic procedures), or the workers need to be replaced (for example, as Reagan did when faced with an ATC strike). Every strike starts somewhere. Perhaps I've misinterpreted what Kim is suggesting here though. ++Lar: t/c 21:29, 11 March 2009 (UTC)

(answering Kim) What you may end up with in that case is flagged revs at some point in the future, along with an embarrassing legacy in the press that will take time to recover from. If the community is so wedded to the consensus idea then try deciding the first question first - shall we have flagged revs, followed by the actual form later. Or better yet, don't let the community decide. WP:NPOV is an editorial issue laid down from above, why shouldn't this be the same. ARBCOM needs to stand up and say that:

• Flagged is turned on, and it will not be turned off
• For X months we will try this variant
• After that period it will be reviewed

No-one can argue with them, because there is no recourse for appeal except to Jimbo. 203.24.135.66 (talk) 21:42, 11 March 2009 (UTC)

If you want to replace the volunteers as Reagan did you have to get them from Citizendium, and the problem with volunteers is, once they walk away changing the procedures is useless as they already left the project. And 4 months ago this was a solution looking for a problem, BLP was just one of the options, if you want to improve, improve Cluebot. Mion (talk) 21:49, 11 March 2009 (UTC)

Sorry? A solution looking for a problem? BLP has been an issue for years. Cluebot is a solution for vandalism, not well written defamatory content. 203.24.135.66 (talk) 22:00, 11 March 2009 (UTC)

if you do not want to end up with that bad legacy, enable flagged revisions NOW, and that won't happen. Don't blame those who want flagged revisions, blame those who oppose it. And tell those editors to work harder in fixing the vandalism, because THEY WANT TO DO IT THAT WAY and block flagged revisions; do not blame me, or Lar, or many others because we are fed up with those that block flagged revisions and whine that there is to much vandalism. -- Kim van der Linde ^{at venus} 22:12, 11 March 2009 (UTC)

Wait, so people are saying that users should stop reverting vandalism until flagged revisions are implemented? That would just make things worse. So, unless things go a particular way a group of users want it, they're going to bitch and not do anything till it's implemented? That doesn't help the encyclopedia. If anything like that happens, Jimbo might as well just pull off a Teddy Roosevelt, and prevent editing overall until the issue is fixed, which would be worse then lots of talking. Deavenger (talk) 22:02, 11 March 2009 (UTC)

Except for the BLP subjects, who may welcome such a decision. 203.24.135.66 (talk) 22:06, 11 March 2009 (UTC)

@203.24.135.66 current progress indicates that the cluebot AI type can handle defamatory content in the near future, and as this encyclopedia is not finished today, there is no reason to introduce measures with a lot of side effects as the market will give us the tools/algorithms in the coming years(http://www.wolframalpha.com/). Mion (talk) 22:12, 11 March 2009 (UTC)

I guess we disagree. I think that action is required now, not in a few years.

203.24.135.66 (talk) 22:23, 11 March 2009 (UTC)

Then start the project Cluebot BLP and start writing...Mion (talk) 22:27, 11 March 2009 (UTC)

Right, as if a bot ever can catch vandalism like this in Hyacinth Macaw: they have red wings. They have not, no bot will know that. Get real, no bot can fix this. -- Kim van der Linde ^{at venus} 22:35, 11 March 2009 (UTC)

Piece of cake for a bot, as databases keep opening up, comparision with descriptions in other databases will filter out the anomalies, as per NOR original desciptions are not allowed all descriptions can be cross checked. Mion (talk) 22:41, 11 March 2009 (UTC)

I have an idea, till such a good working bot has been developed, we enable flagged revisions, and when the bot is working, it can do automated flagging on content it has checked. In that way, we do not have to wait for something that is CLAIMED to be possible, but for which no working prototype has been shown. -- Kim van der Linde ^{at venus} 22:53, 11 March 2009 (UTC)

Such a bot that could detect good content from bad could write the encyclopedia on it's own, couldn't it? I think I saw this story in a film somewhere. It didn't work out, as I remember. 203.24.135.66 (talk) 22:58, 11 March 2009 (UTC)

Indeed, let the stuff explode. I write good content, and as soon as the flagged revisions are implemented, the good content is quickly enough found and flagged. Vandal whacking is like coorborating with a evil empire under the guise that it could be worse if you would not help them. And no, I am not going to bitch about the vandalism that piles up in the period between now and when flagged revisions are implemented. It piles up already anyway. Sometimes, you have to take a prinicpled stand to actually make a change. if you want to fight vandals, be my guest, there is enough to do for you, but I rather spend my time on more constructive things, like writing good content. -- Kim van der Linde ^{at venus} 22:17, 11 March 2009 (UTC)

"OK, if you won't let us fight terrorism properly then you can all get blown up," Indeed, so, put the flagged revisions on, and fix all the details after that. Waiting longer is like having no sensible way at all to do anything. Vandal whacking is like throwing terrorists out of the country but have the border open to everybody without immigration. Not useful. Flagged revisions is like having immigration to keep the terrorists out. -- Kim van der Linde ^{at venus} 22:12, 11 March 2009 (UTC)

I don't quite understand the argument here. FlaggedRevs doesn't prevent vandalism. It prevents people from seeing it, and as a result may reduce some of the motive. But there will still be vandalism with FlaggedRevs and it will still have to be reverted. Mr. Z-man 21:39, 12 March 2009 (UTC)

Arguably there will be less vandalism, which is goodness. Further since it's not visible, it can be undone at leisure. So there is less pressing need to fix it immediately. ++Lar: t/c 05:00, 13 March 2009 (UTC)

Try it, it won't hurt!

