User talk:Jimbo Wales

Question about flagged revs

76156 So, we're turning on flagged revs with a ~60% consensus? Don't you think this is a classic example of "no consensus"? I see no agreement, no discussion coming to a largely resolved opinion, and no realistic chance of anything changing in the short term. I'm really confused where the momentum you sense to turn this on for the English Wikipedia is coming from. Matt Yeager ♫ (Talk?) 04:01, 17 January 2009 (UTC)

It was actually a straw poll, so I guess consensus doesn't matter too much here. Chamal ^talk 04:08, 17 January 2009 (UTC)

Normally, foundation (small-f) changes require consensus in the region of 70% before they're even considered. Jimbo: please reconsider. There is literally no consensus within the community to even try it out. And please don't make this into our Treaty of Lisbon, where it keeps coming up every time we say "Non." Sceptre ^(talk) 18:25, 17 January 2009 (UTC)

? Issues this big hardly ever come up, but, on a smaller scale, we got rid of spoiler warnings after (actually during) a straw poll which gave removal 58% support (after a previous poll gave the "wrong" answer). PaddyLeahy (talk) 19:10, 17 January 2009 (UTC)

FR needs to be tried. I personally doubt it will scale well, I think the same vandalism and disinformation problems will occur, only more sneaky (and locking in previous inaccuracy). But it's there, it needs to be tried and either accepted or rejected based on experience. Franamax (talk) 14:00, 17 January 2009 (UTC)

Obviously, a large minority of the community feels differently. I don't know if this is the best place to argue your POV too.

Also, Jimbo, may I ask for another major foundational shift that occured with consensus of barely over half the community? Considering that this isn't some silly thing about rollback, but something that might make people quit the project (~20% of German regulars quit after FR was implement, if I remember correctly), don't you think it is more prudent to repoll in a few more months and see if a better community consensus can be found?in a year and see if community consensus is different. NuclearWarfare _(Talk) 19:22, 17 January 2009 (UTC)

Nuke, that's the wrong way to go about it. This is not gonna become another one of those ideas which keeps getting voted on until whoever started it gets the "right" answer. My mother taught me when I was a kid, "No means no." Once is enough. Thor Malmjursson (talk) 19:35, 17 January 2009 (UTC)

Also, "X bad thing happened, if I remember correctly" doesn't really help me much to explore the issue. Do you have a reference?--Jimbo Wales (talk) 19:43, 17 January 2009 (UTC)

Let me strike that part. The German IRC channel informed me that it was a lower percentage than that, though they don't have active figures. However, User:Dapete/Report_on_Flagged_Revisions,_December_14,_2008#Influence_on_edits_and_user_registrations, a translated version of a mailing list post, suggests that in the past year, registration has gone sharply down, by as much as 50%[1]. NuclearWarfare _(Talk) 20:07, 17 January 2009 (UTC)

Are there any before/after numbers to suggest this had anything to do with Flagged Revisions? Or was the decline before this was implemented? Remember, too, the Germans have decided to use Flagged Rev's *everywhere* *by default*, which is not even under consideration here.--Jimbo Wales (talk) 01:41, 18 January 2009 (UTC)

As an aside, I consider our BLP issue to be so important that I think it is actually unethical to not use a tool which holds great promise for helping with the problem, now that it has been successfully tested elsewhere. Anyone who would like to see this tool not go into practice needs to start by convincing people that either (a) it is ok for the BLP vandalism problem to continue or (b) there is a better way to solve it. Anything else, for me, is just a total non-starter.--Jimbo Wales (talk) 01:44, 18 January 2009 (UTC)

Category:Living people currently has 333,547 articles. That's 12.34% of our articles (about one out of ten). Enabling FlaggedRevisions on such a large segment of the site will undoubtedly create backlogs that we will be unable to manage. I propose that we enable it for articles that have demonstrable issues, but we absolutely should not do anything site-wide or across a huge segment of our articles.

And having spoken to a number of people about this issue, a lot of them tell me that the places where they see the worst BLP violations are in articles that are not in Category:Living people. Enabling FlaggedRevisions on all BLPs is a poor solution to the problem that will do far more harm than good. --MZMcBride (talk) 02:01, 18 January 2009 (UTC)

Nonsense. "Undoubtedly" we should remember that we're not prognosticators here. I don't think you can pretend to know what would happen, but I will make a wild prediction that the sky won't fall. German Wikipedia is flagging over 800,000 articles with a much smaller contributor base. This proposal deserves a trial, not FUD. Cool Hand Luke 07:20, 18 January 2009 (UTC)

Excuse me? It may come as a surprise to you, but this isn't my first time on the merry-go-round. We currently have new page patrolling on this project. After 30 days, the data expires and the pages can no longer be patrolled. It has been backlogged literally since its inception. One admin has been working nearly daily to keep the data from falling off the end, and even then it requires a bot to keep up with the high volume. And this is just new pages (which are restricted to logged-in users).

If we were to implement FlaggedRevisions for all edits, it would create backlogs. This is a certainty, and I know this not because I'm a seer, but because that's how things work on this site. I've spoken with a number of different people about this. At one point I even tried to get FlaggedRevs implemented through sysadmin fiat. But I've come to realization that doing it site-wide simply will not work for this project. It's time that others caught up to this reality. --MZMcBride (talk) 07:32, 18 January 2009 (UTC)

Wether or not enabling FlaggedRevisions will create a backlog is solely a function of the number of authorized approvers and the number of articles affected. Since the first factor could be easily brought to tens of thousands (by selecting everyone with an active history that doesn't include vandalism and with edits that last above a certain number of article revisions) and that would enable a very dynamic system with better quality assurances than we have today, I don't see a need to stay in the extremes (no moderation/moderation by a tiny minority) when it would lead to worse results.Herbys (talk) 02:16, 26 January 2009 (UTC)

Addendum: One last thing. All of this talk about "fixing" BLPs. Guess what? Flagging the content doesn't do anything. In fact, it nearly does harm when content is flagged as "accurate" or "verified" when it actually isn't. You want to fix the BLP problem? Go to Wikipedia:Database reports/Biographies of living persons containing unsourced statements and start editing. That's the first 500 entries in a list that's over 17,000 entries long. Until people like you start helping out and fixing the actual problem rather than putting cute little flags on the content for somebody else to fix, nothing will change. --MZMcBride (talk) 07:39, 18 January 2009 (UTC)

You're excused. The "first time" was an absolutely pointless venture; there's no compelling reason to update revisions since it made no difference to the displayed content one way or another. Different story with flagged revisions on a minority of our articles; we have strong ethical and technological motivations to flag them because we care about the content on living people and the edits won't show up otherwise. We've got more users, and we're proposing to flag less articles than de.wp, which is 98.7% flagged up-to-date; it deserves a trial. Worst case scenario: you're right and we drop it. Cool Hand Luke 07:40, 18 January 2009 (UTC)

::::::I'm sorry if this offends you, Jimbo, but it seems to me that you are hell bent on getting this mutant offspring of a wiki extension implemented here even if its the last thing you do. When are you gonna start to listen to the users? I have spoken to Administrators who think you are making the biggest mistake of your life. Over 40% of those polled have laid some exceptionally strong arguments, but you seem intent on walking on those who have opposed this and going ahead with it anyhow. A clear cut use of The Parliament Act, if I may say so. Go ahead and implement it if you must, but I for one will accept it if you agree to take it out of use should it prove to be, in your own words, "a non starter". Thor Malmjursson (talk) 02:20, 18 January 2009 (UTC) Struck out - Utter nonsensical ravings and gobbledegook. Reworded below :) Thor Malmjursson (talk) 04:46, 18 January 2009 (UTC)

How could I be offended? It's as if you've accused me of being a Purple Martian. What you are saying about me has so little resemblance to the actual facts about me, that I can't be offended, but rather bewildered. Do you want to try again to take a more nuanced approach?--Jimbo Wales (talk) 03:51, 18 January 2009 (UTC)

(Outdented) - Sorry about that, Jimbo. I hate confusing people, but sometimes things in my brain don't reach the keyboard in the right order; a mild case of PEBCAK if you will. What I was trying to say is, it looks like you are going to force this through no matter what comes, and that you are not listening to your editors properly.

The reference to the Parliament act comes from a British law about something known as a "suspensory veto"; if a law goes before the House of Lords and gets rejected, the Prime Minister can invoke the Parliament Act, Part I, to force it through against their wishes. This is what your attempt to bring FlaggedRevs in looks like from my point of view.

The matter of FlaggedRevs has been discussed at length on the English IRC channel, where I have spoken to several administrators who think the implementation of it here is wholly wrong. It is, in the words of one I spoke to, whom I will not identify, "an abortion waiting to happen." I wouldn't go that far, but I disagree with it in principal.

I would however, accept Flagged Revisions if the condition was added that in the event of it being a failure or causing a lot more problems than it solved, you would agree to remove it from this Wikipedia.

I hope this is a little clearer than my last ranty attempt to say what I was thinking. Thanks for understanding. Thor Malmjursson (talk) 04:46, 18 January 2009 (UTC)

Thanks for clarifying. On the other hand, in this particular case, Parliament seems to have passed it - by a wide margin. And yes, of course, a trial period is meant as just that: a trial period.--Jimbo Wales (talk) 12:43, 18 January 2009 (UTC) Addendum: I have thought through the Parliament Act analogy several times, and I still don't get what you're saying. There is no "House of Lords" which has rejected it, and I'm not forcing anything through against the wishes of the community. We have a straw poll which shows that there is support by a wide margin, and I'm suggesting that it is about time to go through with it. Saying that I'm not listening to users doesn't make sense to me either... I'm not sure how me going against years of discussion and a community poll would be "listening to users" more than what I'm doing.--Jimbo Wales (talk) 12:48, 18 January 2009 (UTC)

There is no "wide" margin. At the time of writing, there is exactly sixty percent support for a trial. If just one more person opposed it, you'd have a majority on which the US Senate couldn't invoke cloture, and that's on the low end of supermajorities. Sceptre ^(talk) 16:49, 18 January 2009 (UTC)

10% seems pretty wide to me. It's enough that the poll is unlikely to have got the wrong result through sampling errors. --Tango (talk) 18:39, 18 January 2009 (UTC)

We don't make decisions by majority opinion. Cheers, — Jake Wartenberg 18:54, 18 January 2009 (UTC)

Try and find one significant proposal that passed with 60% support or less. No RfA has ever passed with that percentage, and you can count on one hand the number that came below 70%. The straw poll does not show consensus for turning on flagged revisions. Hut 8.5 19:15, 18 January 2009 (UTC)

The worst that can happen is we turn it back off. What is the big issue? Prodego ^talk 05:05, 18 January 2009 (UTC)

The big issue is that while we have majority support, we don't have consensus--and that's the way we have always made our decisions. I think that everyone involved should stop and evaluate not whether or not you think flagged revisions are good, but if you think going ahead despite lack of consensus will be good for the community. A lot of editors are becoming disenchanted with the project; we are losing them all the time. This is not going to help, and that is my main concern right now. It should be yours too, Jimbo. The issue of the extension itself pales in comparison. Regards, — Jake Wartenberg 16:40, 18 January 2009 (UTC)

Jake makes a good point about losing good editors currently, but I think we will continue to lose editors if we don't turn on flagged revisions and don't make effort to deal with BLP problems. Using flagged revisions (and whatever other means we have for BLPs) is so essential. Here is an example [2] of how flagged revisions would have helped. I only noticed the IP edit inserting blatant BLP violations in that article, when I noticed another user reverting it six days later! I believe this edit ~1 week later [3] was made by the subject, and on this BLP talk page [4], you see that the subject of the article has previously complained to OTRS. I feel horrible that the IP edit remained for six days! It's an embarrassment to Wikipedia. Flagged revisions would certainly have helped, and is worth whatever effort we need to put into to patrol/approve the edits. The sooner flagged revisions is turned on, the better. If somehow, it doesn't work, then we can always turn it off later, but we must try it and I believe it will be helpful to have it. --Aude (talk) 16:53, 18 January 2009 (UTC)

I am also of the opinion that a trial is better than no trial at all. We don't know until we see some data, we either discover that it's not right for Wikipedia, or we miss out on the biggest opportunity this place has ever had. The BLP situation has become so extensive, that I can completely understand why Jimbo feels it is crucial to investigate every possibility in dealing with it. Of course being unopposed to a trial does not equate to being unopposed to the actual full implementation flagged revisions. Looking at the opposition, a lot of people seemed to have missed that point. --.:Alex:. 22:26, 18 January 2009 (UTC)

I have personally put much thought in to the pros and cons of flagged revisions as currently defined. One of my greatest concerns is the nebulosity around exactly what the trial will incur, and whether it is merely a trial that can be reverted or a fiat. This point in particular needs to be addressed affirmatively so that the community can at least move forward without doubt hanging in the air. While the straw poll is merely that, a "feeler" for the community's opinion, as it currently stands, there are many who oppose and consensus is not reached. A trial on a limited number of articles also cannot address the problems of scale, so I don't believe that it should be called a trial, since it would not be an effective measure of viability.

I am fully cognizant of the liability the project faces when dealing with BLP's that are less than factual, but my concern is that flagged revisions will ultimately do more harm to the project in the long term. Wikipedia's core principle is "the free encyclopedia that anyone can edit". It's on the Main Page, and has been virtually since it's inception. Flagged revisions are being used on the German Wikipedia, with mixed results, as stated by some of the administrators from that project. Flagged revisions would be effective if the care is taken to check for vandalism before sighting, but with a project the size of this Wikipedia, I have my serious doubts. A significant backlog seems almost inevitable, with one solution for German Wikipedia being the use of bots to sight articles, which obviously has potential defects as compared to an editor sighting the article. Other concerns are how it will impact the project to those who want to contribute, but will discover that Wikipedia is no longer the "instant gratification" that allows their hard work to appear immediately. There is also the question around who will be provided the necessary permissions to sight articles. I am not at all against the concept of some method to reduce liability, but liability will not miraculously vanish with flagged revisions. The sole solution to removing liability for inaccurate BLP's is by shutting down Wikipedia. I respectfully disagree with those that believe that flagged revisions is the "magic bullet" that would remove all liability.

The other concerns are more on a forward looking basis around how Wikipedia will present itself to potential and existing editors. With the implementation of flagged revisions, nobody other than those privileged with sighting flags will be able to edit the live version of an article. This in particular holds many dangers in the approachability of the project to an editor. I would certainly not wish to devote time and effort to the writing of content when there is no guarantee that my work will ultimately be shown. German Wikipedia has experienced a significant reduction in the number of edits to their project since the implementation of flagged revisions, and considering that this Wikipedia is also currently experiencing a decline in editing activity, I am concerned that this alone would ultimately be a fatal blow to the project.

