Wikipedia:Flagged revisions/straw poll

The following discussion is preserved as an archive. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

---

I'm calling an absolutely non-binding straw poll to help gauge the ground and provide some statistics on people's opinions of flagged revisions. Please participate, and explain your reasoning. – Thomas H. Larsen 23:53, 19 September 2008 (UTC)

a) Flagged revisions should be enabled over all articles

Vote (a)

1. Necessary for Wikipedia to gain credibility as a reliable resource; and reliability is more important than openness. – Thomas H. Larsen 23:53, 19 September 2008 (UTC)
2. Yes, we are already recongnized, it is now important to make us more credible, flagged versions will reduce the number of people that do not understand wikipedia and finds a page with questionable content which is hard to revert with BOTS today. It is more important that having IP editors see their edits directly, if they make a edit it should still show up quite soon. I also support b), c) and d) (and g)), but if I can only #vote once this is it. It is also important to declare what version would be seen by a IP editor, I assume that would be the last flagged version, but it is not clear in this poll. --Stefan ^talk 04:43, 20 September 2008 (UTC)
3. First choice. I think this would go a long way to stem the tide of vandalism and increase the public's perception of the reliability of Wikipedia. (I had a professor, of all people, tell me in class the other day that he goes out of his way to add inaccurate information because of perceived reliability issues.) b), c), and d) are my second choices, with the caveat that once flagged revisions get applied to a page, it should take overwhelming consensus and at least an admin to remove them. shoy (reactions) 05:53, 20 September 2008 (UTC)
4. It seems to work on the German Wikipedia, I see no reason why it shouldn't work here. --Zvika (talk) 09:50, 20 September 2008 (UTC)
5. My support goes to this option which I know from the German Wikipedia. It functions very well there. Flagged revisions will significantly increase the reliability and thus credibility of Wikipedia. --Eleassar ^{my talk} 17:56, 20 September 2008 (UTC)
6. If Wikipedia ever hopes to grow into a mature project we need this feature implemented across all articles. I can't see any reason why we would want to enable it for some, but not all articles, given the benefits that this feature provides. Kaldari (talk) 07:39, 21 September 2008 (UTC)
7. Enable the revisions for all pages, but make sure that the most current version of the page is what is displayed, not the flagged one. Judgesurreal777 (talk) 12:36, 21 September 2008 (UTC)
8. Ideally, this is the best option. SirFozzie (talk) 16:14, 21 September 2008 (UTC)
9. I see a lot of benefit, and very, very little negative.--ragesoss (talk) 16:52, 21 September 2008 (UTC)
10. Let's give it a try! I'd be happy with either (a) or (d). Walkerma (talk) 19:24, 21 September 2008 (UTC)
11. Yes. I don't see any real advantage to the other ones with vandals that vandalize random pages. Xclamation point 22:28, 21 September 2008 (UTC)
12. Will enable us to reduce the number of semi- and full-protected pages. Tim Vickers (talk) 00:38, 22 September 2008 (UTC)

    at the cost of effectively semi protecting every page for rather a lot of users.Geni 02:39, 22 September 2008 (UTC)

    No, that would be if this prevented people from editing. This does not prevent people from editing, it only delays their edits from being displayed to the public by a short period. As an analogy, if it takes me five minutes to get a sandwich, rather than being provided a sandwich instantaneously, I have not been "effectively prevented" from getting a sandwich. Tim Vickers (talk) 18:17, 22 September 2008 (UTC)

    If however you have been offered british rail levels of service (it will take far more than 5 minutes) you have effectively been prevented from getting a sandwich.Geni 15:33, 24 September 2008 (UTC)

    So you don't oppose this on principle, and would support this if revisions were flagged quickly? Tim Vickers (talk) 16:39, 25 September 2008 (UTC)

13. If we care about creating a reliable, free encyclopedia, flagged revisions are a good addition to Wikipedia. Awadewit (talk) 17:15, 22 September 2008 (UTC)
14. Fuck yes. Assuming this is an attempt to build an encyclopaedia collaboratively and not a public circle-jerk. Why not take the motivation out of vandalism and put it into writing good content? WilyD 18:15, 22 September 2008 (UTC)
15. Soonest. -- Jeandré, 2008-09-22t20:16z
16. I have been an advocate for some time. — Carl (CBM · talk) 21:11, 22 September 2008 (UTC)

    This one or (d) are both fine with me. See comments at (d) and discussion for (e). Switching to (c) to allow experimentation on reviewed articles. - Dan Dank55 (send/receive) 22:09, 22 September 2008 (UTC)

17. Finally. Maybe someday I'll be able to cite Wikipedia in class </somewhat-sarcasm-but-it'd-be-cool-if-it-happened>. ffm 23:10, 22 September 2008 (UTC)
18. Yes please, as soon as possible. Lumos3 (talk) 13:19, 23 September 2008 (UTC)
19. This would help alot. LegoK^ontribsT^alkM 04:21, 25 September 2008 (UTC)
20. Sure. It's been working pretty well over at Wikinews so far. Cirt (talk) 13:01, 26 September 2008 (UTC)
21. Hell yes! Plrk (talk) 17:44, 26 September 2008 (UTC)
22. Worth trying. PhilKnight (talk) 23:55, 27 September 2008 (UTC)
23. Absolutely overdue. Turn it on yesterday. rootology (C)(T) 13:26, 2 October 2008 (UTC)
24. Very much needed. Just make sure the criteria for allowing an editor to flag allows for plenty of established editors. AlexJ (talk) 19:38, 3 October 2008 (UTC)
25. I most definitely believe we need flagged revisions. When I talk to people outside the project, the most common comment is "none of it is reliable because it can be, and is, constantly vandalized." I took a few years off vandalism patrol because it got to me, the non-stop flow of crap into this project was staggering. I recently returned to patrolling, believing that the rise of anti vandalism bot's and IP's unable to create articles, would have slowed things down. I was sorely mistaken. Nothing has changed in years, mass deletions and replacements ("this page is haxed, lolz") are still rampant. I can't imagine the amount of progress this project would see if the the hundreds of active users constantly reverting and warning could instead focus on editing. Yes, people would have to take the time to pick a flagged revision, but it would be far less work the check a popular page twice a day than revert crap 20 times and warn 5 IP's four times each before banning them for 24 hours. - Trevor MacInnis (Contribs) 00:59, 4 October 2008 (UTC)
26. Overdue by far. If enabled it will help Wikipedia become a more reliable and trustworthy source, and cut down vandalism at the source. Ironholds 02:29, 5 October 2008 (UTC)
27. While the exact specifics still must be hammered out (and I of course prefer my own proposal from wt:flagged revisions) this is a good idea. Hopefully we can agree to proceed with this and put to general discussion and vote some of the many good proposals. Random89 02:50, 6 October 2008 (UTC)

Discussion (a)

The problem will be that we won't have enough users to flag all articles, or they'll be unreliable. We'll be completely lost in the number of articles, and we won't be able to focus on the articles which really needs flagged revisions. We can't compare efficiently with dewiki, because enwiki is much bigger, we have more disputes and persistent vandals. I suspect that we won't have immediate consensus for massive enabling, for all the issues mentioned, but we may have enough support with more reasonable goals. Cenarium ^Talk 00:03, 22 September 2008 (UTC)

I would oppose this per my comments below. Cenarium ^Talk 02:41, 22 September 2008 (UTC)

I have full faith in our ability to have all articles flagged in, say, two years, based on how quickly 1.0 assessment caught on. I'm not saying there aren't maintenance issues to be considered, but "not enough users" is not, imo, an issue. Nifboy (talk) 04:50, 22 September 2008 (UTC)