It is common sense to take a method and try it. If it fails, admit it frankly and try another. But above all, try something. —Franklin D. Roosevelt. -- FayssalF - ^{Wiki me up®} 17:13, 12 March 2009 (UTC)

since there is no way to get consensus for any type of flagged revision thing prior to implementation, is there a mechanism to get someone to just implement it? and then as we sort out the details of who/what/when/how/how long/were/etc, we begin to hide the egrigeousness that burns out OTRS and Oversight'ers faster than we should? --Rocksanddirt (talk) 20:53, 12 March 2009 (UTC)

Yes. WP:ARBCOM as the sole body here that cannot be usurped, need to rise out of their present mandate and just do it. 203.24.135.66 (talk) 21:06, 12 March 2009 (UTC)

They already have a mandate to implement this, at least for BLPs. ++Lar: t/c 05:00, 13 March 2009 (UTC)

Passive flagging: More work, no benefits

Ok, if I understand it correctly, passive flagging mean that revisions are flagged, but that it does not have any effect on what the visitor gets to see. Correct? So, what is the benefit? All I can see is a load of work, for the case that the flagging is activated because of issues. So, why not leave the initial flagging till the moment it is activated, as a requirement for the admin who turns on the flagging? -- Kim van der Linde ^{at venus} 17:23, 13 March 2009 (UTC)

It helps focus our resources to make sure all edits are checked. If an edit is flagged then other people can skip and check other things. Also, we don't have to use it. --Apoc2400 (talk) 18:20, 13 March 2009 (UTC)

You needn't to flag all revisions. It's the advantage of being passive: there's no hurry and it's not terrible to have a backlog. It allows to monitor changes to all blps that have been patrolled at least once and to list all that have never been patrolled. At the present time, we have no way to do that so efficiently: we can't cope up with the level of editing blps receive. We only have recentchanges and a few disparate tools for that, a passive flagging would provide a structured and optimized monitoring system (by reducing the changes to check). Cenarium (talk) 22:08, 13 March 2009 (UTC)

It'd be nice, but I'd like something a bit more substantial.--Tznkai (talk) 22:14, 13 March 2009 (UTC)

In cases of blp violations on an article, an administrator can use an 'active flag' on this page, temporarily or indefinitely, as a protection measure. The precise use would be circumvented by a special protection policy. (I'll also comment in the Another idea section.) That's the idea behind Wikipedia:Flag protection and patrolled revisions. Cenarium (talk) 22:22, 13 March 2009 (UTC)

If there's no effect on what the reader sees and its just an internal note to other editors, then there's no motivation to actually do the work. Mr. Z-man 22:20, 13 March 2009 (UTC)

Our blps suffer from being too little watched and blpvios remain unnoticed... It'll allow to bring them to the attention of reviewers. Note also that reviewers are autopatrolled, saving us some work. Cenarium (talk) 22:26, 13 March 2009 (UTC)

It also wouldn't hurt. Perhaps we should split flagged protection and passive flagging into two separate proposals and run polls for both at the same time. As long as there is not a long list of "hardline" proposals as well we avoid the "I'm just voting no to everything related to flagged revisions" effect. --Apoc2400 (talk) 22:38, 13 March 2009 (UTC)

We have, essentially, watchlists and (linked)recentchanges to keep track of changes to blps. And only a few edits out of the total are actually reviewed, and we can't improve that much given our resources. Special pages like User:MZMcBride/BLPs or Category:Unreferenced BLPs list only those that have been noticed, and aren't dynamic or complete enough to provide an optimal monitoring system. A passive flag will allow to bring the attention of unnoticed or unaddressed ones (Special:UnpatrolledPages), and monitor changes to blps patrolled at least once (Special:OldPatrolledPages). A random BLP may be patrolled once every 5 or 10 edits, depending on visibility and editing volume, but it'll still allow to reduce the availability to readers of a blp violation, and notice those that suffered from repeated blp violations, so that we can flag protect them. Actually, it's less effort to check a diff (that may contain several intermediary edits) and click on a patrol link if it's ok, or investigate further if not, than checking edits one by one on the watchlist or via history. And it'll save time for other users who would have checked the edits. Cenarium (talk) 22:50, 13 March 2009 (UTC)

Wikiproject based implementation?

Ok, I am thinking, what about a wikiproject based implementation. Suppose, that the wikiproject XXX has agreement among the regular editors who do most of the checking that they want to implement flagging, so that they have more time for more sensible things than vandal whacking. Would that be acceptable for the community? -- Kim van der Linde ^{at venus} 17:24, 13 March 2009 (UTC)

Nay. I'd rather not formalize Wikiprojects as part of our bureaucracy. Wikiprojects have no governing policies or even norms, and zero consistency in quality and participation. Not to mention a nightmare waiting to loom at the intersection between multiple projects. Another drama filled noticeboard would actually be less dramatic.--Tznkai (talk) 17:36, 13 March 2009 (UTC)

No, wikiprojects don't own articles. --Apoc2400 (talk) 18:21, 13 March 2009 (UTC)

Support if done right. I don't think all the negatives above apply - most intersections are not turf wars, and I don't think Kim was saying WikiProjects "own" articles, only that they should patrol them - but WikiProjects also care about keeping a specific group of articles in good condition. I would propose that WikiProjects be asked to help in patrolling articles under their purview, as a way of supplementing the centralised system being proposed. However, even a fully WikiProject-organised system would get my vote, if organised well. Walkerma (talk) 03:01, 15 March 2009 (UTC)

Another idea

Not my preferred proposal, but I'm trying to find a consensus point here.

Flagged revision capability is turned on wiki-wide, accessible from the protection button. Administrators may turn on flagged revisions for any individual article to address.

• BLP violations on high profile articles (high visitor traffic)
• BLP violations on low activity articles. These articles are subsequently tagged for a new type of clean up in some sort of attempt to get them adequate expansion.
• Extensive vandalism or edit warring that would normally be addressed with full protection.
• By consensus for any article.
• By and large, flagged revisions will be treated similar to protection.