To summarize, flagged revisions are an imperfect cure to a difficult problem. We already have existing tools in place that can mitigate the damage being caused that do not alienate the core group of editors that have made Wikipedia what it is. I must therefore respectfully urge that flagged revisions not be implemented in it's currently defined iteration, and that all concerns raised be weighed on their merits. Best regards. --Chasingsol^(talk) 08:21, 19 January 2009 (UTC)

"A very large majority, at least two-thirds" is needed according to Erik Möller [5]. DuncanHill (talk) 14:27, 19 January 2009 (UTC)

There clearly is NO CONSENSUS under ANY definition of the word 'consensus'.

It is not in the interests of the community to trample on the views of large and passionate minority who wish to maintain the principle that all editors have an equal right to edit, and equal responsibility for what they produce. This change will increase divisions, create new wounds and rub salt into the existing ones. It cannot be in anyone's interests for this to go ahead, and doing so will show contempt for a whole layer of people who have devoted their time and energy in good faith, believing that what WP told them: 'this is an encyclopaedia that anyone can edit' was true. —Preceding unsigned comment added by Riversider2008 (talk • contribs) 13:27, 22 January 2009 (UTC)

Entries such as Senator Edward Kennedy, and Senator Robert Byrd (whose entries were recently reportedly vandalized) and other such notables probably should be much more routinely LOCKED DOWN. This would prevent Wikipedia from becoming into the public eye and the calls for change that follow. Maybe, just maybe locking down a lot of the much more visited, or highest profile, or current entries would save us all from revisiting this matter in the future. Or at least the very near future, anywayz. Lesbrown99 (talk) —Preceding undated comment was added at 01:50, 25 January 2009 (UTC).

Meanwhile, in Germany...

Higher up in this section, NuclearWarfare posted a link claiming that registration on the German Wikipedia had dropped with 50% after Flagged Revisions were enabled. As far as I can tell, they were enabled as a test on the German Wikipedia in May, and fully enabled in August-September 2008. This Wikimedia statistics page suggests that if anything, registration has gone up since then, not that it was halved.[6] The rise is probably caused by IPs preferring to register after FR are implemented, but I just wanted to show that the figures given above in the discussion seem to be incorrect. From the same site, it also appears that the number of edits on the German Wikipedia has not significantly changed.[7]. While the first number (new editors) does not show the change FR may have on the existing base of editors, the number of edits seems to indicate that in the end, there was no apparent drop in the number of active editors. Fram (talk) 10:40, 23 January 2009 (UTC)

Has any consideration been given to versioning? Underline the new stuff but make it easy to see what has been changed. & have it signed. New stuff appears immediately but can be subject to edit. Johntnash (talk) 16:50, 26 January 2009 (UTC)

The importance of BLP's

I apologize for putting up a new section, but I had hoped to be able to converse with you normally. You mentioned above that "I consider our BLP issue to be so important that I think it is actually unethical to not use a tool which holds great promise for helping with the problem, now that it has been successfully tested elsewhere."

I have two objections with regard to this. The first and more relevant is that I strongly, strongly disagree with BLP as an important issue (per se). If there is untruthful information on some living person's BLP, we are completely, completely free from violating the libel laws of the State of Florida and the United States so long as Wikipedia doesn't officially give it a stamp of our approval. Since the legal issue is non-existent, offense is our only reason that I can see to specifically protect BLP's (with FR, etc.). People may well get offended at their own personal pages. These people honestly don't matter much. The Seigenthaler incident, in the end, did nothing to damage Wikipedia directly and only slightly lessened its credibility among the vast majority of our readers. We spend way too much time pandering to the views of the few who are "privileged" enough to have merited WP articles and far too little time working on getting all of our articles done. I very firmly believe that all our articles should be treated equally. A typical Wikipedia reader is far more likely to be seriously upset by a false article on his country, religion, or culture than on some random famous person. Do we really believe that Seigenthaler is more worth protecting than France? Which one's a more likely target for vandals? Which one will offend more people?

The other point is a strong belief that I and many others share, which is that FR is at this point (at least) the wrong way of protecting BLP's. It goes against our fundamental principles and beliefs. I don't think I need to bring up the arguments since many, many people have already done so. And that leads me to wonder: there were a lot, a lot, a lot of well-thought out objections to the FR proposal. (And a lot of well-thought out supports, of course.) You say "I consider our BLP issue to be so important"... well, you yourself may. Does that mean our attempt (and ultimately failure--I'm sure you would agree we didn't find a consensus by any means) to find a consensus on the issue didn't matter, because you had already decided for us that it was a big enough issue? What if we had been split down the middle? Did you represent an opinion equal to that of 10% of all Wikipedia's editors? 20%? 60%? 5%? I'd like to know directly from you what perspective you have on the consensus or lack thereof that we reached, as well on the much more relevant issue of why you think BLP's are so important. Matt Yeager ♫ (Talk?) 22:44, 19 January 2009 (UTC)

• It's not about legally allowed, but ethically allowed. Nor is about what is offensive, but rather what damaging. When information on Wikipedia gets you detained at an airport in Montreal, that's a problem. When you don't like the way we tell the history of Montreal, then it's appropriate to say "tough". WilyD 22:55, 19 January 2009 (UTC)

After all the years we have been discussing BLP-related issues, I find this comment and others of a similar ilk to be profoundly dispiriting. Matt, you have missed the point so very gravely that I don't know where to begin, although referring you to my comments from two years ago at Wikipedia:Requests for comment/Doc glasgow#Outside view by Newyorkbrad and Wikipedia:Requests for comment/QZ Deletion dispute#Outside view by Newyorkbrad, as well as the decisions in RfAr/Badlydrawnjeff and RfAr/Footnoted quotes might be one place to start. In the vast majority of cases, a Wikipedia article on an individual will be the very highest-ranking search engine result when a search is conducted on the name of that person. This affects the lives of the people we write about on a daily basis. To suggest that Wikipedia does not have profound obligations to do its best to keep these articles free of defamatory, gossipy, and privacy-invading material is to suggest that we are without obligation to consider the real-world impacts of our actions and the work we are doing. It would be intolerable for a project of our impact and influence to operate in such a manner; and it has long been a matter of imperative public importance that we should not do so.

Neither flagged revisions nor semiprotecting BLPs nor any other single step that could be taken will solve all the many interrelated problems that we class under the heading "BLP", ranging from the random-driveby-vandalism problem to the POV-pushing-attack-article problem to the invasion-of-privacy problem. As it happens, I gave a talk on this topic yesterday at the New York meet-up which ended somewhat inconclusively: neither I nor nor anyone else can "solve" the issue of how Wikipedia sometimes unfairly and negatively affects the lives of some of its article subjects—any more than we can solve the problem of how the Internet in general has this effect, for this is an Internet-wide problem and not just a Wikipedia one. Although additional concrete steps toward improvement need to be taken sooner rather than later, I have not made up my mind at all about what the best solution should be in terms of the details. But it, frankly, horrifies me that there are still dedicated Wikipedians who think that the best answer is that as a matter of principle we shouldn't give a damn. Newyorkbrad (talk) 23:03, 19 January 2009 (UTC)

I unreservedly agree with Newyorkbrad in regards to his comments. The problem we face is an incredibly difficult one to solve, short of shutting down Wikipedia. As I discussed further above (and a permanent record can be found at User:Chasingsol/FlaggedRevs), there must be some action taken. One suggestion that was mentioned to me was segregating all BLP articles to a separate namespace where they could be subject to stricter oversight, rather than attempting to apply it piecemeal to existing articles. Others include semi-protection of BLP's, the use of flagged revisions, or disallowing BLP's at all. There is no "right" solution to this quandary, only ones that can mitigate some of the concerns. We must very carefully balance our moral duties to those we have articles about, but must also not alienate the very group of people who have made Wikipedia in to what it is. --Chasingsol^(talk) 23:50, 19 January 2009 (UTC)

I neither agree with Matt's position, nor NYBrad's. I think BLPs are an area where we should be especially vigilant in enforcing our normal editorial policies such as NPOV, NOR, and V, and correct problems arising with those sooner rather than later. We should be far more aggressive with unsourced information or non-neutral presentation in a BLP than we should be in hydrogen. That only makes sense. What does not make sense, however, is what Cerejota states. We have lost sight of the fact that we don't censor. If information is available from highly-reliable sources, we should not redact it simply because someone may dislike it. We should present it neutrally, factually, and without undue weight, but if it is significant enough to be in the article, we should not shrink from presenting it.

When BLP was first getting trotted out, I was every bit on the side of it. I remember reassuring people that we would never tolerate its use as a sledgehammer in content disputes when information was reliably sourced, that we would never censor information already available in reliable sources in the name of "privacy". I feel an idiot, because I've watched exactly those things happen. Those who feared such things were exactly right. We've got to rein this thing in. That doesn't mean we need to eliminate it, but we do need to limit its scope. We can't let powerful tools get out of control, and we're in severe danger of that here. Seraphimblade ^{Talk to me} 04:20, 20 January 2009 (UTC)

I don't think anyone is calling for removing or disallowing information from highly reliable sources because someone may dislike it as the ideal of enforcing BLP. "Dislike" doesn't enter into it - potential for harm may, and an editorial judgment on what facts need to be included can certainly be brought to bear in order to limit that potential. That is the the diffuse edge of BLP concern, however - more important by far are unsourced statements of controversial fact, and I'm sure you've seen many of those. Your comment inaccurately restates Brad's (and others') position on information about living people. The policy and tools to enforce it may be being misused or misunderstood, but that doesn't indict the policy itself or the concept of due care behind it. Avruch ^T 13:25, 20 January 2009 (UTC)

I wholeheartedly agree with Matt's second point here, and somewhat with his first. That Wikipedia is an encyclopedia anybody can edit has been a fundamental tenet of the Foundation since inception; schisms have been caused by our intransigence on this matter. When did we decide that the ends justify the means, and betray our core values? Certainly we should show extra vigilance towards articles such as BLPs, where vandalism can have profound negative real-world consequences, but it is my understanding that the average lifetime of vandalism on Wikipedia articles is on the order of seconds six minutes, and I imagine it is even shorter on important, watchlisted BLPs. The push for adopting an overbroad technical solution disproportionate to the problem bears the characteristics, in my opinion, of a moral panic, one which will severely stunt the growth of the project in the long run.

More worryingly, the community as a whole does not want this solution. Newyorkbrad, I respect both your opinion and that of your fellow arbitrators, but your citing of a variety of RfArb decisions above only strengthens the argument that these concerns and proposed solutions are being imposed from the top, without the support of broader community consensus. Mr. Wales, you are the founder of the project, and I do not question the propriety of you declaring unilaterally how your hardware is to be used. But I do question the wisdom. TotientDragooned (talk) 06:45, 20 January 2009 (UTC)

Vandalism only lasts seconds? I don't think the findings of a a comprehensive vandalism study back-up that assertion GTD 13:10, 20 January 2009 (UTC)

Thank you for the link. TotientDragooned (talk) 13:24, 20 January 2009 (UTC)

I'm more worried about the controversial unsourced statements appearing in "unimportant" BLPs that aren't widely watchlisted - just as damaging Fritzpoll (talk) 13:44, 20 January 2009 (UTC)

We fix them when we get to them. There is no deadline. Perhaps we need to make this point to our readers more obvious, say, a large banner on BLPs in the main page and a constant link in the "navigation" section of the toolbar. It ain't broke, so don't fix it.--Cerejota (talk) 13:54, 20 January 2009 (UTC)

And perhaps we need to care more about our readers and about the subjects of the articles, and less about the editors (in cases where that's the choice). Things like flagged revisions may drive editors away, but the same things may also in the longer run attract readers, who prefer to get good versions of articles to read, and not the current vandalised version. And as for getting vandalism on BLPs quick: yes, a lot gets reverted very fast by our RC patrollers. But the things they don't catch can stay along for a long time, sometimes years. A recent example of a week-old blatant vandalism in the lead of a BLP was this [8], only reverted one week later [9]. And while looking at that old diff again and doing a search, I just discovered this [10], which hasn't been reverted in the last four days (I will do so now). Things are definitely broken and need to be fixed. Fram (talk) 14:05, 20 January 2009 (UTC)

I'm still concerned that Flagged Revisions are being bandied around as the "magic bullet" to solve all BLP issues. This will NOT solve the problems. We have a moral duty to ensure that information in BLP's is accurate and well sourced, but the implementation of Flagged Revisions is not going to make that magically happen. There is no guarantee that a sighting of a BLP will be done with care, or even without malice. With several hundred thousand BLP's, it is folly to believe that this will solve the issue. As can be ascertained from the uproar that is currently occurring, it seems certain to also alienate a very large group of valued editors. With all respect to Jimbo, the German Wikipedia is NOT the shining example of a good implementation of Flagged Revisions it is made out to be. If it worked as intended, with human sighting of articles, then why on Earth are bots being used to sight revisions? This is the wrong way of going about solving this difficult problem and is already fracturing a severely cracked community. --Chasingsol^(talk) 14:12, 20 January 2009 (UTC)

BLP and images

Wikipedia has a lot of work to do wrt to biographies of living people. Even the terminology we use unnecessarily limits the frame of argument. What we're dealing with isn't an issue restricted to biography articles, or pages within the umbrella of a biography (talk pages, noticeboards, etc.). The issue is understanding and respecting our ability to do harm to living people in a variety of ways, whether through direct articles, references in other articles, discussion on talkpages, noticeboards, dispute resolution, deletion discussions, etc.