Two years and that's actually good?! That's even worse than I assumed and I'm the first total opposer on this straw poll. It will definitely take a very long time to find out was this worth the trouble and are there new problems emerging. Admiral Norton ^(talk) 12:09, 22 September 2008 (UTC)

Two years is a fairly conservative estimate, but I also think there's an Internet time problem: We've long since passed the point where we can radically improve Wikipedia overnight. Flagged revisions went live on de.wiki back in, what, May? and we're still talking about implementing it here? Nifboy (talk) 10:46, 23 September 2008 (UTC)

I think this is, if anything, exactly the opposite of the problem that'd develop. Right now we have a huge pile of recent pages patrollers, many of whom are not particularly skilled at encyclopaedia writing. Assigning them (effectively) to reviewing new revisions, there's easily enough manpower for it. The real concern is that there'll be too little vandalism and whatnot to deal with, and we'll be left with unemployed editors with nothing to do. WilyD 11:15, 22 September 2008 (UTC)

We will never have unemployed editors. shoy (reactions) 22:09, 23 September 2008 (UTC)

RC patrollers don't review all recent edits, but only a part that is likely to be vandalism as guessed by the software used. And if a page has never been sighted before, they'll have to check all the page before sighting to make sure it's OK; and I'm sure that many times, they'll sight while it's not OK (just checking the latest edit, and sight). So it's another reason we need reliable users, so we need to restrict this user group (like Rollback). And again, we won't have enough people out there to sight every page. Even so, they'll sight randomly, and pages with a real need to be sighted (not necessarily the most viewed) won't be given due attention. But the most appalling is that a mass of good non-surveyor (not only IPs!) edits will be missed for days. Cenarium ^Talk 11:17, 24 September 2008 (UTC)

In practice, I believe the articles most in need of review would be reviewed soonest, while more esoteric substubs would linger until some "final sweep" catches them. But this isn't a problem - the natural instinct to review articles when you see them means the most-seen articles would be reviewed early. — Carl (CBM · talk) 21:16, 22 September 2008 (UTC)

Can you assure me that any non-automatically sighted edit to a page which has already been sighted will be reviewed in less than a few minutes ? A progressive "rollout": (b), then (c), (d) and then finally (a) would help to see what works and what doesn't work and allow to adopt appropriate strategies to counter the problems. While a all out, massive enabling may create enormous problems that we haven't necessarily anticipated, and then the community may decide to shut down flagged revisions. Cenarium ^Talk 11:17, 24 September 2008 (UTC)

It seems a fair number of people support this cockamamie nonsense because it will make Wikipedia more reliable. I submit that Wikipedia will never be, EVER be reliable. Not in the sense that editors here seem to think anyway. No project, flagged revisions or no, that allows anyone, absolutely anyone, to sign up and edit will ever be seen as reliable. Not to say there isn't good content around here but seriously, it just won't happen. This is very, very un-wiki, and based on what I have observed and been privy to at en.wikinews, will be a total clusterf*ck.--IvoShandor (talk) 10:21, 25 September 2008 (UTC)

It is extremely clear based on all of the support comments that people are getting caught up in this and misunderstanding things. More important than reliability is this "a world in which every single human being can freely share in the sum of all knowledge." (Quoted from Wikimedia's Home Page.) It sounds like this proposal wants it to say. "Imagine a world in which every single human being can freely share in the sum of all knowledge (pending the review of a registered editor who has at least 60 days of editing experience and has taken the steps to get a username.)" That is ridiculous. Scottydude _review 16:38, 27 September 2008 (UTC)

No. This proposal merely wants it to say, "... a world in which every single human being can freely share in the sum of all reliable knowledge," instead of "... a world in which every single human being can freely share in the sum of knowledge, vandalism, crank theories, spam, and POV edits." Quite frankly, I disagree with you when you state that reliability pales in importance to bringing knowledge to all people; that simply isn't a comparison, as knowledge might be defined, if you like, as "reliable, accessible information". If information isn't reliable, it isn't knowledge – at least in my books. – Thomas H. Larsen 00:43, 29 September 2008 (UTC)

I just don't think we should make such a drastic change as screening edits to a "wiki" encyclopedia. As IvoShandor said, Wikipedia will never be completely reliable, sure there are efforts to be made, but screening revisions and creating a hierarchy isn't for Wikipedia. That should be left to Citizendium. I am not trying to sound as if I am against the improvement of the content here (and I apologize if I sounded that way.) I am trying to say, however, that flagged revisions is the wrong way of doing things. Scottydude _review 19:45, 29 September 2008 (UTC)

Part of the 'buzz' for vandals is that their edits are displayed live instantly on the encylopedia. In theory, if vandals' edits are not going live onto the encyclopedia, most of them won't bother. Therefore the number of edits that need filtering out will be greatly reduced over a fairly short amount of time. AlexJ (talk) 12:58, 4 October 2008 (UTC)

Six months ago I would have agreed with you. Today, I cannot understand why this page is full of speculation about the impact of flagged revs when it has been live on the German Wikipedia for months. We really need a report from those who implemented the German flagged revs, which was supposed to be a trial run. In March, the Wikimedia Foundation received its largest-ever donation which was specifically supposed to help bring in quality maintenance including Flagged Revs. There really should be an obligation on the WMF to use some of that funding to make (and publish) an assessment of the trial run, to help other WMF projects decide whether to implement flagged revs and if so in what manner. Until we have such an assessment, I don't think we should implement anything here. PaddyLeahy (talk) 19:44, 4 October 2008 (UTC)

Clarification needed

I suppose that this poll concerns only the proposal of sighted revisions as described at Wikipedia:Flagged revisions/Sighted versions, with possibly some adjustments. Cenarium ^Talk 13:52, 23 September 2008 (UTC)

b) Flagged revisions should be enabled only over some articles, decided on a per-article basis (i.e., requiring consensus to be gained on an article's discussion page for the article to be flagged)

Vote (b)

1. I think we should start with this. Allowing flagging of every article immediately could potentially get out of hand. With this it would be easier to scale where possible. Though any of the options that involve using it are better than not using it at all IMO. Mr.Z-man 00:02, 20 September 2008 (UTC)
2. Over all articles would be unworkable at this time. First off, we wouldn't have enough surveyors or they would be unreliable. And during disputes, flag wars would occur, so we need to protect a page from flagging. Not on a categorical basis yet, we would be flooded. And flagging revisions would be too easily abused by vandals if we lower the restrictions (do not want to say how). I've proposed at User:Cenarium/sighted revisions to create a user subgroup of surveyor to handle disputes and restrict some sensitive actions to them, so that we can safely increase the number of surveyors. Cenarium ^Talk 01:39, 20 September 2008 (UTC)

   It is said on the talk page, but I think we'll agree that if the discussion is on some noticeboard, provided this is linked from the talk page, it is acceptable. Cenarium ^Talk 17:41, 23 September 2008 (UTC)