This I believe, is very similar to WP:Flagged protection except that the controlling policy would be targeted at BLP problems and long term vandalism, and is in no way linked to semi-protection. The essential goal here is not to control edit warring or vandalism per se, but to ensure the outward face to the public remains stable and safe.--Tznkai (talk) 17:49, 13 March 2009 (UTC)

This is weaker than what I have proposed previously, but is a step in the right direction, and so is a compromise that I can support. The only thorny question is - who can review? Fritzpoll (talk) 17:56, 13 March 2009 (UTC)

I would prefer a higher standard than rollback, but lower than adminship, but at this point I'll accept anything so long as sighting is removable and logged.--Tznkai (talk) 18:05, 13 March 2009 (UTC)

Agreed - we can decide these details in flight. If the initial system doesn't work, it can always be changed. Let's just get on with it Fritzpoll (talk) 18:11, 13 March 2009 (UTC)

Can work, and I am in favour of giving everybody with rollback also flagging opportunities, and you can loose those them easy as soon as you are caught misusing. I am pretty siure that once it is implemented and people get used to it, they will love it. -- Kim van der Linde ^{at venus} 18:17, 13 March 2009 (UTC)

Yes, it should be easy to loose the review right. You don't need it for editing, so it's not a right. --Apoc2400 (talk) 20:09, 13 March 2009 (UTC)

It shouldn't be contained in the rollback usergroup, because sometimes, users are granted rollback while they are not even autoconfirmed ! It's really a tool-like right. Of course, we can revisit this later. But for now, I think they should be separate. When a user asks for rollback at WP:RFR, and an admin thinks the user would make a nice reviewer, it could be proposed or granted to the user in the same time. I still think we should keep the autopromotion to reviewer rights relatively high, then consider lowering it later (because, when we have given the rights too largely, it's hard to repair this, but it's easy to lower thresholds). It could also be requested at WP:RFR. Cenarium (talk) 00:06, 14 March 2009 (UTC)

I like this proposal also. Let's implement it now. Deavenger (talk) 20:50, 13 March 2009 (UTC)

That would be for the policy matter, but which implementation should we use ? I propose Wikipedia:Flag protection and patrolled revisions because it's a versatile implementation, that can be adapted to consensus. For example: autoreview is enabled by default for autoconfirmed users, but can be disabled on a per-page basis. The original flagged protection implementation also had no reviewer user group (it was any autoconfirmed user). Consensus will also say where we should use semi flag protection. Consensus will say where we should disable autoreview, and where we should use full flag protection (disputes for the most part). But this is a policy matter, that I don't want to explicitly state in the proposal, for now. Let's wait to see if we have consensus for it implementation, then only discuss policy. Do you agree with this approach ? I should also mention that I merged the idea of a passive flag with the proposed implementation, although we can also consider them separately. Cenarium (talk) 22:54, 13 March 2009 (UTC)

A controlling policy has to come along with technical implementation, even if the controlling policy is vague. If flagged revs is going to be "sold" to the public, we need to have a clear explanation of what we're doing.--Tznkai (talk) 22:56, 13 March 2009 (UTC)

The problem is that if we propose a too strong policy along with the proposed implementation, then it'll be opposed, to the point we won't have consensus. So I prefer to start by gathering consensus for an implementation, of course with an outline of how we're going to use it, and when we have consensus for the implementation, agree on a specific policy. So that we should proceed in this order:

1. Gather consensus for a certain implementation, outline its use, then:
2. Gather consensus for a specific policy on how to use it, then:
3. implement the implementation, with use determined by the policy.

Now, we should consider whther Wikipedia:Flag protection and patrolled revisions is an acceptable implementation for what we want to do with FlaggedRevs. Cenarium (talk) 23:14, 13 March 2009 (UTC)

If we learnt anything from the recent polls it is that people want to know what they are voting for. Please do not leave anything for later consensus. If we do, people will assume the worst and vote against. Of course anything can be changed by later consensus, but the proposal must have a defined starting point. --Apoc2400 (talk) 23:25, 13 March 2009 (UTC)

That's true. In fact, we should really present this as a trial, and note that, say, two months after the implementation, a community discussion will happen on whether we should continue with the implementation, modify it, or completely disable it. With the understanding that the policy should be the same for the two-month trial period, then that after that, of course, we are free to change it based on consensus. Based on an idea of Apoc2400, I think we could outline the use like this:

• "For the trial period, the requirements for applying semi flag protection to an article are the same as for semi-protection, but it can be applied more liberally to biographies of living people as long as the review backlog is short."
• "For the trial period, the requirements for applying semi flag protection with autoreview disabled to an article are those for semi flag protection with additionally, evidence that the disruptive actions are due to autoconfirmed users (sockpuppetry considered)."
• "For the trial period, the requirements for applying full flag protection are the same as for full-protection."

Then if the implementation is accepted, we can work out the details of the policy, before the implementation. Cenarium (talk) 23:54, 13 March 2009 (UTC)

Too complicated: lets keep it simple, with only one kind of flagging for now. Most people don't know flag revs do, let alone what semi flag protection would be.--Tznkai (talk) 01:02, 14 March 2009 (UTC)

If I made this more complicated, it's in order to satisfy everybody. If we don't allow autoreview for autoconfirmed users by default, people will find it too restrictive. If we allow autoreview for autoconfirmed users, people will find it too lenient. And we just can't treat disputes the same way we deal with normal vandalism, since reviewers would presumably be involved in the disputes. We can avoid going into unnecessary details, but those three levels are needed for a viable implementation. Cenarium (talk) 01:17, 14 March 2009 (UTC)

I disagree, but its entirely possible its because I don't fully understand your proposal. I think its either too complicated, or you haven't explained it clearly, or some combination of the two.--Tznkai (talk) 01:32, 14 March 2009 (UTC)

If we use semi flagged protection for disputes, and if reviewers are involved in the dispute, it won't have any effect: since their edits will be automatically flagged or they'll flag their edits, or unflag the edit of others, so this would trigger flag wars. So we need to use a higher flag: that only admins can use: they 'validate' a revision when it's consensual or non-controversial compared to the latest validated. That would be full flag protection, the equivalent of full protection.