Perhaps the limiting lexicon of this problem explains why its scope and seriousness hasn't penetrated its way to all dedicated members of the Wikimedia community. When we blithely allow pictures of random naked women on the English Wikipedia and Commons, refuse to require any sort verification of model age or right to publish, and defend to the death our right to hold on to sexually explicit images so that users can put them in personal galleries like "Hot" or "Appreciation of the female form"... It's clear that, as a community, we are not yet fully serious about securing living people against our potential for harm. I'd love to see Jimmy or Brad recognize this problem as serious and begin to argue as strongly for progress in image management as they have for article management. Avruch ^T 00:38, 20 January 2009 (UTC)

God damn, some of those guys have got no taste in women. Tattoos and piercings just make me want to vomit. Aside from that, yes, I can see why this sort of thing fuels your critics' flames, as in many cases the model is not identified and consent cannot be guaranteed to have been given.--Able-bodied Creature (talk) 00:46, 22 January 2009 (UTC)

I agree with you on all the details here, and so I might not be aware of the scope of the problem. I would say that a personal gallery of sexually explicit images with the title "Hot" (to pick one example that you gave) would be deleted immediately and the person who did it likely blocked. Am I wrong?--Jimbo Wales (talk) 01:25, 20 January 2009 (UTC)

I'm also not fully familiar with the problem as you (Avruch) describe it, in part because I have not really dealt with image issues as a speciality during my time on Wikipedia (the closest I came to these issues was in the two Betacommand arbitration cases). In the wake of the Virgin Killer situation a month or so ago, I did suggest adopting some sort of policy about images of minors, and proposed one, which did not enjoy a groundswell of support. (In general, I am tired of seeing the overused slogan "Wikipedia is not censored", which is intended to mean that we do not consider ourselves required to limit certain types of comments in order to shield the readership, misunderstood to mean either that (1) we can't take considerations of appropriateness in a given context into account in formulating our own editorial judgment; and (2) even more distressing, that we can't choose to limit our content in order to protect the article subjects (or image subjects) from unwarranted invasions of privacy.) I'd be interested in hearing from others with greater experience of this issue, but perhaps a subthread should be created so as not to divert attention from the original topic. Newyorkbrad (talk) 01:39, 20 January 2009 (UTC)

Unquestionably. There are a number of such galleries. Admiring the female form. Hot. Nude women & "undressing". Fav Sexy. Girls. I haven't spent a lot of time looking for these, and I'm sure there are a few more. I'm not saying they're everywhere; I am saying that they aren't effectively monitored, and that there are hundreds of images without sufficient provenance - just check out this category. You'll find that most of these pictures are at Commons, and relatively few of them currently have a home on this project, but the core issue confronts all of Wikimedia and not just its most prominent project. Also, should be noted that any of these images can be displayed on en.wp at any time and from time to time they are. Avruch ^T 01:52, 20 January 2009 (UTC)

I might be wrong, but isn't the commons be, besides all the legalese, "realistically useful for an educational purpose". I doubt random collections of naked women fit the criteria of "realistically useful for an educational purpose", however, pictures that can be offensive to users for a variety of reasons do have a "realistically useful for an educational purpose". For example, Ejaculation has a frame by frame picture of a man ejaculating. That image is offensive to many (although I suspect it is titillating to others) but it is clearly serving an educational purpose. I agree with Newyorkbrad that WP:CENSOR is overused, but it is also disrespected as a matter of course, for religious, political, and even fan-cruft reasons. The founding spirit, if I am not mistaken, of WP:CENSOR is to ensure liberty of thought over personal values - and one of the biggest problems of WP:CREEP has been people lose track of the spirit and start to see this as a legal game. I feel that we should always keep that in mind when considering not accepting content. For me the Virgin Killer case was one where the WMF took the correct position, a principled one. It would certainly been easier to remove that one image in order to guarantee access in the UK, but our position was defended. The one thing, behind all controversies, drama, and everything else, that I have always liked about wikipedia is an uncompromising position on freedom of knowledge. I'll hate to see that be eroded, simply because consensus moves that way. Sometimes only a minority defends liberty, does that make liberty less worthy of defense? --Cerejota (talk) 03:51, 20 January 2009 (UTC)

If there are images with educational value, no one is asking for them to be deleted. You're misunderstanding what I'm asking for if you see it as a censorship issue - asking for age verification, verification of right to publish and monitoring for encyclopedic use is not the same as censoring content. It's protecting image subjects who are unlikely to be able to protect themselves. At least BLP subjects can look themselves up in Wikipedia, or Google their name, to discover what untoward things are being said about them; how should an image subject go about doing that? Search for "naked_woman_on_beach.jpg" and click through a thousand images, once or twice a year, to make sure she hasn't been uploaded yet? As you can see, we won't get too many OTRS complaints - but that doesn't mean we aren't harming people by publishing explicit images of them. Avruch ^T 13:25, 20 January 2009 (UTC)

My point was, and I didn't make it explicit, that galleries named "Hot" in the commons could be discussed for deletion, in other words, we have mechanisms to deal with them, and people then decide, by consensus, if they have educational value. The problem I see with proposals that take away that responsibility away from the community, is that we are delegating our responsibilities as editors either to bureaucrats or to machines. There is already a significant erosion in editor commitment to the spirit of the wikipedian five pillars, we do not need to add another nail to the coffin: the deletion process for the most part still work. Really what people need to do is stop trying to fix what is not broken, and actually use the tools at their disposal - there shouldn't be any shortcuts to discussion and consensus seeking - we stop doing that, like with flagging, this place will become stale and disrispected: everyones roots for wikipedia to work, but the jury still undecided, believe it or not. On the matter of BLPs, we already have stronger policies (WP:BLP) and for the most part they work. If we need stronger policies, then we modify WP:BLP. In a more philosophical point, all of these discussions are based on fear, fear of vandals, fears of pov pushing, fears of children exposed to unsuitable material. I am sorry, but I am not a coward, and I refuse to act on fear. I am a wikipedian, and I am bold. --Cerejota (talk) 13:50, 20 January 2009 (UTC)

Perhaps its my fault for not being clear, and for not disclaiming specific concerns or goals first, but I'm not commenting out of fear or proposing a specific bureaucratic process. I'm not interested at this point in the exposure of children to explicit images, or of vandalism or POV pushing, as it relates to images. POV pushing is a problem with BLPs, but I think assuming that fear is behind what others have expressed as a concern is belittling and misses the point.

What I would like to see come from this discussion (in this newly created subthread) is Jimmy and Brad and others taking the problem of image management seriously, and I would like to see them incorporate it as a part of their campaign to raise awareness and provoke thought and progress in the area of BLP. Specifically, sexually explicit images should be more rigourously evaluated for encyclopedic usability and we should look into how we might do better at verifying uploader right to publish and model age. Avruch ^T 14:51, 20 January 2009 (UTC)

You were clear. We simply disagree: I am not assuming it is fear, the straight up language used is a language of fear: "protection" is only needed when one fears something and feels the need to protect oneself from it. If one doesn't fear something, there is no need to protect oneself. And while you are not proposing any concrete changes, what you propose will require those changes - we already have a process for dealing with inappropiate content, and a process for dealing with vandalism, and a process for dealing with POV pushing. If you suggest any change to that process you are, well, proposing that they be changed. Its pretty clear.

I am not so sure that there is anything wrong in how we handle image content, today. If people do not use the deletion process in the Commons and other wikipedias, is not the process's fault, it is the inaction of people's fault. If people vandalize, they get reverted. If people POV push, they get pushed back. Yes, dumbass is the number one language in an environment of massive collaboration, but from the collection of idiocies emerges a pretty amazing encyclopedic collection of knowledge.

I do not want Jimmy and Brad to go down a road that I know their instincts won't take them (erm, Jimmy is, well, lets say, aquainted with - soft - porn if his BLP is to be trusted), simply because people feel we have a "responsibility to protect the children". We don't have that responsibility, nor should we have it. We are not collective parents to the collective children. We are not the Borg. Of course, those offended can ultimately get the fork out of here. But the day we censor crap by bureaucratic order, rather than community consensus, its the day many of us will indeed getting the fork out of here.--Cerejota (talk) 21:08, 20 January 2009 (UTC)

Pardon me, I don't mean to be rude, but it seems that Jimbo does seem to attract a lot of public conversation on his talk page, so I thought it might be ok for me to comment. Regarding the explicit photos: I've always considered Wikipedia an educational tool. If it makes sense to display a picture to further the education of a topic, then in line with the not censured mindset, it seems that it would be acceptable. If showing a picture is only an attempt to "shock" someone, then it really isn't what the community wants. I say this in line with a conversation (RFC) on the urination topic going on (but it seemed somewhat relevant). I'd personally hate to see wikipedia turn into a porn site, using education as a front for the graphics. I'm guessing that most folks agree with me, but I just thought I'd like to comment. Thank you for your time. Ched (talk) 15:04, 20 January 2009 (UTC)

• I tend to agree with Avruch, but the problem is that every time this topic comes up, there is little definition about how to determine sexually explicit material that is education versus superfluous. I probably have about four photos that could be deleted as they lack encyclopedic value, and I wish I hadn't uploaded them. Unfortunately, the motivations of people who make proposals for this often do have an agenda that would affect relevant and educational material that people just don't like because they are personally creeped out or offended by the human condition (which runs counter to our educational goals). Clearly my most controversial photos are the ones taken on the set of a big budget pornographic film shoot where the actors and film crew signed releases on file with Lucas Entertainment. The arguments against these images are because they are explicit, but also because they are gay. So would the photograph at the top of Pornographic film fall in line with one of the ones that should be deleted? No genitals are shown, and the focus is not on the actors, but on the totality of the set. Most people who oppose it seem to not like it simply because it's a gay film. Images such as these are often swept up in the anti-WP:CENSOR initiatives, even though everything Avruch described above does not apply to them. This is what I see as the main problem with prior attempts to limit explicitness. So, where is the happy medium? That's the light in which I'd personally appreciate hearing Jimmy and Brad's opinions. Nobody is really making those determinations. And I'm happy to supply a list of my own images that I think should be deleted, simply because they will probably never have much use on any project (IMO). All that said, people also make the mistake to think Wikimedia Commons is here to serve only Wikimedia projects, which pisses off the Commoners; in fact, the prior sentence just did that. So what is Commons? --David Shankbone 16:13, 20 January 2009 (UTC)

The Commons is not only for wikipedia, but as I quoted above, it is for educational material. It is explicit and clear in that goal. It isn't flickr or youtube or redtube or imagefap. That said, I share your concerns, but still trust the right desicion to be made - there are enough of us who can take offensive content and defend it over our offense because we see its encyclopedic value. --Cerejota (talk) 21:08, 20 January 2009 (UTC)

This image on Commons is particularly choice, with the added bonus that the extended description makes sure to mention she was with her children at the time. Avruch ^T 02:28, 21 January 2009 (UTC)

A deletion discussion was opened, and I argued for deletion.--Cerejota (talk) 19:50, 21 January 2009 (UTC)

I could've told you it wouldn't be deleted. It's a cultural issue, really - no wider use of existing processes and policies will cause a significant change unless it is preceded by a concerted effort at changing perceptions and priorities in the Commons (and en.wp image) culture. Avruch ^T 22:28, 21 January 2009 (UTC)

You'd be surprised... Giggy (talk) 04:08, 22 January 2009 (UTC)

Wow, concensus to keep is evident at the deletion discussion for the image but User:Giggy deletes the image anyway. Way to go, who cares about following proper procedure or the consensus of the community. All of you those "OH NOES IT'S TEH PR0NZ, PLEASE DO SOMETHING FOR THE CHILDREN" people disgust me and you really should head over to Wikipedia Review where you will find plenty of close-minded, like-minded, bigots to vent your wikihatred with. Stop ruining the free encyclopedia that anyone can edit. L0b0t (talk) 15:11, 22 January 2009 (UTC) refactored so as not to ruffle feathers. L0b0t (talk) 19:18, 22 January 2009 (UTC)

WP:IAR is the most important non-WP:5P policy in Wikipedia, specially under snowy conditions. Part of being bold is knowing this: Giggy acted fearlessly clearly using this criteria. You think s/he acted wrong? There is dispute resolution available, use it. BTW Giggy is not "protecting the children" s/he is protecting the subject of the photgraphs, two entirely different criteria.

Lesson: the system works, and it doesn't need fixing... It is slow as hell sometimes, but I prefer a crap picture like this one to fall through the cracks for a while than have a system in place that can be misused to block legitimate content simply because it is unpopular or it is controversial.--Cerejota (talk) 10:36, 24 January 2009 (UTC)

The system does need fixing when "editors" can post rumors and innuendo about someone that is libelous and malicious at best, and can defame a person in such a way as to prevent them from obtaining employment or earning income. There are several profiles of musicians that are in this category, and while the most current page may have had the offending rumors pulled from it, the information still occurs in the history pages. There has to be a way to flag this information for immediate removal when it is simply rumor and hearsay and not based on any fact. If an accurate source for charges of things like "rape" or "molestation" are not verifiable, they really need to not appear on the pages of Wikipedia. OneCourtesan (talk) 21:55, 25 January 2009 (UTC)

Why I am asking Flagged Revisions to be turned on now

Note: This section was directly linked from BITS, a blog of the New York Times and a high-traffic web site, on January 23, 2009 5:46 pm EST. The BITS article was linked from Techmeme, a high-volume web site. 16:48, 24 January 2009 (UTC)

This nonsense would have been 100% prevented by Flagged Revisions. It could also have been prevented by protection or semi-protection, but this is a prime example of why we don't want to protect or semi-protect articles - this was a breaking news story and we want people to be able to participate (so protection is out) and even to participate in good faith for the first time ever (so semi-protection is out).

We have a tool available now that is (a) consistent with higher quality (b) will allow us to allow more people to edit it a wider range of circumstances and (c) will prevent certain kinds of BLP harm.

• We now have a community poll indicating approximately a 60/40 support for the future. This is a very wide margin, with 20% separation between the pro's and con's.

• The proposed configuration is significantly conservative as compared to that of the German Wikipedia, which has been successful with all articles flagged. They do, however, have an approval delay of 3 weeks at times, a figure which I regard as unacceptable. Our version should show very minimal delays (less than 1 week, hopefully a lot less) because we will only be using it on a subset of articles, the boundaries of which can be adjusted over time to manage the backlog.

• The proposal is for a time-limited test.