3. We should take it slow in the begining. Particulary because a sizeable number of editors and users will only learn about this with time and will certainly have constructive criticisms. If it is implemented suddenly, those users who are not aware of this discussion might feel left out. Not all valuable editors keep up with the going-ons within wikipedia.--Shahab (talk) 09:17, 20 September 2008 (UTC)
4. This is the only way we can get flagged revisions to increase the role of unregistered and new users. Hut 8.5 17:14, 20 September 2008 (UTC)
5. I think that this is the best option, because it will ensure that quality revisions end up being sighted. Of course, a "flagging 3RR" will have to be instituted.--Danaman5 (talk) 22:38, 20 September 2008 (UTC)
6. This is the best option. Basically any article that is currently indef. semi-protected (Bush, Obama etc.) should have this applied. The categorical based system is really easy to abuse (think about it) and applying it to every article will simply result in logged-in users WP:OWNing articles over IP editors. ~ User:Ameliorate! ^{(with the !) (talk)} 10:49, 21 September 2008 (UTC)
7. Definitely the best choice, and no more "unwiki" than our current practice of slapping indefinite semi-protection on some pages. Also would effectively allow a test implementation to be done organically. --erachima talk 17:10, 23 September 2008 (UTC)
8. Definitely, basicly same thing a protection and sanctions. Third choice Pie is good ^{(Apple is the best)} 19:29, 29 September 2008 (UTC)
9. Good for would-be semi-protected pages and some high use pages. Aaron Schulz 20:14, 2 October 2008 (UTC)
10. This seems to be somewhat more friendly to IP editors IMO. FunPika 09:53, 6 October 2008 (UTC)

Discussion (b)

• Seems like an opportunity to have the same fight we've been having for the past year, but over and over again on each talk page. Couldn't we come to some agreement here on what kinds of pages we want to run the experiment on? - Dan Dank55 (send/receive) 17:12, 23 September 2008 (UTC)

Cenarium just added a comment above, and I can live with that: "discussion is on some noticeboard", and then each page that wants to discuss it just links to the noticeboard. As long as you're discussing it in one place instead of 100, that could work. - Dan Dank55 (send/receive) 17:55, 23 September 2008 (UTC)

The idea is that a lot of articles will need sighted revisions for similar reasons, so we can discuss them in a centralized place. However, this shouldn't be used for entire categories, since each article must prove that it needs it (repeated blp vios, vandalism, etc). Yhis will avoid generalized discussion and focus the debate. It's a reasonable beginning, and should be attempted before we enable this on a lot of pages. For example, we could discuss the pages, especially blps, related to the US presidential election, another discussion to decide it for TFAs, for ITNs, etc. Cenarium ^Talk 11:29, 24 September 2008 (UTC)

c) Flagged revisions should be enabled only over some articles, determined on a categorial basis (for example, biographies of living persons)

Vote (c)

1. This would prevent many a complaint and law suit. Zginder 2008-09-20T18:08Z (UTC)
2. Switching my vote to (c), so that we can experiment on some set of articles that people care about and that get a lot of eyeballs; I'd suggest we limit it to featured articles, good articles, and A-class articles, but the exact selection isn't that important. The Foundation is particularly concerned about WP:BLP issues, and I am too, but we could pull biographies into the experimental phase by giving as many as possible some form of review. That's the way to get good data so that we'll know how to proceed. I don't have strong feelings about what we do; anything will give us data. I respect the concerns of the opposition that if we're going to experiment, we have to agree beforehand that this is an experiment; that is, I'd like to see some kind of consensus on what we're looking for from the data and acknowledgment of possible upsides and downsides before we proceed. I'm not "caving"; I think there's a risk of crashing and burning if we do too much too fast before we experiment. After the experiment, if there's a consensus to proceed with (a), then I'm fine with that. - Dan Dank55 (send/receive) 14:23, 23 September 2008 (UTC)

   3 side benefits: while the experiment is running, people might have a greater incentive to get their articles up to GA or whatever quality and submit them for review, in order to get whatever "protection" (Sighted revisions, Quality revisions, whatever) we're offering at that level; and of course, if you want someone to do a Sighted or Quality review, it's more efficient, and we'll get more useful data, if you already have an experienced reviewer looking at the article. (This suggests two options: add articles to the experiment only as they are reviewed, or throw all the FAs, GAs and A-class into the pot at once, on the theory that we've got notes from the previous reviews to help us.) Third, if you launch into (a) and it fails, you're back to the "do nothing" position, but if we run a Sighted and/or Quality test on reviewed articles, nursing it along so it doesn't fail, and then we try (a), and (a) fails, we've got a much better fallback position. - Dan Dank55 (send/receive) 14:55, 23 September 2008 (UTC)

3. I think it premature to implement this on an en.wikipedia-wide basis. We should identify the article categories we think are the most likely to offer unreliable or potentially defamatory content and practice with them. I suspect that this will still be a large number of articles, which should help us see the potential problems without impacting the entire encyclopedia. Karanacs (talk) 17:13, 23 September 2008 (UTC)
4. This is my preferred option. I think we need to see how this will work for a subset of articles before rolling it out across the whole of en.wikipedia, and I share the concerns expressed in (b) that discussing this on a per-article basis risks proliferating the current disagreements a hundredfold. Scog (talk) 09:51, 24 September 2008 (UTC)
5. I like Dank55's idea of trying it with high-quality articles first. The reason is that those may be the ones that need it the most. In my experience, anonymous editors very rarely add anything constructive to a good article (I'm using good in a generic sense here), because there is not that much that can be easily be added to the article anymore, and the article has a higher profile that attracts vandals. In contrast, I often see anonymous editors adding useful information to stubs. This is because it is much easier to find something to add, and it is less attractive to vandalize a stub. There are also high-importance articles that are not "good", where flagged revisions could be helpful too. But we could leave those for later. --Itub (talk) 15:59, 24 September 2008 (UTC)
6. Yeah, one of the gimmicks that makes Wikipedia what it is, is "almost-live" coverage of current events. But, on a discretionary basis, this could help. ViperSnake151 01:24, 25 September 2008 (UTC)
7. I like Dank55's idea of trying it with high-quality articles very much, so I vote for it SyG (talk) 17:42, 26 September 2008 (UTC)
8. I support the principle of flagged revisions whole-heartedly per the arguments given above in (a), but I also see the sense of Dank55's reasoning. It's wise to do an initial test run on a small subset of articles before launching into the full set of articles; we can diagnose and fix problems that crop up, and the results may allow us to choose more wisely between alternative systems of flagged revisions. I also like the idea that (1) the prospect of flagging might encourage people to develop their articles, and (2) anonymous people can still contribute freely to most articles, e.g., the son who logs on to announce his father's death.

   High-quality articles are an attractive subset, because they're stable (and, hence, unlikely to be added to constructively), as well as the articles of WP0.7; both subsets seem likely to shape public opinion of WP. If we can point to a subset of articles and say, "You can count on these articles being good any time you look at them.", that would be great, since many outside people say that they don't trust any article in Wikipedia. The rate at which constructive edits are made to high-quality articles is low, but they're highly watched as well; that suggests to me that actual improvements wouldn't have to wait long before going "live". Another idea: if the goal is to reduce the impact of stray vandalism, we could test flagged revisions on the most-vandalized articles; in theory, that would give the greatest benefit for the fewest articles flagged. Willow (talk) 18:02, 26 September 2008 (UTC)