Now, we have the choice between automatically reviewing edits by autoconfirmed users (if the previous revision is), or not. If we choose to do so, people will say that it's too lenient or inefficient for some cases, that it can't help with persistent sockpuppetry, POV-pushing, etc. So we would have to use more of full flag protection, which would be bad because it would restrict editing for more people and overburden admins. So people would oppose.

On the other hand, if we disable autoreview for autoconfirmed users, people will say that this is even more restrictive than semi-protection, that in most cases, this is not necessary, etc, and they'll oppose.

So I propose to autoreview autoconfirmed users by default, but to have the option to disable this: so that it pleases everyone. Cenarium (talk) 01:49, 14 March 2009 (UTC)

I don't think flagged protection should be used to solve disputes. If autoconfirmed users are edit-warring then we have normal full-protection. Also, those who oppose low requirements for becoming a reviewer will also oppose auto-flagging for autoconfirmed users. On the contrary, the requirements to have your own edits auto-flagged should be higher than for flagging other peoples edits. In my opinion, we should have only one level of flagged protection, rather low requirements for become a reviewer, but edits by autoconfirmed users should not be auto-flagged. Edits by bots, admins and perhaps reviewers can be auto-flagged if the previous revisions is flagged. --Apoc2400 (talk) 09:42, 14 March 2009 (UTC)

The idea of flagged protection has been accepted by most traditional opponents of FlaggedRevs because it is more permissive than semi-protection. If we don't autoreview autoconfirmed users, they will oppose, especially if it's intended to be used more widely than semi-protection. But there is an option to deactivate autoreview for non-reviewers, so why only wanting one when we can have both ?

Those who oppose low requirements for reviewers won't oppose that much to autoreview precisely because there is the option to turn it off. And they should also note that edits by autoconfirmed users can also be checked via patrolled revisions. Flagged protection is really a measure of protection, not that much of monitoring.

Disputes are not limited to edit-wars, and full protection is not only used for disputes. Some full protections are long-standing, and thus harmful, having an alternative to full protection would be very appreciable. It would also be an entirely new way to handle disputes, more wiki-like, new revisions would be validated when consensual or non-controversial, it would help in the development of consensus. Maybe it won't work all the time, but it's a progress, and certainly worth a trial. Cenarium (talk) 15:05, 14 March 2009 (UTC)

Can you explain what objections you have to the idea I posted up above?--Tznkai (talk) 18:37, 15 March 2009 (UTC)

The requirements appear relatively low for using flagged protection (or imprecise), and thus we'll have opposition straight away for that. I think we should start progressively (my interest is to have the support from as many people as possible, to have consensus for an implementation). The requirements for the two-month trial should be the same than for normal protection, if possible with a more liberal use for blps. And after the trial period, we should discuss the requirements again, and build a special flagged protection policy. I'd oppose the use of a single flag also, for the reasons mentioned above.

But we can run several trials in the same time with the implementation I proposed. I designed it specifically so that it can be largely adaptable and so used for a variety of trials. You can try to gather consensus for your proposed trial. Cenarium (talk) 19:42, 15 March 2009 (UTC)

Wikipedia: The free encyclopaedia that anyone who behaves can edit

Wikipedia: The encyclopaedia that anyone can edit is a frequent used argument against flagged revisions, but is it actually true? No, when you do not behave, you can get blocked, banned, and your edits will even be actively undone if you evade your ban. So, the slogan should be rewritten to Wikipedia: The free encyclopaedia that anyone who behaves can edit. So, who will be affected by flagged revisions? Those that we smack around already anyway. if you add libel to the project, we ride your ass. if you vandalize, we ride your ass. And if you behave, you quickly enough get the opportunity to flag revisions yourself. So, the major difference is that if you are new, you have a period in which your edits are not immediately visible for the world at large, but when you behave, it will be fast enough that way. Is this small window really the major reason many hold flagged revisions up? -- Kim van der Linde ^{at venus} 18:05, 13 March 2009 (UTC)

I speak for the anonymous IP editors, whose interests I am concerned for in that they are the farm from which we get new editors.--Tznkai (talk) 18:10, 13 March 2009 (UTC)

That's true, but this would be a method for filtering out those whose intentions are good. I'm sure a well intentioned IP editor understands the need for some control as well as you do. 122.49.135.223 (talk) 22:35, 13 March 2009 (UTC)

In my opinion good intentions and self-control may not go together. An editor may have good intentions but be quick-tempered. In fact I was quick-tempered at the early stages and only now I try to keep cool in the heated debates. SkyBon^{Talk\Contributions} 12:35, 26 July 2009 (UTC)

You may claim to speak for them... but I don't think you're properly representing all of them, you need to speak on behalf of the vandals and unsourced content editors, and defend their actions, or you only speak for a subset. Statistics I've seen but can't recall the location of at the moment show that contribution types skew... anons are responsible for more problematic edits on average, per edit, than registered editors are. ++Lar: t/c 11:37, 15 April 2009 (UTC)

I'm not going to bother editing Wikipedia if my changes aren't visible to me and others immediately, and I think this will apply to most people, really. So for the articles this is implemented for, this basically means shutting off edits from and alienating the large group of "Hey, I have some information that could improve this article! I think I'll just go ahead and add it."-users, who only edit once or twice, and only because it is so easy to edit.

It has been my impression that these users really are the lifeblood of Wikipedia, both being a source of a large fraction of the total information here, and also forming the pool from which more experienced editors are recruited. This group is also responsible for most vandalism by far, 'tis true, but this doesn't mean that its contributions aren't important. In my opinion, a fit analogy for implementing this widely is "sawing off the branch one is sitting on".