To the Wikimedia Foundation: per the poll of the English Wikipedia community and upon my personal recommendation, please turn on the flagged revisions feature as approved in the poll.--Jimbo Wales (talk) 23:00, 21 January 2009 (UTC)

Let me be the first to register my opposition to this. DS (talk) 23:03, 21 January 2009 (UTC)

To elaborate: we barely have enough people handling flagged new-articles. The backlog is almost a month long, and it would be longer if it wasn't for me personally working on it, and for me personally nagging people into creating software tools to speed up the task. Flagged revisions will suffocate under its own weight. DS (talk) 23:08, 21 January 2009 (UTC)

I should not generalize, but the backlog on patrolling new articles (that's what you mean, right?) is perhaps created because many people, like me, don't see the benefits of it. However, I do see a lot of benefits for flagged revisions, and will contribute to keep the backlog on those as small as possible. Comparing the two is in my opinion not correct. Fram (talk) 07:44, 22 January 2009 (UTC)

"I do see a lot of benefits for flagged revisions, and will contribute to keep the backlog on those as small as possible. Comparing the two is in my opinion not correct" I agree 100% with that statement and will also work hard to approve these flags as quick as possible. I also agree with this new policy so long as it only applies to BLP's which get a lot of editor attention, so that the flags will be seen and approved in a day or less. As an aside, I think this will help to turn casual editors into regular contributors if we are nice to them and give them positive feedback and polite constructive criticism on their talk pages when we approve or decline their submissions DegenFarang (talk) 14:47, 25 January 2009 (UTC)

A problem with the newpages patrol feature is that not enough users are automatically patrolled, only admins and bots, so it creates backlogs. We could create a usergroup just for that but that would be of too little use compared to the added bureaucracy, while with flaggedrevs on, pages created by reviewers could be automatically patrolled. So this would help to reduce the newpages backlog too. Cenarium (Talk) 11:51, 22 January 2009 (UTC)

Jimbo - is there any chance we can wait a little time so we can figure out how we're going to use it? Say 2 weeks? Ryan Postlethwaite^{See the mess I've created or let's have banter} 23:04, 21 January 2009 (UTC)

You and I don't always see eye to eye, and I've had some harsh criticism in the past, but bravo. Cheers for using your power for good here. rootology (C)(T) 23:06, 21 January 2009 (UTC)

60% isn't really consensus. As much as I respect your opinion, I ask that this is postponed, at least until a more clear consensus is developed. –Juliancolton ^{Tropical Cyclone} 23:08, 21 January 2009 (UTC)

To hell with "consensus". The majority has spoken, we don't need to wait until Wikipedia donations are drained by some silly lawsuit because we (we, as editors) couldn't see the forest for the trees. JBsupreme (talk) 01:44, 22 January 2009 (UTC)

• I have a feeling that the Washington Post article is going to be blown as big as Virgin Killer by other media using this message as the basis. - Mailer Diablo 23:12, 21 January 2009 (UTC)

(edit conflict) No - good idea, Jimbo. That was tasteless, stupid and possibly offensive, not to mention the bad publicity it caused us (an example was linked above). Semi-protection would lock out IPs, and that editor could have just made a couple more edits and done the same again, and full protection certainly wasn't warranted - Flagged Revisions is a happy medium. Dendodge Talk^Contribs 23:13, 21 January 2009 (UTC)

There is a line of people, including me, willing and able to volunteer as "trusted users" to ensure that the revision backlog stays short. Fully support the decision to implement this on BLPs. A no-brainer. Cla68 (talk) 23:15, 21 January 2009 (UTC)

I see what you're saying jimbo, but I don't believe it would've been 100% prevented. I mean, it got past RC patrol after all. Wizardman 23:17, 21 January 2009 (UTC)

Not really. It was reverted within five minutes. — Jake Wartenberg 23:22, 21 January 2009 (UTC)

Count me in the line of people willing and able to volunteer as "surveyor" to ensure that the initial set of articles is a reasonable set for the test. Fully support the decision to implement this on BLPs. A no-brainer. ++Lar: t/c 12:01, 22 January 2009 (UTC)

Provided that this is currently to be considered a test, and provided that there is an explicit time limit -- Jimmy, you don't specify one above... -- which I suggest might be two months, I think this is excellent news. [[Sam Korn]] ^(smoddy) 23:18, 21 January 2009 (UTC)

(edit conflict)x2 I think we should extend it to also cover the recently deceased - we really don't want offensive comments upsetting a person's mourning family, do we? Dendodge Talk^Contribs 23:20, 21 January 2009 (UTC)

Can we please try it for a trial period (echo Sam Korn)? I find myself agreeing with Cla68; I hear a never ending stream of complaints regarding just this type of vandalism on biographies. Bastique ^demandez 23:23, 21 January 2009 (UTC)

"please turn on the flagged revisions feature as approved in the poll"... the proposed configuration is a trial. Happy‑melon 23:27, 21 January 2009 (UTC)

I just want to note that we would not turn FlaggedRevs on here on enwiki before working out some very specific parameters for the test first. Keeping an eye on workflow and seeing what can be streamlined or taken out would be very much part of our attention. --brion (talk) 23:29, 21 January 2009 (UTC) (CTO, Wikimedia Foundation)

• As an admin who is only here for about 4380 of the 8760 hours there are in a year, and whose watchlist consists of mostly WP:BLP articles, I welcome this move with open arms. If I were not reverting, warning and blocking vandals, I could be creating new content, and just occasionaly I am able to do that. This example diff, which covers 96 edits over 10 days, shows what we are up against without this option. It's clear that whereas most of those edits may have been in good faith, few persisted. The Washington Post have, as usual for the media, picked upon an isolated glitch or two, and not, if I read the replies to their article correctly, entirely to their credit; and this is the paper of Woodward & Bernstein! --Rodhullandemu 23:31, 21 January 2009 (UTC)

But if you want to keep the Flagged Revisions backlog down, won't you have to patrol for more than 12 hours a day instead of revert for twelve hours a day anyway? NuclearWarfare _(Talk) 23:50, 21 January 2009 (UTC)

No, because that activity would be rolled up into periodic reviews rather than "as occurs", if I understand it correctly. Whereas we may lose some opportunity to warn & block vandals, we would gain by only presenting defensible articles to our readers. Result! Vandals don't care much whether their versions are visible (except in certain circumstances), so if the message is that unproductive edits won't get past a Flagged Revisions approval, I'm all for that. If they're that potentially destructive, they'll fall into the hubris trap sooner or later. Sooner, nothing will happen; later, and their whole edit history is up for grabs, and it's goodbye, Mr. Vandal. --Rodhullandemu 00:01, 22 January 2009 (UTC)

With regard to the article about Kennedy, I personally attempted to revert the article and semi-protect it. Unfortunately, at the time, I was at a local store's internet café waiting for my ride home after college. The upload speed at the connection I was using was not high enough to allow me to revert and protect the page using Firefox within a reasonable amount of time. It is possible that I would have been able to to revert using Huggle, since I had been using that tool before hand fighting normal vandalism, but the request Huggle was sending to the API to retrieve the older revisions of the page kept timing out. So I tried using Twinkle, but it hung up as well; either it didn't parse the JavaScript or the server was not receiving the request. So then I tried doing it by hand. I was ultimately partly successful, but my connection was so slow that I edit-conflicted literally 15 times trying to remove the sentence manually. And after all that, I still missed part of the speculation, and it was fixed by someone else.

You cannot protect pages from Huggle, since it is designed for recent changes patrol. At the time, it was the only option I had to interact with Wikipedia's servers in a reasonable amount of time, as I explained above. While I was trying to get Firefox to send the protect form's request to the servers, people kept re-adding the erroneous information again and again, so I ( looking back foolishly) slowed it down by attempting to remove the information from a different tab while I was waiting for the protection request to go through.

I know I am not the only person on RC patrol, and I know that me saying this now will not help with regards to that article about Wikipedia's inaccuracy. I just wanted to say, for the record, that the problem was not that no one on RC patrol saw the issue, I saw it. The problem was I was physically unable to do anything about it fast enough because of my slow connection combined with the large number of edits being made to the page. J.delanoy^gabs_adds 23:57, 21 January 2009 (UTC)

(ec)RC Patrol is a good tool, especially for potential WP:BLP violations. But still, as you say, some vandalism creeps through; semi-protection for all BLP articles would significantly reduce unnecessary edits. Although I approve in general terms of the open editing model, I wonder how much longer we can sustain it without stronger defences such as are currently under consideration, and I see these as a maturation process - I say this because there is a particularly insidious type of vandal whose edits are largely undetected, and any of which can leave us open to criticism (pls email me for details). I'd rather we were able to live without that. --Rodhullandemu

I personally had the same experience. I saw it when I looked at the page when television was reporting that Byrd had been taken away for medical reasons (they were wrong, it seems). I knew from television that there was no report of Byrd being dead, so I tried to revert. But the site was super slow at that time, other people beat me it, I got an edit conflict, etc.--Jimbo Wales (talk) 00:12, 22 January 2009 (UTC)

• The worst argument in this thread is by DS, who has essentially said it's too much work, so he doesn't want to do it. DS's statement is easy to make for someone who suffers no ill effects from having their Wikipedia BLP vandalized. BLPs are not working well the way they are, and I'm sure Jimmy can attest to the same experience I have every time I meet a notable person: they launch into a list of complaints about their article. It gets tiresome to see the inertia in the community that causes it to ho-hum not care. We like our power, but we want none of the responsibility that goes along with it. That will be the fastest way we will lose the position we have gained on-line. Kudos, Jimmy, for taking a stand. --David Shankbone 00:32, 22 January 2009 (UTC)

• Worth it to note that he (Dragonfly67) is a fanatic new page patroller, perhaps the most active patroller with the longest tenure. When he describes how much work it is, we can rely on his understanding of the process. Avruch ^T 00:49, 22 January 2009 (UTC)

• @Avruch: I'm not knocking Dragonfly, just the argument. There are a lot of things that are a lot of work. I remember back when we had zillions of copyvios on here. That took a lot of work. Flagged revs will mitigate the existing BLP problems because it will at least stem the increase. But that it's too much work? I don't buy that a valid argument to oppose. --David Shankbone 01:04, 22 January 2009 (UTC)

• Flagged Revisions are a good start but what to do with the thousands of BLP's that are completely unreferenced [11] or the thousands that are inadequately referenced [12]. Is it acceptable that these articles are allowed to remain? Is it likely that they will be adequately referenced in a reasonable time frame? Many of these articles remain largely ignored since creation, most of these articles are about people who do not appear in paper encyclopedias. Is it really justified keeping them on Wikipedia? Every single one of them could contain libellous content or could easily be edited to be libellous, who would notice? Even semi-protecting all BLP's would only have limited success, far better to delete all BLP articles that don't have coverage in traditional paper encyclopedias. If the WMF isn't prepared to accept its responsibilities in this regard, then shame on them and shame on you. Remember "Do no harm" whatever happened to that fine ideal? RMHED (talk) 00:38, 22 January 2009 (UTC)

Can we keep the thread to its subject? Why do people have to throw in every issue when only one is being discussed? Please stop trying to derail the thread with other issues, some of which will be mitigated by flagged revs. --David Shankbone 00:42, 22 January 2009 (UTC)

(ec):::I don't see it as being that off-topic; I see the issue as being closely related to the proposal that BLPs should be permanently semi-protected, although I am in two minds on that at present. I agree with the "mitigation" idea, but couple that with a dedicated BLP task force, which may be initially very busy, but whose functions will eventually subsume into normal processes; and that could be very strong in improving the quality of our project where we are most open to criticism, and therefore should arguably be directing most effort. --Rodhullandemu 00:51, 22 January 2009 (UTC)

Good call

The rationale may be a subject of debate, but it is a good call to ask developers to turn on the software that allows us to test whether flagged revisions can make Wikipedia more reliable without compromising the basic principle that anyone can edit it. The discussion has already generated a very promising idea, WP:Flagged protection, which uses the flagged revision software to allow more IP edits by using it instead of semiprotection. I think this is a good way forward. It doesn't compromise our principles as flagged protection is more inclusive than semiprotection. But it also allows us to address the really serious BLP concerns that Jimbo has articulated many times. Geometry guy 00:46, 22 January 2009 (UTC)

• Strong endorse (with reservations). The "it's too much work" argument is bollocks. We have have no right to hold third parties open to the possibility of damage that we currently do. If we can't reduce the harm drastically, then we basically can't justify keeping 300,000 unmaintainable BLPs. However, flagged is NOT a panacea or a magic bullet. We'll need many reviewers who'll be under pressure to approve as much as possible, as quickly as possible. That will lead to mistakes. It is also unlikely that the reviewers will check assertions sourced to complex (or off line) sources. That means determined libellers and hatchet-jobers will still be able to put credible falsehoods (which are the most damaging) into articles. However, this is worth trying as it should somewhat diminish the problem,--Scott Mac (Doc) 00:51, 22 January 2009 (UTC)

• Just do it. --TS 01:13, 22 January 2009 (UTC)
• How exactly do flagged revisions do anything to end the problem of unreferenced or poorly referenced BLP's? Even now these problem BLP's are still being created and as far as I can see fuck all is being done to stop this. If you try to blank these articles you get blocked for disruption, if you list them for deletion you get accused of disruption. According to WP:BLP only poorly/unsourced contentious content should be removed, who determines what is contentious? What might not seem contentious to most readers may well be so to the article's subject. As long as these articles are allowed to continue to exist on wikipedia, then wikipedia has no claim to be a legitimate encyclopedia. RMHED (talk) 01:17, 22 January 2009 (UTC)

		The Defender of the Wiki Barnstar
		For being a leader when we really, truly needed it. Seems we may yet make an encyclopaedia. WilyD 01:19, 22 January 2009 (UTC)

• Yes self-interest is a great motivator isn't it. RMHED (talk) 01:28, 22 January 2009 (UTC)
• +S It's an encyclopedia. Our "good" is information. We have a prophylactic device to prevent the distribution of "bad goods". Um, label it no-brainer. If it all goes thunk in the beginning, that's OK, we can work it out. SUPPORT. Ling.Nut ^{(talk—WP:3IAR)} 01:52, 22 January 2009 (UTC)

• Strong oppose - Although I am opposed to the trial run, I would rather it go through a trial run before it is enabled indefiniately.— Dædαlus ^Contribs 02:11, 22 January 2009 (UTC)
  • In fact, I would like to hear where exactly there exists this consensus of which you speak. I see a general amount of people in favor of implenentation for trials, but I see nothing for a complete implementation.— Dædαlus ^Contribs 02:20, 22 January 2009 (UTC)
    • It would behoove you to read what Jimbo's written before decrying it. It is, in fact, to implement a trial. WilyD 03:13, 22 January 2009 (UTC)

• Concur - on the Dutch Wikipedia we have a couple of years of experience with the predecessor of FR now, 'patrolled revs'. Patrolled revs work almost like flagged revs except all edits are visible. The Dutch experience is that the lag of unsighted revisions is never longer than a couple of weeks, most of the time it is no longer than one to three days. We have a rather small community compared with the English Wikipedia, so I think the lag will not be a problem here. Besides, Wikipedia is not written in a day or a month. There is no haste. There is no problem with a lag of even three weeks. Woodwalker (talk) 08:33, 22 January 2009 (UTC)

• Indeed, thank you very much, for stepping up, and asking for this to be installed. Even knowing you would take a lot of flak for it, I appreciate you doing what is clearly in the best interests of the project. SQL^{Query me!} 04:52, 23 January 2009 (UTC)