9. Way too bureaucratic to implement on every page, but certain categories probably would benefit (e.g. high quality articles, BLPs, and high-traffic articles on the main page) --Patar knight - ^chat/_{contributions} 21:28, 28 September 2008 (UTC)
10. Yeah, this could be used for high quality BLP articles, and other. Second choice Pie is good ^{(Apple is the best)} 19:30, 29 September 2008 (UTC)
11. Good place to start, iron the bugs and look at moving to a) all articles down the track....basically support for all reasons stated by #'s 2, 5 and 8 above, and basically think BLPs would be a good place to start because they are a) quite dynamic, b) often edited, c) frequently vandalised d) most likely to throw up issues which is what you need for a trial rollout. Should also include a more static less vandalised category as a comparison.--ClubOranje^Talk 09:55, 30 September 2008 (UTC)
12. 2nd choice--all BLPs at a dead minimum. rootology (C)(T) 13:27, 2 October 2008 (UTC)
13. I support both this proposal, and additionally a variant whereby a consensus of administrators—as reached on, for example, the administrators' noticeboard—may impose flagged revisions on an article as an alternative to blocking, protection, and/or related disruption-stemming actions. (Additional rationale: permitting enabling of flagging on a category-by-category basis is more sensible from the point of view of easiness of administration. Toggling one switch is easier than toggling a hundred.) Anthøny ✉ 00:21, 4 October 2008 (UTC)
14. I would consider supporting it on all articles eventually, but given enwiki is much bigger than dewiki, I think starting it off on as small set of articles where it is most needed to sort out issues and bugs is a good idea. Pages it can be tried out on include stable high quality articles, heavily vandalised articles, and BLPs. It could be implemented in a similar way to protection, often as an alternate to it. Camaron | Chris (talk) 19:23, 6 October 2008 (UTC)

Discussion (c)

Going by what I see on OTRS no it wouldn't.Geni 02:42, 22 September 2008 (UTC)

• Second choice, after (d), with "All WP:0.7 articles" being the target category, since that's where it will easily do the most good. Nifboy (talk) 16:19, 23 September 2008 (UTC)

d) Flagged revisions should be enabled on any article at flaggers' discretion

Vote (d)

1. If someone goes through the trouble of flagging an article, that should be respected. m:The Wrong Version will exist regardless of flagging, and as such I'm not spectacularly worried about it. Encourage IP users to be bold when articles suck, and encourage our core contributors to seek out and maintain quality where it exists. Nifboy (talk) 02:53, 20 September 2008 (UTC)
2. First choice with it being tied in with admin, and if possible, the group being created that is being mentioned Wikipedia:Village_pump_(proposals)/Persistent_proposals#Proposal:__Tiered_deletion_with_established.2Fexperienced_user_group here. Pie is good ^{(Apple is the best)} 19:34, 29 September 2008 (UTC)

Discussion (d)

Question: I really like the sound of this, but before I vote, can you elaborate on how this would differ in practice from being enabled on all articles? Sorry, I've been away from this discussion for a while. Walkerma (talk) 03:45, 20 September 2008 (UTC)

My understanding is that this option discourages going through and flagging all two-and-a-half million articles that we have. Option 2, meanwhile, is too bureaucratic for my tastes and not bold enough. Nifboy (talk) 04:33, 21 September 2008 (UTC)

e) Flagged revisions should not be enabled at all

Vote (e)

1. Per all the things I have written at Wikipedia talk:Flagged revisions. Admiral Norton ^(talk) 18:29, 20 September 2008 (UTC)
2. The damage outweights the benifits.Geni 08:53, 21 September 2008 (UTC)
3. Per other discussions. This is really the wrong idea. --Patrick (talk) 14:23, 21 September 2008 (UTC)
4. Don't think they'll help. Stifle (talk) 15:27, 21 September 2008 (UTC)
5. I fear that flagged revisions would become a POV pushers charter. Poltair (talk) 19:06, 21 September 2008 (UTC)
6. I was a first highly in support of this proposal, but it gets worse every time I think about it. This proposal only seems good on the surface and at glance, but has many flaws. Will this really help our credibility? We must really discuss this before we change the name of "Wikipedia:The encyclopedia anyone can edit". I quote equazicon: ""This is a phenomenally bad idea that will (possibly) spell the end of Wikipedia as we know it. This is the beginning of a members-only club, everything we swore to hate. This basically means anonymous users can't make live changes to the publicly-displayed pages. Registration is now required. You can not edit this page right now. And so forth. This is it. The end."" Sure anons 'can' still edit freely, but when they edit and do not see their changes live, they will think that their edit was just a mere article suggestion and will not contribute and spend nearly as much time to perfect their edits, if at all. Besides, over 80% of anon edits or something are positive to the encyclopedia. Are we really going to have the contributors ready at hand to review all the unsighted edits, besides, lazy people like me are going to edit an article without reviewing the previous edit making the potentially bad edit now sighted. Also issues brought up in Wikipedia_talk:Flagged_revisions#Situation on the german Wikipedia, Wikipedia_talk:Flagged_revisions#German editing statistics and many other personal experiences tell me this is a bad thing. Wikipedia is already the best thing out there, wonderful tools like Huggle have and will stop the majority vandalism, but this, all this does is sway editors away. I want you to think really hard about your first edit to Wikipedia. . . would you have continued editing if your edit did not show up live, having no knowlege in the process? Would you have edited enough to gain enough intrest to create an account? I know I wouldn't have. I would have quickly lost interest and found another hobby, not knowing what just happened to my 5 minutes of editing. I suggest we take a whole new approach to this. Maybe put a red star next to all anon edits on RecentChanges and pages listed on special:unwatchedpages, so people can regularly check them and keep articles top quality in ways flaggedrevs cannot, or update our vandal fighting bots. Wikipedia was founded on a since of trust (read Jimbo's userpage), I'd like to keep it that way.-- penubag  (talk) 00:26, 22 September 2008 (UTC)
7. Absolutely not. Wikipedia policies should place openness front and center. Those of us who are editors and admins don't see this up front. We see an encyclopedia which has quality problems and vandalism issues. We see simple policies which IP and new editors disregard. We don't realize that wikipedia is a complex beast with a maze of social norms and policy intersections because we deal with these norms and rules every day. Our policies should be primarily focused on allowing seamless participation from new editors and IP editors. The more we limit participation the harder we make it for people who don't understand wikipedia to add to our base of knowledge. That way leads the Heat death of the universe. Our userbase will ossify and our outlook will turn inwards. We should enact policies that limit participation only when they are absolutely required for the community. This isn't. It is, in fact, the opposite. We should not create a review process (however limited at first) where an outside editor has to have their contributions "reviewed" by someone from inside wikipedia. Period. I oppose the enactment of flagged revisions wholeheartedly. Protonk (talk) 03:34, 22 September 2008 (UTC)
8. Not, per Protonk, Wikipedia policies should place openness front and center, NPOV quality improvement is based on the fact that all edits can be directly seen and corrected Mion (talk) 21:46, 23 September 2008 (UTC)
9. No. Adding sighted revisions is an unnecessary addition of bureaucracy. It also directly contradicts our "open editing" pledge. On smaller wikis with less activity, sighted revisions may work, but we have upwards of 150 edits per minute 24 hours a day, seven days a week, 365 days a year. There is no way that we can possibly keep up with that kind of volume. If we are going to introduce sighted revisions across the board, we may as well require registration to edit, because that is the effect that sighted revisions will have on this project. For those worried about a lawsuit, the case concerning libel on Wikipedia that was concluded recently sets a precedent for not holding Wikimedia responsible for edits by people using the various project sites. I was originally going to support allowing sighted revisions on some pages after gaining consensus on the talk page, but Protonk's scenario below seems, to me, very likely. I would be willing to support using sighted revisions if they were used only as an absolute last resort before full-protection in very extreme cases. Consequently, I would only support sysops having access to the +reviewer flag, and if implemented, I would like to see sighted revisions used as an intermediate level of protection between semi- and full-protection i.e. at the admin's discretion, but only in response to an actual need, not just "because". Also, after the original reason for protection was no longer valid, the "sighted" status would be removed. J.delanoy^gabs_adds 17:41, 24 September 2008 (UTC)
10. Worst. Idea. Ever. This is a clusterf*ck over at Wikinews. Just because GodKing thinks so, doesn't make it a good idea. What's the point? To make Wikipedia a better source, face it, a project that anyone can sign up for as an editor will NEVER be reliable in the sense that some think it will be. --IvoShandor (talk) 10:14, 25 September 2008 (UTC)
11. No per everyone's ideas above. Sure this may make Wikipedia more reliable but thats not what this is about. Wikipedia bases itself off of the idea that anyone is free to edit. Although some could argue this will still be true in the future it would not be! Edits imply real time, active contribution. Sure there is vandalism, there will always be vandalism, but that has not stopped the editing of this amazing project before. There are thousands and thousands of quality anon edits, all of whom would now be subject to a quality assurance check. Even worse, those people would be editors just like them except they have an account. That's not fair! If anyone can edit this encyclopedia than no one should be able to hide those edits. There is a huge difference between reverting or undoing and edit and hiding an edit until it is verified. Activating this feature would be a huge step in the wrong direction. Scottydude _review 16:27, 27 September 2008 (UTC)
12. This idea only really works with obvious vandalism, and obvious vandalism is - well - as obvious to the general public as it is to the "approved" editor/reviewer. The genuine credibility gap in Wikipedia is when an expert (or at least someone knowledgeable about a subject) is reading a topic they know something about and they see that there are mistakes. Most approved editors/reviewers will not have that level of understanding of a topic or they'd already be editing the article, so most articles will either be unflagged or - more likely - flagged incorrectly, which is even more damaging to our credibility. I'd rather that someone with knowledge of a topic were able to make an immediate edit and get involved, than that person read an incorrect version that is flagged as "approved" or "published", try to correct the version, only to have to wait for someone to "approve" the correction. Most of the actual core content of Wikipedia is created by anon editors, the logged in experienced Wikipedians mostly write guidelines, format and otherwise do meta-wiki activities. This system would not help with credibility but simply give non-article creating Wikipedians some other meta-Wiki activity to do that would not boost the core content. Let's focus on creating and editing content not on meta-flagging. The more time we spend on writing articles and the less time we spend on thinking up and debating ideas like this the better off Wikipedia will become. SilkTork *^YES! 22:42, 29 September 2008 (UTC)
13. Absolute oppose NOT turning on Flagged revisions. We're not here to repeat a PR mantra like cuckolded zombies serving in an army from Wales, storming the gates of some e-libertarian Nirvana. We're here to write an encyclopedia that is accurate, reliable, and not harmful to living people. Not enabling flagged revisions in my opinion actually interferes at this point with the writing of this encyclopedia in the service of some ideal that is secondary in value. Will it "hurt" for a while? Sure, maybe. So did 3RR and blocking when they were first introduced, but they eventually helped get rid of other impediments to making an encyclopedia. Turn it on for the good of Wikipedia, and to reiterate, I strongly oppose this option of not turning it on. rootology (C)(T) 13:31, 2 October 2008 (UTC)
14. I've seen them in action on the German WP site, and find them annoying at best. Lugnuts (talk) 12:08, 4 October 2008 (UTC)
15. From what I understand, if we enable flagged revisions, we might as well just go and say good-bye to IP editors. Despite what some think, the vast majority of IPs and accounts are here to do good, and these people not being able to see their edits when they make them will be a net negative rather than a positive: we'll get less article-writing and more people driven-away. In addition, I strongly dislike the idea of having to review and/or flag a page every time one is edited: we'll have a constant backlog of unflagged pages, and if we spend our time flagging pages, very little article-building will get done. I prefer blocks and protections over this. Acalamari 19:24, 4 October 2008 (UTC)
16. Flagged revisions would do much more harm than good. It's just not feasible on a project as large as this. Kolindigo (talk) 15:36, 6 October 2008 (UTC)
17. With a wiki of 2,500,000+ articles, and the smallest fraction of regular editors, keeping up-to-date with flagged revisions across enwiki will be extremely difficult. I'm worried this will prove more of a hindrance than a help to this wiki, mainly because of its immense size. PeterSymonds (talk) 20:22, 8 October 2008 (UTC)