Finally, let me point at Nupedia, and point out that it didn't take off (becoming Wikipedia) before the heavy editing restrictions were lifted. 84.208.81.75 (talk) 07:31, 25 August 2009 (UTC)

Wikipedia with flagged revisions would not be directly comparable to Nupedia. The big restriction at Nupedia was that only credentialed 'experts' could edit, which greatly restricted the pool of potential editors. The English Wikipedia will still be open to anyone to edit, but it may be a while before the edits become visible to readers. Will that discourage some 'good' editors? Yes, quite possibly. Will it dramatically reduce the crap edits (vandalism, promotional edits, jokes, trolling, POV edits, etc.)? I'm sure it will. -- Donald Albury 13:40, 25 August 2009 (UTC)

"It has been my impression that these users really are the lifeblood of Wikipedia", most IP users are vandals; not opinion but actual fact, that has been analysed time and again. Wikipedia has restrictions at the moment, IPs can't create articles, can't edit protected pages and can't vote on certain community pages. The flagged revisions will take a layer of protection off, meaning IPs can edit more, but that their edits will not be visible immediately. This means that they can still edit, and can edit more pages but the thrill of seeing graffiti on the wall for vandals will be curtailed, which will mean less need for editors to waste their time reverting vandalism, which means more time to actually edit. Wikipedia doesn't need hundreds of new articles, it needs the currnet ones ot be improved. Darrenhusted (talk) 16:26, 25 August 2009 (UTC)

Can you link to one of those analyses please? Thanks! --Falcorian ^(talk) 16:45, 25 August 2009 (UTC)

Actually "most IP users are vandals" is wrong. It has been demonstrated that most vandals are IP users, but that's not the same thing. Only about 20% of edits made by unregistered users are reverted. I should also point out that only about 0.1% of articles are protected, and nearly half of those have some sort of expiry date. Implementing flagged revisions on all articles would open up a tiny number of pages while closing off a huge number of others, so claims that it will actually open up Wikipedia are also wrong. The number of people blocked for vandalism on the German Wikipedia didn't drop after flagged revisions was introduced, so it won't necessarily reduce the number of vandals either. Hut 8.5 18:11, 25 August 2009 (UTC)

Noone said FRs would be implamented on all articles. It'll mainly be a replacement for (semi-)protection. ♫ Melodia Chaconne ♫ (talk) 18:22, 25 August 2009 (UTC)

Flagged revision in the German Wikipedia

I think many doubts could be dispeled if you look at the enormous success that flagged revision have in the german wikipedia!--Don-golione (talk) 00:03, 14 March 2009 (UTC)

Which succes in the German wikipedia ? The current waiting time is 4.6 d/11.6 d for an edit to show with IP contributions down to 12 %, and the waiting is for the new statistics on DE. Mion (talk) 10:06, 14 March 2009 (UTC)

4,6d-11,6d are waiting times for articles that are on nobodys watch list. I voted against flagged revision in 2008, but now I gotta admit that it is a great tool against vadalism. Especially when I compare it with the english wikipedia.--Don-golione (talk) 13:21, 14 March 2009 (UTC)

I prefer an encyclopaedia that is very accurate but out of date by days or months rather than one that has a chance of being inaccurate but updated by the second/minute. Remember when encyclopaedias were out of date by years or out of date at the time of printing? FlaggedRevs will make WP far more accurate but MAY in SOME instances mean that edits will be delayed. -- Alan Liefting (talk) - 00:34, 15 March 2009 (UTC)

That is a baseless statement, there are no statistics that prove that the German wikipedia is more accurate than the English wikipedia, on the contrary, similar to the Encyclopædia Britannica Eleventh Edition, the content from the German wikipedia with its out-of-date content makes its use as a source problematic. Mion (talk) 10:52, 15 March 2009 (UTC)

Who would use Wikipedia as a source? 203.24.135.66 (talk) 22:20, 15 March 2009 (UTC)

Sorry but are you suggesting the German wikipedia is 100 years out of date because of flagged revisions? Nil Einne (talk) 19:07, 4 April 2010 (UTC)

Wikipedia:Flagged protection and patrolled revisions

It is proposed to run a trial of Flagged Revisions at Wikipedia:Flagged protection and patrolled revisions. The proposal is divided in two parts:

• Flagged protection: an article can be 'protected' by an administrator so that the version viewed by readers by default is the latest flagged version. This is a modified version of the original flagged protection proposal.
• Patrolled revisions: a 'passive' flag used to monitor articles, especially blps, for vandalism, blp violations, pov pushing, etc, that can be used for all articles, but has no effect on the version viewed by readers.

The proposals are independent but supplement each other. They involve the creation of a 'reviewer' usergroup. This implementation can support secondary trials. The main trial should run for two months, then a community discussion should decide the future of the implementation. Cenarium (talk) 22:51, 15 March 2009 (UTC)

A poll has started at Wikipedia talk:Flagged protection and patrolled revisions/Poll. Cenarium (talk) 18:33, 17 March 2009 (UTC)

This poll is scheduled to end on 1 April 2009. Cenarium (talk) 21:57, 28 March 2009 (UTC)

Alternative: Time-limited invisibility

Please see Wikipedia talk:Flagged protection and patrolled revisions#Alternative: Time-limited invisibility. - 7-bubёn >t 15:02, 18 March 2009 (UTC)

Rewrite

I have rewritten this page to keep it in line with recent developments, and added info on WP:FLPPR, as this proposal has gained consensus for a trial implementation. Cenarium (talk) 16:48, 11 April 2009 (UTC)

Poll on reviewer autopromotion for flagged protection and patrolled revisions

There is currently a poll on the autopromotion of reviewers at Wikipedia talk:Reviewers#Poll on autopromotion, for the trial implementation of flagged protection and patrolled revisions. For information, see general documentation and overview. All users are invited to comment, and to participate in the elaboration of a reviewing guideline as well. Cenarium (talk) 13:49, 12 April 2009 (UTC)