• Oppose. What could have been prevented, exactly? (A) A couple of articles' inaccuracy for five minutes. (B) Some bad press. (A) is so trivial I have to assume it's (B). But the genius of Wikipedia is that the principle of "anyone can edit" (and have the satisfaction of seeing their work go out to the world at once) works, even though the media chatterers don't understand or respect it. So kowtowing to bad press seems wrong-headed to me. Our future depends on those ignorant of Wikipedia's potential stumbling on an article, fixing it, and getting hooked. FlaggedRevs throws a wrench into that process, thus doing the naysayers' job for them. In the long run, I see Wikipedia becoming a less reliable and well-edited resource once this cat is out of the bag. (P.S. I am a college professor and would like to reiterate my argument about the naysayers specifically for the academic context: the professors who wring their hands about the unreliability of Wikipedia are those who don't understand it. Those who do understand it & are open to appreciating its value, or at least many of us, know that it works for precisely the reasons (even vulnerabilities) that has the ignoramuses wringing their hands.) Bottom line, why are Jimbo & others thin-skinned about bad press from writers who are rarely qualified to do justice to what Wikipedia is & ought to be in the first place? Wareh (talk) 00:05, 24 January 2009 (UTC)
  • Sorry, but I have the feeling you don't understand how FR works. Even when IP-edits become invisible to others (that's just one option), the IP itself will see his edits immediately. Besides, (A) is a misjudgement of the seriousness of vandalism and, more seriously, what I call "bad edits" (sneaky POV, sentences becoming incomprehensive, etc). Whenever I revert vandalism, I check the other contributions of the vandal. Often I find one or two edits that have not been noticed for days, or even weeks. Woodwalker (talk) 15:12, 24 January 2009 (UTC)
    • His argument is essentially that we'll be discouraging people to make contributions by enacting this change. That is a valid argument against and one that should not be left to hypotheticals as there has already been a trial on the German Wikipedia. So what was the result for .de? Did new users contribute more, less or the same? If it is more or the same then this is an irrelevant point...if it was less, then we should really consider this carefully and the costs it will have on the whole of Wikipedia DegenFarang (talk) 15:13, 25 January 2009 (UTC)
      • Actually, Woodwalker does raise one detail ("the IP itself will see his edits immediately") that had escaped me. That is a thoughtful solution to one of my major objections, and I'm glad to hear it's part of the proposal. I still have concerns about what will represent a pretty fundamental change in the original idea of this encyclopedia, and I utterly reject that I have "misjudged the seriousness of vandalism." In my opinion, Wikipedia works by the principle of "10 steps forward, 1 step backward," and if our intolerance for the backwards steps becomes too absolute, we risk holding back progress together with vandalism, which if it happens even a little will mean a net loss of valuable contributions. Wareh (talk) 22:06, 25 January 2009 (UTC)
• Oppose. Well, I already declared Wikipedia dead, so I guess I can't be upset when people start kicking the corpse. But it's a damn crying shame that Jimbo's the one wearing hobnailed boots. --The Cunctator (talk) 01:02, 24 January 2009 (UTC)
• Neutral As much as I support FR instead of semi-protection, there are gaping holes in what you propose to do, Jimbo, please specify if you would like WP:Flagged protection or something else turned on. I would also like a poll on whatever you are proposing.--Res2216firestar 21:14, 24 January 2009 (UTC)
• Concur. I have reverted many vandal edits (most but not all from IPs) that would have been prevented by a flagged revisions process. I think that quality and accuracy of content, particularly on recent news and "important people", is more important than instant dubious edits. Trusted users that meet edit criteria should still be able to make edits without FPs. Peter Campbell 23:02, 25 January 2009 (UTC)
• Support: That's a change Wikipedia really needs. But I do believe we also need the many edits of anonymous editors, though it's impossible to check all the edits properly. So what if we start only with the featured articles? That might be a consensus. NCurse _work 23:19, 25 January 2009 (UTC)
• Oppose: I would argue that this issue is only significant to individual biographies'. Why should we impose this sort of change to all pages in wikipedia? There should be some additional security features on biographical pages, but not everywhere else. SCmurky (talk) 18:56, 26 January 2009 (UTC)

Yank the band aid off quickly

You're wrong Jimbo, particularly in your assessment of "consensus," but if you're going to mandate this be done then please do so quickly and firmly so we can move past the (apparently futile) straw poll and get down to work. --ElKevbo (talk) 01:50, 22 January 2009 (UTC)

Discussion for Reviewer criteria

No time like the present to start: Wikipedia talk:Requests for permissions#Flagged revision reviewers. rootology (C)(T) 02:04, 22 January 2009 (UTC) ²

Jimbo didn't read the article

I really like the idea of flagged revisions, and I'm sure that discussion has spread somewhere else. However, I wanted to comment here about Jimbo's original post where he uses the Post's article as an example of something that went wrong: Are you nuts? Jimbo, splash some water in your eyes and read that article again. It took less than five minutes for the vandalism to be removed, and that's damn impressive. Instead of being proud of your users you instead take it as an opportunity to push for flagged revisions. What? -- Ned Scott 04:14, 26 January 2009 (UTC)

Time based "flagging"

Here's a proposal which is simpler than manual flagging and could prevent most of the vandalized content from appearing: for frequently vandalized articles set a time based constraint, so that the edit appears publicly, say, only after an hour. This way there is a time window within which vandalism can be reverted and it won't appear publicly. -- anon —Preceding unsigned comment added by 81.182.19.76 (talk) 06:05, 26 January 2009 (UTC)

This is a good idea. Manual flagging will be fine for popular articles, but not for those articles which are edited very rarely (which represent the vast majority of Wikipedia's content). Esn (talk) 08:03, 26 January 2009 (UTC)

Premature announcement

While I understand that Wikia's press relation staff encourages you to react to events and make immediate announcements, and that this is the way that politicians work (proposals of news laws after a news item), I think that, again, on this occasion, you should have refrained from making this announcement to the press. The media cannot distinguish between "I'm the boss and this proposal shall be implemented" and "I'll propose this to the community, and they'll decide" (because they are used to governments and corporations operating on hierarchical lines_, and, as a result, the worldwide media has announced that editorial controls would soon be implemented.

Throughout the years, when the media played up incidents, you have announced a variety of new measures, some of which were never implemented. This has been an embarrassment for some of those in contact with the press, who had to explain why we had not delivered something we had announced years earlier.

Another problem is that this reaction blows this editorial incident out of proportion. When you think about it, it is not infrequent that the media announces the death of live individuals or other erroneous issue. In France, the chief of the news reporting in a major radio channel ordered the announcement of the death of some TV personality be broadcast whereas they had not run the necessary checks. Do you think the channel enacted rules imposing "moderators"? No. Do you think this person was sacked? No, he was promoted to a higher management position. A TV anchor erroneously announced the retirement from political life of a major political personality. Was he sacked? No, he was given a slap on the wrist. Did the TV channel enact new rules concerning fact-checking? Of course not.

Some media, or at least some of the journalists involved, have an editorial line of discrediting Wikipedia by playing up incidents at Wikipedia while keeping silent or apologetic about incidents in professional media. You will not change that by enacting new rules in reaction to their claims ­- they'll simply conclude that they were in the right, and they'll do it again. The only way you could get them to stop would be to close public editing altogether, maybe reserving it for academics and journalists. Since that's not going to happen, further attacks will necessarily ensue.

In short, while I personally approve of the "flagged revision" system, I think that announcing their implementation in reaction to what is, in reality, a minor problem, is counterproductive. David.Monniaux (talk) 11:57, 26 January 2009 (UTC)

I want to second David.Monniaux's comment; we're already receiving requests from media outlets who are confusing the situation and don't know whether this was an official announcement, or what our plans are. For the press contacts, it makes our job more difficult when these things are not discussed with the Communications Committee beforehand.⇒SWATJester ^{Son of the Defender} 16:05, 26 January 2009 (UTC)

Add a category of 'validators'

Strongly support

I'm all for "flagged revision". Every day I check a range of articles I care about, mostly to reverse some really childish change from someone who thinks it clever to be destructive. It wastes small amount of time and also irritates me intensely. Especially when they keep making the same changes to the same article.

I also see the advantage in allowing new or unregistered users to contribute, provided it really is a contribution. Mostly it would take 10 seconds to figure out

You could have a new category of validators. People who don't want the committment or controversy of being full editors, but would be willing to passes a change if it serious and not vandalism. Not required to check it is true unless it is clearly a significant matter, like someone dying. Nor judge the style. I'd be willing to do that and I'm sure a lot of others would as well. --GwydionM (talk) 18:53, 26 January 2009 (UTC)

Just make it so...

Jimbo, time to just get 'er done - WP:CONEXCEPT.

Traveling

I will be traveling for the next 24 hours (approximately). So please don't get agitated if I am slow to respond.

Those who are in the minority who are opposed to this are invited to make an alternative proposal within the next 7 days, to be voted upon for the next 14 days after that, a proposal which is clearly aware that you are in the minority and that does not attempt to simply re-hold the same vote. I ask you to seek some detailed policy around the use of the feature that you think both you and the supporters can agree upon. Simply engaging in FUD and screaming is not going to be helpful, but I trust that outside of a few, most of the people opposed can actually work cogently with others to find a reasonable and responsible compromise position.

One possibility, and I ask you to simply consider this, although I do not support it. Suppose the plan were to simply replace the current semi-protection feature with the flagged-revisions feature? So that everything would be as it is today, with the added simple benefit that anonymous ips and new users would be able to edit things that today they are not able to edit?

Suppose further that, because the feature is softer, it could be used in a slightly broader set of cases. What set of cases should those be?

As I see this feature, and I think that those who disagree with me have mostly not studied how it works, it is softer than semi-protection. Keep in mind: this feature will allow us to unprotect the front page of English Wikipedia - and leave it unprotected - for the first time in many years.

We have a long history of working hard to extend the wiki way. Those who are interpreting this in the opposite direction are mistaken.

Well, anyway, I'm out of here for 24 hours. I'll respond when I get back, but be forewarned: I am not at all interested in discussing whether or not I'm being dictatorial by accepting a 60%-40% vote of the community. That's just insulting to me, and not in accordance with our longstanding practices of careful and thoughtful communication.

(This is one reason why voting is evil: it leads people to dig in their heels rather than work for a mutually beneficial compromise. I urge you now to do that.)--Jimbo Wales (talk) 20:06, 22 January 2009 (UTC)

Thanks for the note, and just a quick question though I understand you might not get to it right away. You say those who voted "no" in the poll should "seek some detailed policy around the use of the feature." But was this not the next step anyway? My impression that all that has been decided as of now is that we will turn the feature on, and the next step was now to decide what kind of trial runs of flagged revs we would do. This is obviously how the issue was couched in the poll (i.e., the poll was not about doing flagged revs or not, but rather just about making it possible to test it, which was as far as a number of the support votes went). So are you basically directing all of us to go hash out how we are going to test the flagged revs feature or are you suggesting something else? This is part of what I was asking you about above in the section titled "more clarity needed as to next step for flagged revs," so it would be great if you could reply to some of the questions there, though no worries if it takes you a day or two to get to that. And regardless I promise I won't call you a dictator. --Bigtimepeace | talk | contribs 20:19, 22 January 2009 (UTC)

I am instinctively hostile to the idea of flagged revisions, but I supported a trial. Further, I think the version that Jimbo proposes is something I could support, assuming a trial reveals no fundamental flaws. At the moment the flagged protection proposal is even more conservative, proposing only to augment semiprotection by flagged protection, and not proposing to extend its scope. I would emphasise, however, as I commented above, that there has been no community !vote on approving an implementation, only a !vote on whether trials can go ahead. Geometry guy 20:28, 22 January 2009 (UTC)

On a minor point, you keep hinting that much of the opposition is FUD. With all due respect, your preferred configuration (only on SPs) was not the proposed trial configuration, with the wording of the poll opposers were left hanging in the air because supporting the trial left any and all final uses of FR open to possible implementation, with the details somehow to be thrashed out after it was switched on. That leaves a huge scope of later effort, much of which was easily defineable before the poll. Which has been a huge problem, because if you actually break down the 'magic 60%' in detail, you will see support for configurations that ultimately could not all ultimately be achieved, being both technologically and philosophically incompatible. I'm all for JIMBOSEZ, but I think what you have been trying to say has not been specific enough for many opposers like me to be able to get on board, and we have had to throw up all conceivable problems that we can see. And a whole bunch of the supporters have also many times demonstrated they had no comprehension what you actually supported either, and realy were just coming at it from a totally general JIMBOSEZ angle. I think if you were intending to get involved and had some idea of your preferred specifics nailed down already as you appear to have had, I think that it should have been abundantly clear to everybody voting what those specifics were, and much off topic conversation would have been avoided, and the support cited would have been much defineable. A 'beware FUD' box at the top of a discussion page didn't do it for me. MickMacNee (talk) 20:32, 22 January 2009 (UTC)

I find Mr Wales's dismissal of opposition as "FUD" (a term with which I was unfamiliar until I saw his comments on this subject) to be offensive and demeaning. Please reconsider. DuncanHill (talk) 22:13, 22 January 2009 (UTC)

DuncanHill, if I had actually done that, I would certainly reconsider it. Instead, what I did, is invite those who voted 'no' in the poll to work with those who voted 'yes' to find a proposal that more people can support. There is no doubt that there has been a lot of 'Fear, Uncertainty and Doubt' sowed in these debates, and my point is simply that such is not going to be helpful. I encourage people instead to start reaching across the aisle and working with people on the other side to address issues thoughtfully.--Jimbo Wales (talk) 00:09, 23 January 2009 (UTC)

You obviously don't recall your "anything else is just FUD" comment which has been at the very top of the talk page for Flagged Revisions for rather a long time. But then, you didn't make your announcement of your decision there did you? You chose to do it on your own talk page, which I think doesn't say much for your attitude to the debate. DuncanHill (talk) 01:01, 23 January 2009 (UTC)

I remember it quite well, thank you very much. I'm sorry you didn't understand it. I did not dismiss opposition as FUD, I dismissed lies about what the feature does as FUD. I know you aren't one for taking a step back to assume good faith, but I ask you to please do so at this time, or else just step off the topic entirely.--Jimbo Wales (talk) 01:43, 23 January 2009 (UTC)

I don't for one moment doubt the good faith of pretty much all of those who have expressed support for the trials. I do question the propriety of one of its foremost proponents being the one to decide to close the debate. I do question your choice of your talk page as the place for your announcement. And I do feel that you really need to think about the way in which you react to those who question your actions. DuncanHill (talk) 01:57, 23 January 2009 (UTC)

I try to treat objections with seriousness, kindness, and concern. In return, most people are happy to work with me. Others simply choose to spit in my face. I accept that.--Jimbo Wales (talk) 02:40, 23 January 2009 (UTC)