Discussion (e)

I can't help but agree with the comment of penubag above, it will create a massive inequality of treatment between IPs or new users, and surveyors. Were it restricted to articles with a real need for flagged revisions, the good would outweighs the bad. But enable this on all articles ? We won't even be able to review one edit on ten, this will definitely mean the end of free editing. Cenarium ^Talk 02:18, 22 September 2008 (UTC)

However, it is not a reason for me to oppose entirely: if we keep the number of sighted pages under control, we'll be able to review an edit by a non-surveyor in a matter of minutes. The general line is: sight if it's not vandalism or a major violation of our principles. If a user un-sight edits for WP:ILIKEIT reasons or whatever, the rights will be removed. If a user tends to WP:OWN an article, same will happen. And least but not least, if an article attracts a lot of inappropriate sightings, I think we should set an higher level of flagging, 'confirmed', restricted to some users such that IPs view the latest confirmed version instead. This would be equivalent to restrict sighting in terms of effect, and would be useful in disputes and contentious articles where the sighting system alone could be easily abused. Cenarium ^Talk 17:05, 24 September 2008 (UTC)

As a note, I've read Jimbo's admonition about flagged revisions and I don't buy it. there is a function of large organizations called Path dependence. Once you get started down one track, decisions made between following versions of that track and other completely different options are no longer made in a vacuum. This is the last time the community will get a fair shake on discussing the idea of flagged revisions and we should discuss it completely. While we have the "promise" that flagged revisions will only be used on currently semi-protected articles (where they would obviously represent an improvement), remember that Social Security numbers were originally never going to be used for anything but SSA benefit allotment. Once the tool is in place, it becomes easy to use it. Articles with less than persistent vandalism will be "flagged" with less review and control than articles currently semi-protected and we will devolve that function further and further out in the community. Right now we have articles which need to be semi-protected and we semi-protect them. That makes for a cumbersome situation. It should be cumbersome. We should still have to go through a song and dance to get articles semi'd. Openness is very important to this project and we should reject back-door methods to reduce that openness. Protonk (talk) 04:06, 22 September 2008 (UTC)

• [I apologize for the tone; I rarely get on a high-horse these days. I don't disrespect the opposition, at all; I just feel that this is one of those rare times that we're dealing with an absolute, right-and-wrong issue.] There are only a few statements in all of wiki-dom that are more important than all the others, even more important than policy: the mission statements of the WikiMedia Foundation. On the English homepage (http://wikimediafoundation.org/wiki/Home), the first two sentences read: "Imagine a world in which every single human being can freely share in the sum of all knowledge. That's our commitment." So, for you naysayers, either you disagree with the mission statement, or else you have a plan for how you're going to arrive at stable versions that can be distributed on DVD (such as WP:V0.7) without anyone ever having to specify a preferred version of articles. Which is it? (I understand that opposition to this particular proposal doesn't necessarily imply opposition to stable versions, but the arguments I see do imply that opposition.) For instance, One Laptop per Child is using our stable versions to distribute versions of Wikipedia all over the world; are you really in favor of distributing the page versions with the donkey dongs? Or are you offering to vet all the page versions yourself, once a year? If not, then we need some way to make it easier for the entire community to select appropriate page versions. - Dan Dank55 (send/receive) 22:19, 22 September 2008 (UTC)