Question

Has there been a consensus on when/if we're having a trial for this? Just wondering. Jwkpiano1 (talk) 03:28, 19 May 2009 (UTC)

No. Nothing ever happens at the Grand Hotel. -R. S. Shaw (talk) 20:31, 17 June 2009 (UTC)

Do we have all the experts here

Do we seriously consider that, all the reviewers will be experts in all topics or is it just the keeping off the vandals that is the underlying objective? We must consider that, an expert in say on Mining topic shall sit by his PC and wait for his stuff getting approved or may be wait for someone to visit his edit. We may lose on other things which one might want to share out of sudden instinct and spur of moment and may be a valuable piece of information. I franky do not believe that the pace with which the world is building around, the bureaucracy imposed by flagging shall limit the pace of development to the capacity of bottleneck. A lot of articles and headings need revisions and expansions and who has the time to get stuff approved. Roles of admin as defined should be crusaders or barriers is athing that needs decision. Trust or distrust needs to be decided as well. Vertical.limit (talk) 08:55, 17 June 2009 (UTC)

While the terms of promotion to reviewers is I believe unclear, there is almost no chance that reviewers will have to be in a specific topic area. Also there is no need for anyone to sit by their PC, non-reviewers will simply submit referenced material as normal, reviewers will simply approve it when they see it. I'm not sure, but perhaps you are under the mistaken impression that when an article is awaiting review it can't be edited. This is not the case. Nil Einne (talk) 06:31, 18 June 2009 (UTC)

Analysis on German Wikipedia

Has there been an analysis done of FR on the German WP? They have had it in operation for just over year so there should be enough data around in order to do it. -- Alan Liefting (talk) - 11:08, 21 June 2009 (UTC)

Software vs Wetware

This won't work.

It's a lovely idea, and the technology is a wonderful idea. But it won't work.

The reason it won't work is that it will require too great of a change in the mindsets of the Wikipedia community.

Flagged Revisions will drown us, do you understand? IT WILL DROWN US.

Remember when Patrolled Newpages was introduced? We're barely keeping up with that - and we'd have drowned if it wasn't for me personally putting hundreds of hours in it. Do you realize that the instant that Patrolled Newpages was introduced, we immediately got a 720-hour backlog? And do you know how long it took to clean up that backlog?

Answer: IT NEVER GOT CLEANED UP. No one worked on it. Pages expire from the unpatrolled queue after 720 hours. It took months before I noticed the enormous backlog and started acting on it.

Most people only bother checking the first page of Recent Changes / New Pages. Some people will take the time to check the second. A few will check the third. Almost no one checks beyond the fourth. Except me, and the people I've pushed into doing this.

I have whinged people into creating software tools to make the task easier. Brion Vibber redesigned the interface specifically to deal with me, because the method I was using to go to the last page was putting an enormous strain on the server. And even with all this, the backlog is still hundreds and hundreds of hours. Most of the crap gets filtered out within the first 72 hours... but not all of it. And it only takes one piece of garbage to create another Siegenthaler incident.

If you have 1000 new articles, and 50 people volunteering to spend an hour checking 20 articles each, then at the end of that hour maybe 700 of those articles will be checked, if we're lucky. Because no one knows which ones have already been checked. Because no one bothers to click "patrolled".

This is the problem with Flagged Newpages - where we only need to check the article once. We are nearly drowning in it as it is.

With Flagged Revisions, we will need to check every version of every article. We will not have enough people doing that. We simply won't. We don't have the resources to cope with the smaller version; the larger version will be even less feasible. And before you object to my argument, go and patrol five hundred newpages. Click if they're good. If they're not, then click anyway, and then tag them for deletion-or-fixing-or-whatever. And work from the back end.

And then do it again the next day.

And the next.

And the next.

Then come back here and tell me that we can deal with Flagged Revisions.

And remember - patrol is a logged action, so I can see if you've actually bothered doing it. DS (talk) 18:38, 21 June 2009 (UTC)

Except that it's not going to be implemented on most pages. Just a small number. ♫ Melodia Chaconne ♫ (talk) 20:06, 21 June 2009 (UTC)

Waterboarding instead of mass drowning, then? Incidentally, I note that you have patrolled precisely 0 articles. DS (talk) 20:56, 21 June 2009 (UTC)

One interesting property of Special:Newpages is that having the backlog has zero effect on the wikipedia. None. Nada. Zilch. It's strictly an "optional reminder that someone should take a look". I occasionally whack away a few hours from the end of the unpatroled backlog when I'm feeling like doing something mind-numbing, but it's the changes I make that have effect, not the fact that I'm working through the backlog.

In the case of flagged revisions, the result matters to what people are seeing. You still may be right, but there is a difference. --Alvestrand (talk) 01:34, 23 June 2009 (UTC)

Note AFAIK that it isn't true you need to check every version of an article. If 25 people edit an article, and then you review it, you see the difference between the current version and the last reviewed version. You don't need to look at the intermediate revisions unless you want to track down some dodgy edits or whatever but that's no different from nowadays. More importantly, flagged revisions is almost definitely going to cut down on the number amount of vandalism so there'll be less to actually review anyway. Nil Einne (talk) 19:11, 4 April 2010 (UTC)

NY Times reports on Flagged revisions for BLP

I just saw this in the NY Times online:

Officials at the Wikimedia Foundation, the nonprofit in San Francisco that governs Wikipedia, say that within weeks, the English-language Wikipedia will begin imposing a layer of editorial review on articles about living people.

The new feature, called “flagged revisions,” will require that an experienced volunteer editor for Wikipedia sign off on any change made by the public before it can go live. Until the change is approved — or in Wikispeak, flagged — it will sit invisibly on Wikipedia’s servers, and visitors will be directed to the earlier version.