You think I've spat in your face? Are you honestly suggesting that? DuncanHill (talk) 02:47, 23 January 2009 (UTC)

Yes. At the very least, it's unethical of Jimbo to close this decision himself. Jimbo, whether he is the founder or not, is partisan in this debate. He may close it as "no consensus" or "No", but he cannot in good conscience close it as "Yes". Sceptre ^(talk) 22:17, 22 January 2009 (UTC)

I have no idea what you mean. 60% support is more than enough. But rather than push the issue, I'm calling for those who are in the 40% minority to have a stab at making an offer that the other side can still support, but which you support as well.--Jimbo Wales (talk) 00:09, 23 January 2009 (UTC)

As far as I am aware, the poll is still open. You need to spend a little more time reading, Sceptre. Geometry guy 22:19, 22 January 2009 (UTC)

It's still open in name, but it's effectively closed now. And when it comes to the poll's closure, I'm saying that Jimbo, cannot in good conscience, close it as "Yes". Sceptre ^(talk) 22:23, 22 January 2009 (UTC)

"Yes" to what? Geometry guy 22:26, 22 January 2009 (UTC)

I don't understand what Sceptre means, but let me be a bit more clear. The original poll is now, as far as I am concerned, closed. It closed with very very strong support for turning on FlaggedRevs, around 60%. That's enough. However, because I would like to see a much higher level of support, I'm inviting those who opposed to put forward a compromise proposal in an effort to get to a higher level of support.--Jimbo Wales (talk) 00:09, 23 January 2009 (UTC)

I know this discussion is all over the place now but I'm interested in getting a direct answer on this issue. What exactly do you mean by "compromise proposal?" As of now there is no specific proposal for how to use flagged revs, just a decision per the poll that we will have trials. To my mind the next step was to start discussing some of the proposals at Wikipedia:Flagged revisions/Trial/Proposed trials and figure out what kind of trial or trials we will run. Do you agree that that's what we should do next? Or do you have something else in mind when you suggest those who oppose work for a compromise? I'd really like you to clarify what you think we should be doing next, because as of now I still don't understand what "turning on FlaggedRevs" means to you and that's a problem. I can say that as one who opposed this I'm very much open to discussion and compromise and imagine many others are as well. --Bigtimepeace | talk | contribs 01:18, 23 January 2009 (UTC)

I understand where you are coming from, and I generally agree with you. A significant proportion of those who are objecting are objecting for reasons that I think can be easily addressed with a more specific and modest proposal, including particular timelines and decision metrics.--Jimbo Wales (talk) 02:42, 23 January 2009 (UTC)

So do you honestly believe there is no consensus for some form of trial? There is an overwhelming number of people that believe the answer to that question is "Yes," and a small minority that seem to feel there is no consensus for any trial. As the small fringe group--presumably including you, here--YOU need to demonstrate to the majority there is no consensus from the poll for some trial. We're waiting! ;) rootology (C)(T) 22:36, 22 January 2009 (UTC)

If there are no objections within 14 days, your request will be considered valid. If there are objections, please try building consensus. If necessary, you can also resort to a poll (a very large majority, at least two thirds, is generally necessary).

Note that custom configurations will take longer to process, and might sit in the technical support queue for several weeks.

Our developers will _only_ look at the information attached to the BugZilla request, so please make sure that everything relevant is at least linked from there.

— Erik Moeller, [13]

Emphasis mine. Now, I may have only a AS-Level in maths, but I think I know that 67 is greater than 59. There is a very weak consensus, but this a consensus that would struggle to delete an article, let alone trial something that could potentially change the encyclopedia as we know it. Sceptre ^(talk) 22:40, 22 January 2009 (UTC)

Please demonstrate that there is no consensus for a limited trial from the various polls and surveys ran the past month. Can you? rootology (C)(T) 22:41, 22 January 2009 (UTC)

Brion set the required consensus for turning it on in any capacity at 67%. This poll has 59%. Therefore, this poll has no consensus to turn it on. Quite simple. Besides, when has 59% been consensus enough to do anything? The only thing I can think of that passed with that rate was Carnildo's RFA. And that's still rather controversial. Rollback hit the bar to be implemented (just, at 304:151); Flagged Revisions has not. Sceptre ^(talk) 22:57, 22 January 2009 (UTC)

"Yes" to what? Geometry guy 23:00, 22 January 2009 (UTC)

"Yes" to a trial and/or implementation. Sceptre ^(talk) 23:02, 22 January 2009 (UTC)

Ah okay so the 40% !vote is an objection to a trail and/or implementation. Now what is the percentage objection to a trial? Geometry guy 23:05, 22 January 2009 (UTC)

Still around 40%. The straw poll was for trialling. Some people, like myself, are opposed to even a trial of the feature. Sceptre ^(talk) 23:13, 22 January 2009 (UTC)

You don't serious believe that. How many objections were on the basis of a trial rather than a matter of principle or an objection to a particular implementation? Some like yourself object even to a trial, but nowhere near 40%. Do you want to go through the votes, or shall I? Geometry guy 23:27, 22 January 2009 (UTC)

Since you didn't try, I did: result here. Geometry guy 21:41, 23 January 2009 (UTC)

Bringing Carnildo into this is a red herring, because as another old timer you know pure consensus hasn't scaled in ages as well as I do. And you're still looking at raw numbers, which everyone keeps pointing out ARE NOT consensus. 59% by head count? Great. That's overlooking the fact that a lot of the opposes were not against some form of trial. Please demonstrate that there is no consensus for a limited trial from the various polls and surveys ran the past month. Can you? rootology (C)(T) 23:04, 22 January 2009 (UTC)

Because the standard falls way below what Brion set as a consensus to turn it on. And as a developer, he can overrule Jimbo on software implementation matters. What part of this are you not getting? Sceptre ^(talk) 23:13, 22 January 2009 (UTC)

I'm not getting where you've demonstrated a lack of consensus for trialing. Please provide evidence. I'm challenging your interpretation, as you've apparently challenged Jimbo's. Can you demonstrate a lack of consensus for a trial based on the various polls? Please do. rootology (C)(T) 23:26, 22 January 2009 (UTC)

I would like to note for the record that Brion does not set policy for English Wikipedia.--Jimbo Wales (talk) 00:09, 23 January 2009 (UTC)

But Brion can overrule you on anything that results in a change to the code. Ultimately, it's his decision to implement the trial, if he believers there is consensus to do so (and based on his earlier "at least two-thirds" statement about implementing FR, I honestly hope he doesn't.) Sceptre ^(talk) 00:22, 23 January 2009 (UTC)

You are simply mistaken about the division of responsibilities. This is an editorial matter, not a software/performance matter.--Jimbo Wales (talk) 01:14, 23 January 2009 (UTC)

Can you turn the software on yourself, Jimbo? 'Cause if not, Brion pretty much has the upper hand on this one. If he decides not to switch it on, technically, that's you shot,right? Thor Malmjursson (talk) 02:32, 23 January 2009 (UTC)

Iceflow I would stongly suggest leaveing biron out of this.Geni 02:41, 23 January 2009 (UTC)

division of responsibilities? Okey. Given the current setup of the foundation you do not have the power to unilaterally order it switched on. Sue does, Biron probably does unless Sue tells him otherwise and the board can pass a resolution to turn it on. So while it can be argued that it shouldn’t really be up to biron it isn’t up you other than as a voting member of the board either.Geni 02:41, 23 January 2009 (UTC)

It would be very chippy to make a fuss about the use of the term 'FUD' when the rest of what Jimbo has said offers a clear way that the community could reach a real consensus rather than the unsatisfactory outcome of a foggy and unclear poll, misunderstood by people on both sides.

I've put up a call for just such 'alternative proposals' here [14] Work for real consensus starts now. Riversider2008 (talk) 22:50, 22 January 2009 (UTC)

Well, specifically Jimmy said an alternate proposal had to be hashed out in the next 7 days, and then we'll vote to implement the alternate, or the principal "Try Flagged on all BLPs" proposal that has been perennnial. There is clear consensus for a trial. It's just which FR trial. rootology (C)(T) 22:55, 22 January 2009 (UTC)

I don't see how you can claim that there "is clear consensus for a trial", when 40% are against it. That just ain't clear consensus. The poll is/was for turning on the capability, or not. - Hordaland (talk) 21:32, 23 January 2009 (UTC)

I suggested this on the arbcom page and I'll suggest it here. For whatever reason, this community is incredibly resistant to change. There comes a point when for a feature of this importance, the Foundation and/or the devs and/or Jimbo just need to make an engineering decision and say this is a feature we are adding to the software, period. It is a part of the offering of this website no different than the watchlist or the show preview button. Content decisions and policy decisions are subject of community consensus, but something of this importance just needs to be implemented by fiat. --B (talk) 01:25, 23 January 2009 (UTC)

That sounds like "the community made the wrong decision, so I'm making the right decision for them". And once you differentiate between wrong and right decisions in straw polls and discussions, you have no right to close them. Sceptre ^(talk) 07:43, 23 January 2009 (UTC)

By the way, I just calculated the support percentage for the poll. As of the closing, it is 65.813953488372093023255813953488%(I just had to do that), or rounded up 66%. Now, this is not counting the neutral votes (I don't know exactly how to calculate that into it), but that is pretty darn close to 2/3 consensus on the issue. But, either way, it looks like it's gonna happen. UntilItSleeps PublicPC 18:42, 23 January 2009 (UTC)

Your maths must be dodgy; I get 60.3% (430/[430+283]). It'd take another 45 people to get it to get it to two-thirds. Sceptre ^(talk) 18:50, 23 January 2009 (UTC)

This is a bit like saying the American people made the wrong decision in supporting McCain so strongly. The great and good are going to go against their decision and declare Obama to be the new president.

As for the percentages, well, it is under 40% oppose, but if you break down the oppose vote, you discover the in principle objection is much lower (less than 33%), and the objection to flagged protection is lower still (20% perhaps?). This needs another vote, but it certainly merits a trial. My analysis is here. Geometry guy 21:41, 23 January 2009 (UTC)

Another alternative proposal: How about keeping a lower portion of the page for edits under 48 hours old? If they are not caught by a veteran editor and are false, let's live with it. But, at least, they will minimize any spontaneous damage and alert readers/users to the fact that these are recent changes. Commuri (talk) 00:20, 24 January 2009 (UTC)

Hello Jimbo,

For a starter, ban all anonymous ("numbered") users. Only users who have a users page with at least a minimum of data about themselves, should be allowed to edit. That should at least solve some of the problems. Users who can be identified would (may) probably be less likely to post nonsense or errors and can be more easely held accountable if they do. Peter Horn 23:48, 24 January 2009 (UTC)

Which would be great if rather a lot of our content didn't come from IPs.Geni 02:27, 25 January 2009 (UTC)

Okey. Compromise. The closest we have at the moment to flagged revisions in terms of seeing if we can keep up is the patrolled function for new pages. Now we can hope more people would step up to the plate to flag stuff but we have no real evidence they will. To this end I propose we allow anon article creation for a period of 2 weeks with the challenge to see if we can keep up with keeping them patrolled (challange advertised by whatever method people can agree on). This has the added benefit of seeing if it produces any useful content and if it reverses the apparent downwards trend in the article creation rate. It also provides an opportunity to settle some unwished business from the last experiment (turning anon page creation off was meant to be an experiment only). There are however risks that we may be overwhelmed so suggest we have a few people who have the authority to request the devs shut the experiment down if it appears to them to be causing significant problems.Geni 02:27, 25 January 2009 (UTC)

Audience

He Jimbo. I stumbled upon a recent comScore press release, and according to them, the wikimedia foundation now has an audience of 27.1% of the total internet audience. That makes it the 5th most powerful website conglomerate in the world it seems, only surpassed by Google, M$, Yahoo and AOL. Thought that might be a useful statistic for one of your speeches :D --TheDJ (talk • contribs) 19:12, 23 January 2009 (UTC)

• Shame then that the WMF still doesn't accept its responsibilities with regard to BLP's. Now when will all those unsourced BLP's be deleted? RMHED (talk) 19:54, 23 January 2009 (UTC)
  • When they are submitted to an AfD and the consensus is delete.--Cerejota (talk) 08:57, 26 January 2009 (UTC)

"Congratulations, Wikipedians. You rule. Seriously, it's a remarkable achievement. Who would have thought that a rag-tag band of anonymous volunteers could achieve what amounts to hegemony over the results of the most popular search engine, at least when it comes to searches for common topics...." A lot of dedicated Wikipedians knew it all along. GetLinkPrimitiveParams (talk) 01:44, 24 January 2009 (UTC)

Wikipedia:Flagged protection

Why are we not discussing Flagged protection more? It's a much better option than FlaggedRevs for all articles or for all BLPs or whatever. --MZMcBride (talk) 06:11, 24 January 2009 (UTC)

I would wholeheartedly second that. Flagged protection does away with the disagreeable parts of having to semi-protect pages perpetually, can protect sensitive articles, and doesn't affect articles that do not require protection/flagged revisions. As a staunch opponent of a widespread trial or implementation of flagged revisions, this is a compromise I can embrace 100%. Steven Walling (talk) 06:29, 24 January 2009 (UTC)

As an opponent of the currently suggested "trial", I would also be fully behind this measure. --Chasingsol^(talk) 07:15, 24 January 2009 (UTC)

Same here. Chamal ^talk 07:18, 24 January 2009 (UTC)

As a supporter of the currently suggested trial, I would also be fully behind this measure. Could this be a way forward?--Jimbo Wales (talk) 09:33, 24 January 2009 (UTC)

Definitely. I'm swaying towards support as well, this sounds much more agreeable and ethical. — neuro^(talk) 12:23, 24 January 2009 (UTC)

I like this a lot. I too support a trial of flagged revs, but this looks like much more workable and well thought out proposal. In fact, I was thinking that this was the sort of direction flagged revs should be going towards in the first place. --.:Alex:. 14:18, 24 January 2009 (UTC)

I was fully opposed to flagged revisions but I'd support flagged protection. -- M2Ys4U ^(talk) 14:20, 24 January 2009 (UTC)

Would it be worth starting a poll/survey/whatever to see what the general opinion is on the idea of Flagged protection? I notice it already appears to be generating a lot of interest. --.:Alex:. 14:50, 24 January 2009 (UTC)

I would say that it would be necessary, actually. This is obviously much less umm... "hostile" towards the anon editors, and is likely to generate much more support from people who opposed flagged revs based on this view. Chamal ^talk 14:59, 24 January 2009 (UTC)

I think the margin of consensus for this one is likely to be a lot higher than the one for flaggedrevs. Better idea, less drama, seems a good plan. — neuro^(talk) 15:03, 24 January 2009 (UTC)