• This is fair criticism. The burden of producing 0.7 is heroic and I don't even pretend to be a large part of carrying it. I don't agree with the dilemma proposed here but I'll provide a counterargument regardless. One of the WMF statements is that this project represent the sum of human knowledge, freely shared. Peruse the set of the rest and we find several regarding the open nature of the wiki as being important. I don't introduce that to exchange one sweeping principle for another but to insist that where we have conflicts between valid interpretations of core principles we must come to a compromise or we must prioritize between the two. In this case I feel that the options presented by the introduction of flagged revisions and the eventual outcome of the any option represent a major blow to open editing and immediacy of response. Maybe not today, where we might implement flagged revisions on only 0.7 selected articles or only on previously semi-protected articles. But once we introduce the proposal, our ability to contain its spread is limited. We have little ammunition to argue against a slow spread of flagged revision articles or a quick extension of flagging to certain classes of articles (BLP's, etc.). Once we have the policy established, arguing against the encroachment of editing rights will be an uphill battle. This discussion, right here, right now, represents the fairest shake opponents of flagged revisions will have to convince the community that the eventual result will be unpleasant. I don't pretend that it will be disastrous. At no point will we cross a threshold and see the encyclopedia radically change. Instead I think that we will have our view move steadily inward, with the core group of editors determining not just presentation but content. Returns to participation will drop and subsequently participation will as well.
• If I compare that drop in participation and change in distribution of participation to the alternatives: shipping a release version with blatant vandalism or wasting person-hours to recheck each release, I don't come to the conclusion that the burden we undertake is enough to outweigh the potential loss. First, we have to accept that flagging revisions will not eliminate vandalism above the schoolboy level any better than Huggle/RCP does now. More sophisticated vandalism will just involve flagging as reviewed a "tainted" revision. Current "your mom is gay" style vandalism will be reverted but (as is the advantage of the flagged revisions) will never make it to mainspace. The likelihood (in my opinion) that a release version of wikipedia contains one or more cases of blatant vandalism will remain the same: 1.00. This doesn't mean that we should throw our hands up. It means that we should be cautious in proposing and implementing policies based largely on a need to combat vandalism as we are in to the region of strong diminishing returns to investment.
• I don't feel that my rejection of flagged revisions is permanent, nor is it an invitation to anarchy. I support (for example), the proposed Wikipedia:Abuse filter which will stop vandalism before it is input based on an algorithm similar (basically) to the one Cluebot uses to revert it. I am also not strongly opposed to the introduction of Wikipedia:Flagged revisions/Quality versions on articles identified by the release team, so long as the community acts to ensure that process creep is limited in the extreme. The policy should be written in the outset to strictly delineate the use of flagged revisions and ensure that the scope does not expand as we (the long time editors) become comfortable with the feature. A danger still exists in declaring quality revisions acceptable for printed release and then allowing that definition to expand. If and only if we ensured flagging was a background process (unimportant and unseen to viewers of the page), reactive to criticism, and revocable by the community should we contemplate enabling them. Even if the flagging is limited to release subjects, we don't want to get into a situation where dalliances in article scope, structure and content would be entertained by the "in group" of editors only because they knew that the changes would never make it out to the "real world". Protonk (talk) 02:31, 23 September 2008 (UTC)

I believe you're saying that some form of "quality versions" might be okay with you; I'd be open to seeing the proposal for what you'd find acceptable. If you're saying it's dangerous to experiment to see what happens, I don't agree. I do agree with you that once we implement flagged revisions on say 30000 articles, it's likely that further changes would be hard to stop, unless we get firm consensus and understanding ahead of time that expansion would be bad. In my innocence, I believe such negotiation is possible. - Dan Dank55 (send/receive) 03:21, 23 September 2008 (UTC)

I think we agree, albeit narrowly. I'm going to take another look at "Quality Revisions", which is the most acceptable version and see where it could be improved. I don't so much think that it is dangerous to engage in an experiment (that is, endorse some small version on a contingent basis), but I don't think it is necessary. Protonk (talk) 03:58, 23 September 2008 (UTC)

Switched my vote to (c) per the concerns of the opposition; see above. - Dan Dank55 (send/receive) 14:50, 23 September 2008 (UTC)

"... [W]e have upwards of 150 edits per minute 24 hours a day, seven days a week, 365 days a year. There is no way that we can possibly keep up with that kind of volume." In that case, I respectfully ask: are we currently keeping up with it? If not, then Wikipedia is too large; if so, then we can cope with FlaggedRevs. Unless my logic is faulty. – Thomas H. Larsen 01:50, 25 September 2008 (UTC)

That's right, we get that many edits, mostly from IPs. Now, how in the world are we going to review them all? This would be the end of free editing. -- penubag  (talk) 05:04, 25 September 2008 (UTC)

That fact that we can keep up with this volume of editing is obvious from the fact that most vandalism is reverted in 5 minutes. If our antivandalism people can watch and filter this volume of editing, why would it be impossible for them to use flagging as part of this effort? Tim Vickers (talk) 16:44, 25 September 2008 (UTC)

The only reason why we somewhat can keep up with the vandalism is because over 80% of it isn't vandalism; now try our reverting bulk by 80+% (besides, no scripts like Twinkle written yet). And anons mostly edit in good faith anyways. Spoiling it for everyone (even for you) just because a few (<20%) ruin it is uncalled for, besides we already have the vandalism covered (even you said yourself, "most vandalism is reverted in 5 minutes"). Also, please read my reason for opposing; I, like you, was highly in support until I thought about it for a while. This proposal will greatly hinder further progress to this encyclopedia because it will not only slow the edits of anons but will stray away potential long-time contributors (as I stated in my opposition reason). Please try to be a little openminded and think a little more :) Thanks. -- penubag  (talk) 01:10, 26 September 2008 (UTC)

If something is reverted as vandalism, then someone saw it, and they can flag the revision after they revert, if necessary. If something isn't vandalism, obviously someone had to review it, else we wouldn't actually know that it wasn't vandalism (Schrödinger's edit?), in which case, clicking a button to flag it wouldn't add a huge burden. Presumably any edits that aren't reviewed are those made by users who would likely have the right to flag edits themselves. If flagging is built into anti-vandal scripts and programs, it would be even easier. Mr.Z-man 16:26, 26 September 2008 (UTC)

Maybe it would be even easier, but the difference beeing that in the current situation the page is semi protected, so everybody looks at the same page, with flagged revision, something else happens, after the flagging, edits are made, but they don't show up, I also think that the 80 % of the good IP editors will stop editing if there edit might, or might not show up, where on the other side proposals are in the making to make it more attractive to edit on WP (besides of this idea will make it or not) another aspect we have to confront is trust from the editors, similar to commercial advertising where you is a visitor see a different advert then another visitor, the same happens at the moment with internet content, as for the edits, already there are websites where you make an edit, and yes the edit shows up in the page, but funny ALL other visitors of the page see the page without that edit and lastly i would like to wait with this straw poll until full statistics are back online. Mion (talk) 17:10, 26 September 2008 (UTC)

g) I don't mind

Vote (g)

1. I'm going to comment here for now, as I support any and all of options A,B,C and D. I strongly support introducing flagged revisions in some form but doubt consensus will be reached for option A so just wanted to say that I will support whichever option people can agree upon. Davewild (talk) 07:57, 21 September 2008 (UTC)
2. Concur with Davewild; I have a vague "order of preference" but A, B, C and D (slightly favouring D) all seem workable. CIreland (talk) 14:35, 21 September 2008 (UTC)
3. Good point, it is more important to get it enabled so we can see how it works in practice than worry over exactly how this is done. Tim Vickers (talk) 01:24, 22 September 2008 (UTC)

Assignment of surveyors rights

By request

I propose that the surveyor rights (users able to sight revisions) be assigned by administrators in a similar way to rollback, requested at WP:RFR, and that rollbackers be automatically assigned to the surveyor group. The users must be autoconfirmed and satisfy certain requirements; essentially no vandalism, no recent breach of WP:BLP and other core policies. This doesn't preclude an additional automatic assignment. Cenarium ^Talk 17:04, 23 September 2008 (UTC)

Automatic assignment

It is possible to assign it automatically, the option 1 is from Wikipedia:Flagged_revisions/Sighted_versions. When the right has been removed, it is not reassigned automatically. Cenarium ^Talk 17:04, 23 September 2008 (UTC)

Option 1

• an account for 30 days
• 150 edits
  • 30 edits to articles
  • 10 articles edited
  • 15 days of edits
• a confirmed e-mail account.