--SteveMcCluskey (talk) 21:38, 24 August 2009 (UTC)

On-line at Wikipedia to Add Layer of Editing to Some Articles -- Donald Albury 22:43, 24 August 2009 (UTC)

I saw that article and am confused; will this policy only affect pages that would previously have been semi-protected? Or all pages about living people?--Lkjhgfdsa (talk) 23:01, 24 August 2009 (UTC)

I understand it to mean that flagged revisions will apply to all BLP articles. This makes sense, given the growing concern for keeping harmful misinformation out of BLP articles. -- Donald Albury 00:00, 25 August 2009 (UTC)

No, the media don't get things right as usual with wp. It'll be applied only for pages meeting the requirements for semi-protection, only a little more liberally for BLPs because there's a limited admin discretion. This is what the community approved : Wikipedia:Flagged protection and patrolled revisions, nothing else. Cenarium (talk) 18:52, 25 August 2009 (UTC)

Not sure if it's entirely relevant, but Mashable (a popular social media blog) just published a story about Flagged Revs: [1]. VI ^talk 23:17, 24 August 2009 (UTC)

Slashdotted as well. Titoxd^{(?!? - cool stuff)} 05:42, 25 August 2009 (UTC)

There needs to be an easy-to-read FAQ targeted toward to the media and the laymen about the changes being proposed. Common questions include: which articles will this effect, who approves the edits, is this permanent, etc. Again, make it clear and concise and it'll help prevent some of the misinformation being spread. Mahanga^Talk 19:25, 25 August 2009 (UTC)

We don't seem to have any official confirmation of which configuration is being implemented, we can't write an accurate FAQ without it. Hut 8.5 20:00, 25 Aug.ust 2009 (UTC)

See information [2]. There is now a lab setup that will be used in the coming weeks as a testing grounds. There is still no fixed schedule as to when the english wikipedia will enable flagged revisions. —TheDJ (talk • contribs) 20:35, 25 August 2009 (UTC)

"Per the wikisignpost twitter, @wiki_nihiltres has a Flagged Revisions primer. May clear up some of the widespread media confusion/misreporting." Mahanga^Talk 14:01, 26 August 2009 (UTC)

• FWIW, there are still close to 1,000 BLP UnCat items, and over 23,000 Bio articles w/o a "Living=" parameter. Does anyone know if they will get added to the FR automatically when properly assessed? (provided that FR gets implemented prior to the completion of those two items). link — Ched :  ?  15:10, 26 August 2009 (UTC)

There has been a post written about Flagged Revisions on the Wikimedia Blog. --Falcorian ^(talk) 16:15, 26 August 2009 (UTC)

Wikipedia Pillar

All edits are equal, but some edits are more equal than others.--194.183.86.147 (talk) 10:05, 19 September 2009 (UTC)

The primary wikipedia pillar should be to inform the reader. Hence some edits are better at fulfilling that purpose than others are. Baseball Bugs ^{What's up, Doc?} carrots 10:10, 19 September 2009 (UTC)

Nope, not true at all :D. You haven't read exactly what Flagged revisions are. Sir Lothar (talk) 18:06, 19 September 2009 (UTC)

+1! SkyBon^{Talk/Contributions} 13:10, 8 October 2009 (UTC)

Whoever said all edits are equal anyway? The Wikipedia:Five pillars definitely doesn't say anything of that sort. It's an extremely silly thing to say. An edit which introduces 'X IS GAY!!!!!!!' 100 times into a page a long with a picture of a penis is clearly NOT equal to an edit that turns a stub into a featured article (not that this is likely even if we include off wiki and subpage work). To be blunt, anyone who thinks these edits are equal should even be editing wikipedia.

If you meant all editors are equal, while that's a common sentiment, most wikipedians would agree it isn't true. It's true that in some cases like determining consensus we generally don't concern ourselfs with who the editor is and per WP:NPA and WP:AGF under normal circumstances it's not acceptable to challenge an editor edits simply because you don't like the editor or his/her history even if the edits themselves weren't problematic. However semi-protection prevents non autoconfirmed and unregistered users from editing. Unregistered users can't create articles. Full protection prevents all but administrators from editing although they generally aren't supposed to (since full protection usually means ongoing edit wars) unless there's clear cut consensus. I'm pretty sure some of the abuse bots have treated anonymous editors differently and I'm even more sure the abuse filter does (I believe it can even be programmed to depend on the IP range). The recent changes page has options to hide different kinds of edits and anyone with any experience with patrolling or simple common sense is far less likely to check out edits by "X LIKES GAY BUTTSEKS!!!" then they are by someone they recognise as an experienced and trusted editor.

Nil Einne (talk) 19:27, 4 April 2010 (UTC)

Survey on an other wikipedia

As for your information, in order to know if Wikipedia-fr users want to plan a procedure in order to have Flagged revision, a survey has been done. The result of this survey (not an official decision, just a survey to start the official procedure for the decision) is that users rejected the Flagged Revisions (for both types: the one used in Englsh Wikipedia, the one used in German Wikipedia).