Well I've gone ahead and done it. Best to just get a general feeling towards the idea before proceeding or coming up with details or anything, although I do think the consensus will be more favorable towards it than Flagged revisions. Feel free to make changes to the opinion selections as you guys see fit. --.:Alex:. 15:20, 24 January 2009 (UTC)

As an opponent of flagged revisions by general principle, I support(ed) flagged protection too; this is the only method in which IPs are guaranteed a greater capability to edit. Sceptre ^(talk) 15:25, 24 January 2009 (UTC)

Flagged protection is just a particular implementation of flagged revisions. Cenarium (Talk) 15:55, 24 January 2009 (UTC)

True, but the way it is used makes a big difference, don't you think? Chamal ^talk 16:04, 24 January 2009 (UTC)

We use flagged revisions as we want it. On de.wiki, they want to use it on all articles for all edits with reviewing restricted to a usergroup, this doesn't make it the only way. The poll was about implementing a passive configuration to allow trials, not how to use flaggedrevs. Cenarium (Talk) 16:13, 24 January 2009 (UTC)

It's like saying any usage of the protection system would be the same, but there's a huge difference between using permanent semi-protection on every page and using it temporarily and only on problem pages. One is clearly far more detrimental than the other. Just because this is another possible way of implementing FR, doesn't mean it's the exact same as other types of possible implementations nor that it would have the same effect. That's something that I am a little concerned will be overlooked, as every compromise should be investigated and not ruled out for being based on "the same thing". There is no one way of doing things after all. --.:Alex:. 16:14, 24 January 2009 (UTC)

^{(Edit conflict x2)}Now, this idea I absolutely love compared to Flagged Revisions. Although it is essentially a form of FlaggedRevs, it seems like it would be much less discriminate of anon edits, and the overhead of the system would be lower as well. Until It Sleeps 16:17, 24 January 2009 (UTC)

Less discriminate of anon edits compared to what? FlaggedRevs is just a feature of the software, the vote was about whether or not we should trial it, there wasn't any particular implementation in mind. FlaggedProtection is just a proposal of how to use the software, that's the stage we're at now - we agreed to trial it, now we're discussing the details. I haven't seen any specific alternative proposal yet (various things have been suggested, but nothing written up formally). --Tango (talk) 20:41, 25 January 2009 (UTC)

I did ask the question "would this mean that if 40 people see the same thing needs correcting, will it lead to 40 of the same edits sitting there waiting for a review?" and that was why I asked it, system overheads and the chance of errors/lost info seems to be a high risk if that was one side effect of FR. I really wanted to ask though, where is FP being discussed ? Cheers--Chaosdruid (talk) 17:06, 24 January 2009 (UTC)

Wikipedia talk:Flagged protection --.:Alex:. 17:45, 24 January 2009 (UTC)

As a very strong opponent of Flagged Revisions, I would be exceptionally happy to welcome Flagged Protection instead. This is perfect, the load will be lower, compared to Flagged Revisions, this is much likely to be less of a drama, more of a safer option, and I think a much more likely option to gain a better consensus on usage. Thor Malmjursson (talk) 22:12, 24 January 2009 (UTC)

• This is definitely a lot better, with a lot less bureaucracy. Having editors reviewing edits limited to existing rollbackers is also perfectly fine with me. - Mailer Diablo 03:08, 25 January 2009 (UTC)

Also agree. As someone who opposed FlaggedRevisions, I would wholeheartedly support Flagged Protection as a way to deal with those especially unstable and problematic articles. Many articles would stay open to many users who want to edit them. MuZemike 04:17, 25 January 2009 (UTC)

This seems like a verrry good idea. Should we start an impromptu straw poll somewhere? —Ed 17 ^{(Talk / Contribs)} 05:35, 25 January 2009 (UTC)

Already started at Wikipedia_talk:Flagged_protection#Feeler_poll. Chamal ^talk 05:39, 25 January 2009 (UTC)

Thanks Chamal :) —Ed 17 ^{(Talk / Contribs)} 05:47, 25 January 2009 (UTC)

Yeah, I'd second the Flagged Protection idea. Aaron Schulz 20:24, 25 January 2009 (UTC)

In order for this proposal to be acceptable to me, at a minimum VoA's idea at the bottom of WT:Flagged Protection must be included; although I can't understand why there is such great opposition to applying the protection to all BLPs. Applying the extension to only a few thousand pages (i.e. the semi-protected ones) has little net-benefit to the project IMO, and would not have stopped the vandalism Jimmy cited above. GDonato (talk) 21:40, 25 January 2009 (UTC)

No form of Flagged Revisions known to me would have stopped the error at the Kennedy article; see #What happened with Kennedy, below. Two of the mentions of his death were inserted by long-established editors (I think through confusion), who would certainly have been eligible for rollback. In one case, an admin reverted one piece of vandalism and missed another. Anyone who thinks FRs will fix everything (or even almost everything) hasn't thought it through. Septentrionalis PMAnderson 04:13, 26 January 2009 (UTC)

As currently proposed, this does nothing for BLPs. It's not an acceptable alternative.

Incidentally, I think we shouldn't be held hostage by the minority at Wikipedia talk:Flagged revisions/Trial/Votes. I note that in the last 48 hours of this strawpoll, 32 users supported the proposal while six opposed. I also notice that about 75% of admins who participated supported. This is as close to consensus as we will get on Wikipedia, and we should do it because it's the right thing to do. Cool Hand Luke 22:34, 25 January 2009 (UTC)

Do what, exactly? Geometry guy 22:40, 25 January 2009 (UTC)

Wikipedia:Flagged revisions/Trial. One of these—no one proposed a trial this weak. That's what we were voting on, not flagged protection. We can do flagged protections too, but we need a trial on previously unflagged, unprotected articles. Cool Hand Luke 22:50, 25 January 2009 (UTC)

Despite coming down on the oppose side, I concur that there appears to be a consensus to go through with the trial. Most of the responses (on either side) do not appear to be well informed. My own vote (or !vote) was based, in fact, on a lack of well-organized information regarding the proposal: the question of the precise scope of the intended test, the conditions under which it would be successful, and other fundamental methodological questions, appeared to be unanswered. Now, it may be that there are answers to these questions, but that they are just poorly advertised and buried somewhere in the WP:TLDR discussions known to those who have pursued the matter. With my oppose vote, I had hoped to tease out a more concrete proposal, since the current one seems to leave many details out. Nevertheless, Jimbo clearly believes that there is an immediate need for this intervention, and immediate necessity obviously outweighs considerations of good experimental design. I only hope this is not a complete disaster. siℓℓy rabbit (talk) 22:47, 25 January 2009 (UTC)

There is no consensus, just a panic. But a properly limited trial will not be a disaster, and Flagged Protection should not be one. Let us hope for prudence. Septentrionalis PMAnderson 04:23, 26 January 2009 (UTC)

I have to agree with (a trial of) Flagged Protection. It works in a simple, logical way and I think that if the community is so "overwhelmingly in favor" that you, Jimbo, feel empowered to make the decision to go ahead with some form of flagging, FP would be miles ahead of FR. Matt Yeager ♫ (Talk?) 07:43, 26 January 2009 (UTC)

Flagged protection is just a particular implementation of flagged revisions, so it's not either/or as you seem to suggest. - Hordaland (talk) 10:38, 26 January 2009 (UTC)

I strongly support both suggested implementations - I don't care which we implement, but we need at least 1 of them. (Would it be possible to implement either normal FlaggedRevs or Flagged Protection on a per-article basis, similarly to the different levels of protection, so we get a hierarchy that goes something like:

1. Full protection (strongest)
2. Flagged revisions
3. Semi protection
4. Flagged protection (weakest)

That would plug the gaps that exist below semi protection, and between semi and full protection - something we have a real need for (the latter more so, IMO). Dendodge Talk^Contribs 17:59, 26 January 2009 (UTC)

Why not continue as is, with text added immediately, but any text added in the last Hour or Day or Week (depending on volatility of Wiki entry) "highlighted" as new. This could be by colour, font or underlining. I realise change of colour etc raises "ergonomic issues", but maybe the form of highlighting could be configurable by the user. The idea would mean that updates could be added but readers could be immediately on guard for odd new additions. The "highlight" could either be turned off automatically, or by an editor. It could even "decay gracefully" getting, say, "less hot", e.g. changing from red to black, over a few days. —Preceding unsigned comment added by Desmoh (talk • contribs) 18:43, 26 January 2009 (UTC)

Userboxes

Hello, I'm (mainly) ru.wp user. Now we are trying to validate rules about user-boxes. Results of the voting will be "let them all live". But there is a lot of forcible arguments, based on WP:NOTBLOG, that all that have no connection with the articles writing should be banished. Means taboo on all that can bring ethnic, national, religious or orientation-based hostility.

But in this prohibitive mood they want to suppress all, that makes no harm - humor for example, because it is useless. I believe that in user-generated content project there is a place for a smile, which brings friendly atmosphere.

Your words are very important for us all. I see you have "42 answer" user-box on your page. Am I right trying to save such userboxes? Carn (talk) 08:55, 24 January 2009 (UTC)

I just now removed that editbox, because I don't care for it. My own view is that userboxes which have any possibility to offend or to be divisive in any way should be forbidden, as you call it "that can bring ethnic, national, religious or orientation-based hostility." (I would say that can be bring any hostility.) The little harmless joke ones, well, I don't care for them personally, but I think it is possible that they can be allowed with little trouble.--Jimbo Wales (talk) 09:04, 24 January 2009 (UTC)

Thanк уou for your opinion, i hope it helps us bring order into userbox chaos.Carn (talk) 09:42, 24 January 2009 (UTC)

• Hello, Mr. Wales, I’m an active ru-wiki user. (1) One can see that there are hundreds of political userboxes on the page Wikipedia:Userboxes/Politics by country and Wikipedia:Userboxes/Politics, e.g, Anti-communist; anti-fascist; For Kosovo's independence; for Scotland's independence; for Tibet's independence (and some more divisive). Do you suppose all persons who place such userboxes on their pages to be «bad Wikipedians»? Do you suppose your own view on that question to be the undisputable rule for the project?

(2) Please, imagine such situation: one en-wiki administrator will delete such userboxes (e.g. communist / anti-communist; for Tibet’s independence / against Tibet’s independence) without any discussion, but with commentary «Jimbo said that such hostile userboxes should be forbidden». What do you think about his actions?! Thanks for your answers! --Chronicler (talk) 16:42, 24 January 2009 (UTC)

I do think that all such userboxes should be forbidden. I don't think an admin should act unilaterally to do it.--Jimbo Wales (talk) 03:58, 25 January 2009 (UTC)

(3) One more important question. In WP:NOT is stated that: «Wikipedians have their own user pages, but they may be used only to present information relevant to working on the encyclopedia». Do you suppose such information as anyone’s nationality; religion; orientation; number of children and so on to be relevant or helpful for Wikipedians? Do you agree that such userboxes should be forbidden? Thanks for your answers! --Chronicler (talk) 16:37, 25 January 2009 (UTC)

It seems to me that if one supposes such common personal information to be inevitably divisive and hostile, he will violate WP:AGF --Chronicler (talk) 16:44, 25 January 2009 (UTC)

Those letters "AGF" stand for "Assume Good Faith". They have no bearing on the opinion that importing statements into Wikipedia inevitably leads to sectarianism. Good faith disagreements on real world matters are notoriously divisive, and a look at our history of arbitration cases shows that such good faith disagreements lead to extreme disruption where they are imported by editors into Wikipedia. --TS 16:55, 25 January 2009 (UTC)

I'm disheartened to find this page even exists, and I've nominated it for deletion at Wikipedia:Miscellany for deletion/Wikipedia:Userboxes/Politics by country. I've given my reasons on the nomination.--Scott Mac (Doc) 18:17, 24 January 2009 (UTC)

Celebrate two years of conflict resolution WikiProjects!

On Jan 24, 2007, WikiProject Sri Lanka Reconciliation was founded as the first WikiProject specifically focused on conflict resolution. It has since been used as a model for WikiProject Israel Palestine Collaboration, which in turn became a model for WikiProject Ireland Collaboration.

I originally wanted to invite people to some form of virtual birthday party, but we couldn’t think of a suitable equivalent to candles, cakes and funny hats. Instead, I put a yellow box on top of our project page, where we will keep some info about celebrating the event and the history of the project.

I think this is also a good time to point you to some of the things we do slightly differently from the rest of Wikipedia, and which I think may be worth considering. These are described in our guidelines. Of particular interest to you may be how we classify reliable sources, and how we built robustness against trickery into our procedures. One aspect of the latter are sockpuppets: In the last two years, I noticed an increased fear of sockpuppets in other areas of Wikipedia, while in our little area, that has ceased to be a problem, because we focus on viewpoints and edit versions, instead of vote counts and reversion counts. (I think the sockpuppet-hunt is particularly important because it threatens Wikipedia at its raison d'être: being the free encyclopedia.) The key to our success, as I see it, is providing a home for experienced editors from both sides of the conflict who help the newbies stay out of trouble, and help them make their points without having to resort to trickery. — Sebastian 21:11, 24 January 2009 (UTC)

French speakers needed for Tunis

I'm in the process of transwikiying this which is a core article. For the many zillions of people who watch this page if there are any French speakers please help translate a paragraph or two, should only take a few minutes and the more people help out the quicker it will get done! Dr. Blofeld ^{White cat} 20:40, 25 January 2009 (UTC)

Alternative Suggestion

How about having all revisions have a one-hour delay before they are viewable by the general public? However, anyone signed in with a legit account can see them (and revert them)? A lot of the worst vandalism gets reverted in less than one hour and this would catch a lot.

--76.205.215.216 (talk) 21:36, 25 January 2009 (UTC) --WickerGuy (talk) 21:37, 25 January 2009 (UTC) Also have all changes quickly e-mailed to any user who has a page on their watch-list. --WickerGuy (talk) 22:25, 25 January 2009 (UTC)

Wikipedia:Wikipedia Loves Art

Hi Jimmy. Thanks for all your help with this project, and with Wikipedia:Wikipedia Takes Manhattan. It would be great if you could possibly upload the video sometime in the next few days. Thanks!--Pharos (talk) 00:35, 26 January 2009 (UTC)

What happened with Kennedy

Please see a review of what happened to Ted Kennedy, which I have posted at Wikipedia_talk:Flagged_revisions#The_Kennedy_test. In brief:

There were seven additions of Kennedy's death, mostly at different points in the article.