Vote

Option 2

• an account for 60 days
• 300 edits
  • 100 edits to articles
  • 20 articles edited
  • 30 days of edits
• No block (TBD)
• a confirmed e-mail account.

Vote

1. Sounds about right. Assumption is that by request becomes an option once a prodigal has learnt the error of their ways and kept their nose clean for a suitable period--ClubOranje^Talk 10:04, 30 September 2008 (UTC)
2. First choice By these edit numbers, all ill-intentioned people will be blocked. Pie is good ^{(Apple is the best)} 22:48, 2 October 2008 (UTC)

Option 3

• an account for 120 days
• 700 edits
  • 300 edits to articles
  • 40 articles edited
  • 60 days of edits
• No block (TBD)
• a confirmed e-mail account.

Vote

1. I would prefer this for a beginning. The requirements are considerably higher, but I think it is needed to make sure we don't make it too easy for vandals and sockpuppets to access surveyor rights. It would include most regulars and if the need is greater, we can lower the conditions. Cenarium ^Talk 17:54, 2 October 2008 (UTC)
2. If, and only if that these criteria are only automatic assignments and that any ex-vandal that was blocked can apply as well if they reformed. Second choice Pie is good ^{(Apple is the best)} 22:47, 2 October 2008 (UTC)
3. I support automatic assignment when this level of contributions is reached, plus "rollback-style" assignment upon request, at the discretion of the granting admin. In a similar fashion, even automatically granted rights can be withdrawn by an admin, open to appeal to the wider community. Random89 17:02, 6 October 2008 (UTC)

Oppose to any form of automatic assignment

1. Being a surveyor would probably put a person in a delicate position, having the powers to fix a plenty of things and having the ability to mess up big. Rollback is probably not a big deal compared to this, and it's still handed out on request only. Admiral Norton ^(talk) 18:54, 2 October 2008 (UTC)

Discussion (assignment)

• Both options seem particularily good, except for the No block thing. Make it something like within the last two months with no behavior breaching policy or something. This criteria makes it seem like that a previous vandal can't be turned around. Pie is good ^{(Apple is the best)} 19:37, 29 September 2008 (UTC)

  These criteria are for automatic assignment, they are not binding for assignment on request. Cenarium ^Talk 17:41, 2 October 2008 (UTC)

Discussion

Difficulties with a single group of sighters

Suppose we have one user group able to sight revisions for all articles with sighted revisions enabled. Then if we want to generalize sighting to a large class of articles, the group will have to be bigger and bigger. It will create several serious problems:

1. Sight wars will frequently occur on contentious articles.
2. Vandals will be granted the surveyor rights more easily and able to complete some sensitive actions.
3. On disputed articles, this will create an inequality between surveyors who can sight their edits and non-surveyors.

What I propose is the creation of a user subgroup of surveyors, and the possibility for admins to restrict sighting of an article to this subgroup. Some other sensitive action will be restricted to this subgroup. I'd like to have your opinion on this, the detailed proposal is here. Cenarium ^Talk 15:04, 20 September 2008 (UTC)

Additionally, this system will allow to moderate disputes: to summarize, neutral "moderators" will sight an edit only if it consensual. Cenarium ^Talk 18:31, 20 September 2008 (UTC)

Update A better alternative to 'sight protection' is to set up a higher level of flagging, confirmed revisions, and create a user group able to confirm revisions. An article can be forced to display the latest confirmed revision. (This is similar to existing functions.) Cenarium ^Talk 14:55, 3 October 2008 (UTC)

And when the surveyor group becomes large, surveyor wars might ensue, so we'll need a group of… This is becoming to look like a ranking system. Admiral Norton ^(talk) 18:32, 20 September 2008 (UTC)

It's no different that autoconfirmed users, rollbackers, and administrators. Cenarium ^Talk 18:39, 20 September 2008 (UTC)

For 1) above, I would assume WP:3RR and any other editing/reverting rule would apply to sighting as well, for 2) not much worse than it is now but it will happen much more rarely, there will be sockpuppet sighters that sight the edits from other socks and so on, what is new? we should not object to good things just because they are not perfect. 3) same as admins who can protect and semi protech their edits now, all must be handled the same way as we handle these issues now. I would assume that you would not be allowed to sight a edit to a contentious article that you made yourself, same as a admin can not protect a page where he himself is a contributing editor or block someone that he himself is in dispute with, all this must be solved, but it will not be any serious problem. --Stefan ^talk 22:09, 20 September 2008 (UTC)

We have circa 2.5M articles, if we don't want huge backlogs for sighting, we'll need to have many, many sighters. Of them, a considerable part will be unreliable for disputes, if we can't protect from sighting, we'll need to fully protect articles in order to stop sighting wars or block people. So if we can't restrict sighting, we won't be able to improve our handling of disputes. But if we can restrict sighting to "moderators", we'll be able to control disputes. And it will solve the above problems much more efficiently. I don't object, but without restricting, I'm pretty sure that it will be unworkable on en. Cenarium ^Talk 00:00, 21 September 2008 (UTC)

Well, in my opinion sighting wars should seen as wheel wars, and result in immediate loss of privileges. There's absolutely no reason why antisocial behaviour by experienced community members should be tolerated at all. – Thomas H. Larsen 05:11, 22 September 2008 (UTC)

I agree. Take it away from them if they mess around with it. Quick and easy. Pie is good ^{(Apple is the best)} 19:39, 29 September 2008 (UTC)

It's not always the best solution. There is a lot of prolific editors who are regularly involved in disputes. Removing their surveyor rights will cause a rise of edits to sight (they won't be automatically sighted anymore). But we should keep the sighting delay as minimal as possible, I think it should be below one minute. If we enable confirmed revisions (temporarily) on a disputed article, this will both resolve this problem and avoid full protection. Consensual or non-controversial edits will be confirmed within minutes. We'll be able to relatively control disputes in an effective manner. It may also be used as an arbcom enforcement, for articles under probation, very high risk blps, etc. It also addresses the point 3, since edits are never automatically confirmed. Cenarium ^Talk 14:55, 3 October 2008 (UTC)

How does this impact anonymous editors?