Manual vandal-whacking enriches Wikipedia

Fixing obviously wrong information is one of the lowest-bar activities able to be performed on Wikipedia. Both my own experience, and the anecdotal experiences of others I know, suggests that performing a fix on obviously wrong content can be a good catalyst for getting people familiar with Wikipedia editing and lead to them becoming a valuable member of the project. Let's not be quick to assume that, even were a method of preventing vandalism that has zero false positives to be available, that it would enrich the project. (Besides which, I like vandal-whacking. It's nice that not everything has to be a struggle to cite policy or find sources; sometimes you can just be right.) - DustFormsWords (talk) 04:53, 6 November 2009 (UTC)

That is an excellent argument. I agree. Whatever happened to Flagged Revisions, anyway? I saw an article in a major newsmagazine today that listed Flagged Revisions on Wikipedia as one of the things that happened in 2009. Did they get that one wrong? All Hallow's Wraith (talk) 16:55, 7 December 2009 (UTC)

I find vandal-whacking to be a deeply frustrating activity. It's so obviously necessary, there are so many of them, and despite the many forms it takes, they show no creativity whatsoever. I would enjoy Wikipedia more with fewer vandals to whack. --Alvestrand (talk) 18:10, 7 December 2009 (UTC)

Looking for more information about legal question

A while ago, I asked a question about flagged revisions putting legal burden on those who green-light revisions. (The discussion is here Wikipedia talk:WikiProject Flagged Revisions#Legal question about flagged revisions.) I feel this question is directly raised by the stated purpose of flagged revisions to protect biographies of living people from libelous content. Yet the only response I got to my question was a single, "We don't know"-type answer. This was disappointing.

I haven't followed the issue every day. I hope some information related to the answer has appeared in the meantime; so I was wondering if anybody has some links to other discussions that try to answer this specific yet fundamentally important question. I find it beyond belief that the Wikimedia Foundation can be on the verge of making a change like flagged revisions yet not have a (very good) answer to somebody worried that they are setting themselves up for a legal trap. Jason Quinn (talk) 23:00, 18 December 2009 (UTC)

FlaggedRevs is a MediaWiki extension, released on June 4, 2008

It was released in June 2008? Why hasn't this been implemented yet, then? Are these part of the marvellous "teachings" of the Foundation that will purportedly make the other wiki communities "more efficient and educated"? SalaciousBreadcrumb (talk) 18:11, 15 January 2010 (UTC)

Well (a) it took a while to get consensus on whether to use the feature and what to use it for on this wiki, and (b) the implementation that eventually gained consensus hasn't been finished by the developers yet. Hut 8.5 19:25, 15 January 2010 (UTC)

Then why don't make a "closed beta" for the whole project, which isn't indexed thorougly by search engines? The most likely answer is "because JimboMimbo can't get money", but that follows greed, not honesty. SalaciousBreadcrumb (talk) 20:02, 15 January 2010 (UTC)

There is already a open beta of the FlaggedRevs extension [3] as discussed in this very wikipedia page since October 1 so to be honest I'm not really sure what you're talking about. Forking en.wikipedia and implementing the beta flagged revisions on one of the forks is a rather silly idea if that's what you were suggesting. Although there's only one developer working on the project it's been stated the reasons for the delay have little to do with funding and adding more isn't going to help. Nil Einne (talk) 19:02, 4 April 2010 (UTC)

Real-live article example

Damir Ibrisimovic has an idea for rewriting the perception article which involves mechanisms for locking data on pages. Sounds like flagged revisions, no?

If possible, we would like to use the Wikimedia labs machinery to create the rewritten article. I logged in and it looks like editing can occur now. It just wouldn't be prime time as it would be on the encyclopedia. But Damir Ibrisimovic might be just as happy using the machinery as it stands. It would also serve as proof of concept at some level. --Ancheta Wis (talk) 02:41, 10 February 2010 (UTC)

At what point do we stop being a wiki?

A wiki is "a website that allows the easy creation and editing of any number of interlinked web pages." I wonder, as we introduce more restrictions and review processes before a created or edited page can go live, at what point do we cease to be a wiki and become just one of the many non-wiki sites out there (e.g. newspapers that invite letters to the editor) that review content submissions and posts the ones that they accept? There are a lot of sites, like MySpace, that only allow users to freely edit certain pages, while others are under the control of site administrators, although you can give them suggestions on what to put on those pages. There might come a point at which Wikipedia is less like a wiki than many of the Internet that we don't normally consider to be wikis. Tisane (talk) 19:34, 25 March 2010 (UTC)

There is no point where that happens for the wiki. Flagged revisions is switchable on and off for individual articles. When/if an article is by consensus finished, would you still want vandals editing it? Would you still want people that have a 99.99% chance that they cannot improve an article because there have been tens of thousands of people editing it already, would you want to allow them to still edit it directly?- Wolfkeeper 20:05, 25 March 2010 (UTC)

The point of the project isn't to create a wiki anyway, it's to create an encyclopedia.- Wolfkeeper 20:05, 25 March 2010 (UTC)

There are plenty of wiki's with far more restrictive policies then we'll ever have. For example, some are only editable by a very small number of editors, without any registration system (commercial sites and some university sites for example, e.g. [4] or the US intelligence agency wiki). Nil Einne (talk) 18:58, 4 April 2010 (UTC)

Don't stop WP from being a wiki

About 5/6 of IP users make constructive edits only.

I honestly believe that the majority of IP users are trying to help Wikipedia. Remember that not every IP is a vandal. Also, the vast majority of Wiki-users don't have an account, so IP edits just help us to see the average reader's perspective on things. If we added this policy, even if it only was for BLP articles, new users would be discouraged from creating an account and contributing to Wikipedia. If anything, this policy should only be used on semi-protected articles, but hopefully we won't need to implement it at all. I strongly oppose this idea. I recommend reading WP:HUMAN and WP:IP!=VANDAL for further information. Ilikepieittastesgood 03:26, 23 June 2010 (UTC)

If 8% of edits are vandalism than another 8% must be constructive edits to revert vandalism which makes 16% of edits on en.Wikipedia a complete waste of time and energy. Proportionally, a lot of the energy of constructive editors is vasted on constantly monitoring pages for recent changes. As articles improve this is only likely to increase in the future. Flagged revisions improves the quality of editing and appearance of Wikipedia, as observable on other Wikis which implemented it, without restricting contributions by IPs. It is actually a more open system than the current one of protecting articles. --Elekhh (talk) 23:55, 23 June 2010 (UTC)

Are you saying it's good for BLP articles or do you want it on the whole wiki? Please clarify. Ilikepieittastesgood 03:28, 24 June 2010 (UTC)