• Two of these were by long-established accounts, confused either about the facts or which way they were revising the text. Under most implementations of FR, these would have been sighted immediately, and therefore visible to the general public.
• In one case, Newbie A added Kennedy's death, then Anon B added "January 20, 2009" as a free-standing paragraph. An admin reverted Anon B's edit to Newbie A's version; under all implementations of FR I know of, this would have sighted the version containing Newbie A's vandalism.
• There were two cases in which one anon vandalized the article and then another anon reverted. This would probably have taken longer under FR, since the second anon wouldn't have seen it to revert. (The last case is similar, but removed by an established editor.)

In all, the article was in error for less than ten minutes.

Under FR, at least one, probably three, instances of the false claim would have been visible to the public, and for almost as long as it was in reality. Flagged revisions is not a panacea, even if the Washington Post says it is; why do you trust their political blogger's judgment on a technical question? Septentrionalis PMAnderson 03:50, 26 January 2009 (UTC)

I think this analysis is simplistic. Why don't we contact the long-established accounts and find out why they added it. If they don't have a good reason, then they could be talked to firmly ("We don't add death reports until we have a source") and if they insist on continuing the behavior, they can be banned. Let's ask the admin if they would have been able to take more time to look at what they were doing if they weren't acting out of the legitimate fear that the article contained misinformation at a critical moment in time? And please don't insult me by assuming that I simply go by what a blogger at the Washington Post says: such an approach to talking to me may prevent you from hearing what I am saying.--Jimbo Wales (talk) 08:37, 26 January 2009 (UTC)

The edit summaries suggest that one or both of the established editors thought they were removing vandalism, and may well have been confused as to which way their edit was going. This does happen; I've done it myself, under less time pressure than mass vandalism. But what matters is that they did add it, for whatever reason; reviewers may be expected to do the same thing under FR. Septentrionalis PMAnderson 14:03, 26 January 2009 (UTC)

• I am glad to see Jimbo's assurance that his decision is not driven by the mass media; he may want to tell Signpost. Septentrionalis PMAnderson 14:05, 26 January 2009 (UTC)

I don't believe you are correct in your analysis of the second point. When the admin reverted the edit, his edit would not have been autosighted as it was made on top of unsighted edits. In order to make the changes visible, he would have needed to review the article manually, whereupon he would have seen a diff of all the changes since the last sighted version, including the changes by Newbie A. As an unsourced and contentious claim, it would certainly have been noticed and in all likelihood removed. But there is no risk that the addition of the death-date would have been 'accidentally' sighted because that's not how the system works. Happy‑melon 08:56, 26 January 2009 (UTC)

This is at least the fourth different claim Happy Melon has made about how "the system works". I am tired of assurances that "nothing can go wrong", without evidence or specifications. Murphy rules. Septentrionalis PMAnderson 14:03, 26 January 2009 (UTC)

Would you mind quantifying that accusation, Pmanderson? Naturally if I have made incompatible claims then I will need to withdraw one of them, but I do not believe that this is the case. If you attempt to reconstruct the edit sequence on en.labs you will see that it pans out exactly as I have described. Could you please provide some quotes of where I have implied the system functions differently? Happy‑melon 15:37, 26 January 2009 (UTC)

Yes, I would mind; I do not want to comb through your edit history to find claims of months ago. At a minimum, I have seen no adequate and precise account of how the system works, and I have seen hand-waving which certainly suggested what I said. Septentrionalis PMAnderson 15:44, 26 January 2009 (UTC)

Well in that case I assume that your accusation has no material foundation, and will ignore it. Given that a picture is supposedly worth a thousand words, how many volumes is the fully functional demonstration that I constructed worth? The only way to get a correct picture in your own mind is to try it for yourself. Happy‑melon 17:51, 26 January 2009 (UTC)

No thank you; trial and error is not good enough. The only way to fully understand a system is to see its specifications. Happy-melon's assurance of the perfection of a system he cannot describe are hot air and evidence of bad faith. Septentrionalis PMAnderson 18:23, 26 January 2009 (UTC)

To deliberately decline the opportunity to verify your claims before (or indeed after) making them is not a very constructive approach to discussion; nor is stating that you effectively cannot be bothered to take the time to ensure that your statements are factually correct. I am more than willing to take the time to explain any part of the FlaggedRevs system to the best of my ability to ensure that people have the fullest possible picture of what it is and what it does. If you have read "hand-waving" descriptions, I hope they have not come from me. The system has been explained on numerous levels, ranging from the raw source files which are as always available for public view, right through to the aforementioned demonstration, and also by innumerable explanations that I and others have given on a dozen talk pages. Which aspects of the system would you like me to describe? And will you actually read and digest such an explanation if I take the time to write it? Happy‑melon 18:37, 26 January 2009 (UTC)

And to begin what I fear will merely be another loop of the above discussion, when have I ever presented the impression that FlaggedRevisions, in any incarnation, is in any way "perfect"? Happy‑melon 18:39, 26 January 2009 (UTC)

A terrible test case

Whilst being a very strong believer that radical reform of BLP editing is required, I think Ted Kennedy is a real bad rest case. 1) Firstly bad edits to an article like this will be (and were) quickly spotted and reverted by someone who knows enough to spot the lie. 2) The reader is likely to check other sources and realise that what he's read is crap. 3) Ted Kennedy has a huge reputation, which a wikipedia article cannot influce much. In short, crap on "high profile" BLPs may be heavily embarrassing to wikipedia from a PR angle, but it doesn't and can't do much harm to the subject. The real problem is when this happens to a BLP of a less notable figure. It is far more likely that the lie will remain for months because the article is underwatched, and any watchers may not spot the lie as they are ignorant about the truth. Further our bio and its mirrors may be the only "encyclopedic" content on the net, thus opinion forming for anyone googling in a way celebrity articles are not. On Kennedy stuff like this is simply high-profile graffiti, on other BLPS it is potentially destructive to the subject, his business and his reputation. Can we stop using Sarah Palin, Ted Kennedy and other household names as the litmus test for our policies?--Scott Mac (Doc) 12:17, 26 January 2009 (UTC)

Then this case shouldn't be used as justification for FR; but it is already being so used. Septentrionalis PMAnderson 14:06, 26 January 2009 (UTC)

Well, this case is an argument for FR. All I'm saying is that there are far far stronger ones.--Scott Mac (Doc) 14:53, 26 January 2009 (UTC)

Then they should be mentioned. The other specific case I have seen mentioned was an article which made a serious accusation against its subject, but cited a source for the accusation. (The source was a revisionist website, but telling that would require a degree of checking which Happy Melon keeps insisting is unnecessary for every edit.) Septentrionalis PMAnderson 15:44, 26 January 2009 (UTC)

Flaggedrevs in the news

Medias have started reporting on flagged revisions, but often misleadingly, saying that it will be used on all articles, for all edits, some says that this will ban anonymous contributions. Now it's done... but maybe it could be clarified that this would only be used on an experimental basis, at least initially, and more importantly, that it'll be used only for certain articles, as a protection measure, temporarily or indefinitely, and that this is also intended to allow more anonymous contributions, by replacing the semi-protection system. Has the foundation considered a press release ? Cenarium (Talk) 12:57, 26 January 2009 (UTC)

This story [15] is currently on the front page of the BBC news website (international version). DuncanHill (talk) 17:46, 26 January 2009 (UTC)

And on the front page on the UK version. I notice it comes on the front page under Americas and is the second story on the Americas page which I thought was rather odd. Davewild (talk) 17:51, 26 January 2009 (UTC)

A press release would be highly appropriate given the size of this brou-ha-ha. WP's normal decision-making system - namely, having an enormous fight - may be well-suited for keeping the place stable and not too chaotic, but here nobody who hadn't actively followed the debacle has a chance of seeing what's going on, and editors are finding out about the situation (whatever it is) from the media. --Kizor 18:10, 26 January 2009 (UTC)

Alternate solution: color-coding based on age of edit and page views

I remember reading a proposal a few years ago on some other site about this idea, but I'm afraid I couldn't find the original page. Anyway, the idea is this:

All immediate edits show up a red text. If the edit is one second old with no page views, it's bright red. As the time of the edit grows, and as the page views grow, the text gradients from red to black. For example, if the edit has been there for a year on an unpopular page, or if the edit has been there for 2-3 weeks on a popular page, it would likely be black.

The whole basis of this is sort of that same "The Cathedral and the Bazaar" mindset of how OSS works: Many eyes make all bugs shallow. If many people have seen the change without a re-edit, it's more likely to be correct, and it would give the edit the privilege to show up black. At the same time, this would still allow the text to show up. It's a warning to the reader that "This edit has not been time-tested, and it may or may not be correct. Check the sources, and verify the information." It also encourages people to source new edits to legitimize them better.

This idea could be combined with the original Flagged Revisions process. As well as factoring time and page views, it can also factor source reliability ("Is this user registered and made reliable edits before?") and approval. The content could immediately turn black after an admin approval. That could also just be the system for approval: red=unapproved, and black=approved.

To summarize, here's the three different ways of approaching this system:

1. Text is color-coded based on time and page views.
2. Text is color-coded based on time, page views, and source reliability. If the context is approved, the text immediately turns black. If not, it still has a chance of turning black based on time/page views.
3. Just like the original Flagged Revision process, except that unapproved text still shows up red, and will turn black after approval.

It all depends on how simple or complex that you want it. However, I strongly believe that any of these solutions would solve for both striving for speedy edits as well as projecting reliable information. (I personally like option #2 myself.)

SineSwiper (talk) 14:39, 26 January 2009 (UTC)

I like this idea, and let me suggest a slight variant:

When an unregistered user edits a page, that change shows up to them straight away and they see the change in black (this allows us to keep taking on new users at the same speed we were before, people still get the benefit of seeing their good faith edits right away). Other unregistered users initially see the page as it was before the change, thus removing the problem of news agencies picking up on every piece of vandalism that happens to a famous person's mainspace article. An established registered user does see the change, and sees it in colour as you describe. This allows 'ad hoc' vandalism reverts to continue- I figure most Wikipedians are like me and don't often check the 'Recent Changes' page, but will fix vandalism they come across while browsing. To simplify the revert, maybe even turn the coloured section in to a hyperlink which when clicked flashes up a box with the options "delete change", "allow change". Clicking "allow change" would make the change visible in black to all users, including unregistered ones. Delete change allows a simple reversion. This idea could even be extended to text that has been deleted by an unestablished user, simply have deleted text remain visible in grey (or perhaps struck out) to established users. Kaid100 (talk) 15:47, 26 January 2009 (UTC)

Not sure if I really like the idea of different views for different people. It makes troubleshooting very difficult, and remember that you're an unregistered user until you re-login. I often forget to login until I notice it when I go into edit mode.

After all, the whole point of the colorization is to keep the edits immediate, while still giving notice to the viewer that the information is new and possibly may not be accurate. Any attempt to remove that immediate view defeats the purpose of the rapidly-updating ecosystem within Wikipedia. SineSwiper (talk) 18:28, 26 January 2009 (UTC)

View as a Consumer of your product

To be blunt folks: I rely on the principal that my sources of information are accurate; if they are not, they are useless and I will find other reliable sources. Debate all you want, but if what I read on one of your pages proves false, misleading or inaccurate the credibility of the entire resource is then in question. It is all well and good to be a community, be a responsible one. As is true in other occupations, “Trust and Verify”. 68.148.21.154 (talk) 16:37, 26 January 2009 (UTC)

We have a number of policies and guidelines on this: WP:V and WP:RS to name 2. If you use Wikipedia as a reference material, only use things with a little blue number (like ). Click on that number and it will take you down to something that looks like this:

That will usually be linked to another website (unless it's an offline reference). Check that site to see whether or not the information is there. Dendodge Talk^Contribs 16:47, 26 January 2009 (UTC)

And don't trust our little footnotes implicitly either. Wikipedia is not a reliable source. In fact, if we're going to change policy on this, a simpler solution would be to accept disclaimers, and say on every BLP page some messages which would admit that: Wikipedia is not a reliable source; it is a work in progress. If this page is in error, please edit it. Septentrionalis PMAnderson 17:02, 26 January 2009 (UTC)

Something like the message that's already linked-to on every article, perhaps? – iridescent 17:06, 26 January 2009 (UTC)

But shorter and actually on the same page as the article. It is clear that the anon Consumer missed that page of legal boilerplate. Septentrionalis PMAnderson 17:12, 26 January 2009 (UTC)

---

The question you should be asking is not "Why isn't Wikipedia a reliable source of information?", but "Why does everybody assume that closed-sourced encyclopedias are reliable sources of information?"

You don't always know that your sources are going to be accurate. Closed-sourced encyclopedias suffer from the illusion of accuracy. Yes, they are probably accurate a majority of the time, but you don't know the biases of the editor of the entry or biases inherent within the company itself. When something IS inaccurate on a closed-sourced encyclopedia, very very few people will question it because they believe 100% of what is in the source.

This is a serious problem in the mentality of people with how they question the validity of information. By assuming that a piece of information could be wrong and double-checking your sources and information, you yourself control how confident you are in that information. No one source should EVER be the sole point of information! Most teachers teach that in high school.

However, Wikipedia allows you a single starting point, with direct sources for most statements, so that you can confirm the data elsewhere. The footnote system is something that encyclopedias don't have, and no other encyclopedia updates as fast or has more entries than Wikipedia. You have to accept that the current system makes it this way. At least Jimbo acknowledges the problem enough to try to change the system (hopefully without sacrificing the positives). SineSwiper (talk) 18:49, 26 January 2009 (UTC)

Problems with the original poll

I actually agree with Jimbo's interpretation that 60% is usually enough to change something like this. However, I would like to state that in my opinion there were serious flaws in the manner that the poll was conducted. I have more than 11,000 edits and have been editing for almost five years and was not even aware that such a poll existed until I read about it on BBC news when the poll was already closed. I think that with proposals that could change such a basic pillar in how Wikipedia operates, there really ought to be a better method of notifying the general Wikipedia population that such polls exist. If this proposal is not adopted I think these votes have the capability of degenerating into elitist farce in which the vast majority of Wikipedia's editors have no real input into something that is supposed to be collaborative.- Moshe Constantine Hassan Al-Silverburg | Talk 18:20, 26 January 2009 (UTC)

It was listed on your watchlist notices at the top, at the Village Pump, at WP:CENT and many others. Short of spamming messages to all talkpages, I'm not sure what else Happy-melon could have done to notify you :) Additionally, the BBC article is very misleading - there is no plan at present to expand FlaggedRevs to all edits on Wikipedia. Jimmy and others (including me) want it trialled on BLPs, and there is discussion about augmenting semi and full protection with it at WP:Flagged Protection Fritzpoll (talk) 18:42, 26 January 2009 (UTC)