If non-logged-in users cannot view the latest revision, then how do they know if a particular error/omission has not already been fixed in the current (hidden) revision? Does this mean that everyone will have to log in to edit? « D. Trebbien (talk) 18:18, 20 September 2008 (UTC)

Also, anonymous users won't be able to see their edit until it's flagged and that could mean for days or more. Furthermore, I can't think of a reasonable system to fix this. Admiral Norton ^(talk) 18:35, 20 September 2008 (UTC)

This is why I'm still opposed to mass sighting for now, we won't have enough surveyors to check the edits, we would be flooded. But on certain articles, it would be very helpful. Cenarium ^Talk 18:38, 20 September 2008 (UTC)

I'm sure it is possible to give IP editors the ability to see all version, but by default they will see the last sighted, but they will be abel to se the history and from there see any version (I have not played with this on de). All this should be possibel to describe in the edit window for a IP editor when he edits. We also need to change some text somewhere on all pages to tell users that this is a 'sighted' version and that if they want to see the latest version they need to look in the history or maybe we can add a new tab for it or something. Not sure what possibilities the current code base gives us. --Stefan ^talk 21:49, 20 September 2008 (UTC)

How many IPs know about page history? We're not talking about experienced editors here. As for the tab suggestion, lemme tell you my experience about it: while I was an IP it took me weeks to summon courage and find out what are these tabs really for. Admiral Norton ^(talk) 22:41, 21 September 2008 (UTC)

If it is decided to show by default the sighted version, then i would assume we could place a link or tab to show the most recent version. This tab would most likely be ignored by passive readers, but would allow IP editors to view the more dynamic content. Random89 17:06, 6 October 2008 (UTC)

It completely disenfrancise them.Geni 02:41, 22 September 2008 (UTC)

A slightly different proposal

I have a proposal for enabling flagged revisions that differs from all of the options above. I think that flagged revisions should be enabled for all articles. However, the default version shown should be the latest revision, regardless of whether an article has been flagged. Administrators should then be able to set articles to display the flagged revision by default either for anons only or for everyone, similar to how semiprotection and protection work now. That would allow high-vandalism articles or particularly controversial articles to benefit from flagged revisions without significantly reducing Wikipedia's openness as a whole. Articles showing the latest revision first could still be flagged; it just wouldn't change what revision is displayed. Having articles show the flagged revision first could be used more widely than (semi)protection, but doing this on all articles would do far more harm than good in the long run. Pyrospirit (talk · contribs) 13:25, 21 September 2008 (UTC)

And what is the real benefit here? It just generates more work for users and diminishes the importance of the actual result. Using flagged revs as semi-protection would be better, although I don't think I'd like that either, as much as I love to play with buttons and menu bars. Admiral Norton ^(talk) 22:41, 21 September 2008 (UTC)

Article WP:OWNers and other statusquoists

I can't think of a realistic solution for this. Admiral Norton ^(talk) 22:41, 21 September 2008 (UTC)

• Um, how about a delete and salt whenever this happens? Pie is good ^{(Apple is the best)} 19:40, 29 September 2008 (UTC)
  • Oops, thinking about the wrong discussion. Anyways, if they mess around with it, just take away the privilege. Pie is good ^{(Apple is the best)} 19:41, 29 September 2008 (UTC)

Still another proposal: only in quality articles

I have another proposal: flagged revisions shall be implemented in articles that have already reached a certain level of quality. That way anons can still edit articles most in need for improvements, but articles having reached a sufficient level of quality would be protected from vandalism and more stable. What the certain level of quality should be (FA, GA, or something else) I don't know. SyG (talk) 18:47, 22 September 2008 (UTC)

I'd suggest this was a type of Flagged revisions should be enabled only over some articles, determined on a categorial basis Tim Vickers (talk) 20:26, 22 September 2008 (UTC)

I'm switching my vote to this; see (c) above. - Dan Dank55 (send/receive) 14:51, 23 September 2008 (UTC)

Non-articles

If we are to enable sighted revisions on all articles, we should also consider them for non articles visible to readers:

• All the template space (with the possibility to remove some templates, like test templates)
• All the portal space (only a few edits from IPs and newcomers, so won't add much to the workload, but vandalism happens sometimes)
• Certain high visibility pages from other namespaces (like Wikipedia:About, Help:Contents). Cenarium ^Talk 14:05, 23 September 2008 (UTC)

Message to editors

It should be made clear to all editors when a page has sighted revisions enabled, in edit mode. Otherwise, newbies won't know why their edits are not visible. For example, using the editnotice, or a built-in message. Something like:

This will have to be designed after implementation though. Cenarium ^Talk 15:34, 23 September 2008 (UTC)

Missing an option

What about "only enable flaggedrevs where semi-protection is currently used"? Aaron Schulz 22:00, 23 September 2008 (UTC)

I don't mind throwing those into the mix, but if those are the only pages we use, then that won't generate any data on how the feature works on non-contentious pages. - Dan Dank55 (send/receive) 22:08, 23 September 2008 (UTC)

Using them in place of semi-protection is the only way I'd be in favor of flagged revisions. I see them as being superior to semi-protection, but inferior to normal editorial review on articles that don't need that protection. I am sure that flagged revisions would have a lower threshold than semi-protection since it's less obstructive to editing. —Elipongo (Talk contribs) 10:06, 24 September 2008 (UTC)

This would give people an incentive for contentious editing on their favorite pages to provoke page protection so that they can get some form of flagged revisions. - Dan Dank55 (send/receive) 12:10, 25 September 2008 (UTC)

Semi-protection is usually put in place to stem IP vandalism, I think flagged revision would work quite well for that. Edit warring would earn full page protection, just as it does now. —Elipongo (Talk contribs) 12:25, 25 September 2008 (UTC)

I agree with you, as long as the flagged revision feature is disabled as fast as semi-protection would ordinarily be removed; otherwise, if someone wants flagged revision, there's an incentive to raise a fuss (unless, of course, flagged revision applies to all articles). - Dan Dank55 (send/receive) 12:40, 25 September 2008 (UTC)

• I'm not convinced that flagged revisions can replace effectively semi-protection on most articles. It may work in complement to sp on articles subject to heavy vandalism, or for certain other articles. But on regularly vandalized pages, it won't prevent IPs or new accounts from editing, and we'll have to revert a lot more often than now. On disputed articles, it won't work if surveyors are involved, so it won't improve our control of disputes. Unless... we set up a higher level of flagged revisions for certain users, charged to deal with changes to disputed articles. Cenarium ^Talk 17:52, 25 September 2008 (UTC)

It was my impression that semi-protection is used to prevent readers from seeing the vandalism, not to save us work. Flagged revisions would keep readers from having to see "poop" or that Nathan Hale lived in Timbuktu, while still allowing legitimate IP editors to work. Why shouldn't an editor from an IP be able to make a legitimate edit to the George W. Bush article just because a lot of unregistered editors like to mess with it. —Elipongo (Talk contribs) 20:54, 25 September 2008 (UTC)

If we give ourselves more work inconsiderably, then we won't have time for other needed tasks. Replace SP with SR on George Bush, and it becomes a total sandbox. We'll revert hundred times a day there and we'll miss hundreds of bad edits to other pages without SP or SR. Cenarium ^Talk 12:59, 3 October 2008 (UTC)

But once again, the purpose of the indef-semi on these articles is not to save us work, but to not show vandalism to the reader. If a certain page receives an overwhelming amount of vandalism, we can still semi it as well as having it sighted. Random89 17:12, 6 October 2008 (UTC)

Opening of indef-fully protected templates

I see many people on this page claiming that sighting equates to a closing of wikipedia or somehow making it less open and free. When i think of flagged revisions, I see the potential to open to editors areas which have for years been available only to admins. For example, several main-page templates could conceivably be opened to general editors to edit and update, such as ITN and DYK, areas in which updates to the visible template often lag behind the consensus and deadline to do so. The implementation of flagged revisions is the very embodiment of "wiki" principles, not the antithesis of such. Random89 17:22, 6 October 2008 (UTC)

Indeed: instead of this now culture, where "edits must appear now, because I want them to appear now", make a reliable encyclopedia. – Thomas H. Larsen 07:42, 9 October 2008 (UTC)

---

The above discussion is preserved as an archive. Please do not modify it. Subsequent comments should be made on the appropriate discussion page, such as the current discussion page. No further edits should be made to this discussion.