Wikipedia:Protecting BLP articles feeler survey/Actual experience with flagged revisions

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

---

Actual experience with flagged revisions

Many people here only cast their vote with a vague idea about flagged revisions. This section is for reports of hands-on experience.

• Sebastian's experience: See de:Wikipedia_Diskussion:Gesichtete_Versionen#Erfahrungsbericht_von_SebastianHelm. Depending on interest, I can provide a partial or full translation here. — Sebastian 19:13, 29 December 2008 (UTC)

• • Am I to understand that flagged revisions have already been tried in German Wikipedia? If so, then yes, I think more information about how that worked and how it worked out would be extremely useful. I scanned this page and didn't see any information about it. Scarykitty (talk) 05:26, 30 December 2008 (UTC)

    
    Yes, they have. Actually, it has already been implemented in a number of Wiki projects. [...] — Sebastian 21:57, 30 December 2008 (UTC)

    I moved the rest of my reply and the section on my experience to Wikipedia:FlaggedRevs fact sheet. — Sebastian 19:16, 31 December 2008 (UTC)

What does Jimmy Wales think?

Jimbo also said: I think that almost any argument, on any topic, which has premises beginning with "Jimbo said..." is a pretty weak argument. Surely the merits of the proposal should be primary, not what I happen to think.

This is what Jimmy had to say about this:

"I would be thrilled with the implementation of flagged revs on all BLPs as a test. I don't think I should force it through, but I strongly support that we experiment with it. What I will do is this: I will gladly serve as a formal point of contact to ask the Foundation directly to implement whatever we decide on. What I recommend is a timed test, i.e. turn on flagged revs for all BLPs for 3 months, and have a poll in the last two weeks of that period to determine whether we want to keep it. I feel confident that we will.

I would also support us simply copying what the Germans have done with it. I know there are concerns about volume, but the Germans are able to deal with it just fine as I understand it. Yes, we have more edits, but we have more community members. So I reckon we can deal with it quite well. However, I'm *thrilled* about it for BLPs and merely *supportive* for all articles.--Jimbo Wales (talk) 23:24, 19 December 2008 (UTC)"

Not all Wikipedians believe Jimmy Wales’s opinion is more important than other Wikipedians’, or is in any way a form of guidance.

Unsigned and undated post

Which is funny, because Jimbo also says:

“ "You can edit this page right now" is a core guiding check on everything that we do. We must respect this principle as sacred. ”

Richard75 (talk) 00:12, 31 December 2008 (UTC)

Exactly. That's what flagged revisions allows instead of having to semi protect or fully protect articles. Nothing in flagged revisions prevents "You can edit this page right now". But protection does prevent that. - Taxman ^Talk 02:44, 31 December 2008 (UTC)

Quite right. "You can edit right now" is fully respected. The edit may not be seen until another human looks at it, in the limited case of Biographies of a Living Person. That delay is the balance we strike between the obligation we owe to our editors and readers on the one hand and the obligation we owe to the human beings who are the subject of a BLP on the other. It respects "you can edit right now" in a way semi-protection does not. David in DC (talk) 02:57, 31 December 2008 (UTC)

I think the principle ought to be "you can edit right now, and your improvements will be visible immediately". If you're a non-newbie user, then semi-protection still lets you do this. Whereas (if i understand correctly) flagged revisions violates this for everyone. This will create a class of "gatekeepers" and will end up contravening the open spirit of Wikipedia. --Cybercobra (talk) 09:35, 4 January 2009 (UTC)

I'm far more open to a trial of this for BLP's than for all articles, but I think it's important to point out what it means to "edit" something. The wiktionary entry says that the verb form of edit means "to change a text, or a document." Part and parcel of "editing" is actually making a change, and seeing it. If your "edit" doesn't appear for days, weeks, or not at all (for whatever reason) then you have not edited anything. If we are going to do these flagged revisions in some form, let's not pretend that what we are doing is not a significant departure from the "you can edit this page right now" principle—a departure which is qualitatively different from semi or even full protection. --Bigtimepeace | talk | contribs 21:17, 12 January 2009 (UTC)

Semi-protection is an edit restraint used in severe cases of vandalism on very few articles in a given timeframe, while flagged revisions impair editing at a large scale. Also, I think the small quote in the right is the only good Jimbo quote put in here. Just because the guy invented this thing doesn't mean we have to worship him. Admiral Norton ^(talk) 16:22, 13 January 2009 (UTC)

I assume that a big (perhaps one of the only) reason why he supports flaggedrevisions for BLP articles is because it would greatly reduce the number of legal threats the Wikimedia Foundation receive, which would therefore reduce the number of WP:OFFICE actions invoked, which is always an extremely good thing. --.:Alex:. 22:18, 18 January 2009 (UTC)

IP Comments

Similar to what happened on the DE Wikipedia the IP editors are not invited to this discussion, good thing is we can learn from it and prevent the same mistake, logged in we get a nice "A survey on options for protecting biographies of living persons is under way. Editors may participate by commenting here" on the watchlist, is there a way to invite the 85 % of good IP editors for the discussion if the option is raised that their contributions go on the queue list ? Mion (talk) 19:37, 24 December 2008 (UTC)

That would mean putting this on the general site header instead of the Watchlist template. I don't know the answer to whether this is good or bad, where to get consensus for such a thing, or if such consensus (which would be decided by logged in users) would be even possible. rootology (C)(T) 19:42, 24 December 2008 (UTC)

1/3 of all edits are done by IP editors (based on User:Dragons flight/Log analysis) for the last 8 years, I think they have enough credit to be invited to the discussion before they are blocked in editing. Mion (talk) 20:46, 24 December 2008 (UTC)

Back to the point, I thought someone mentioned that the general site header is the same for logged in and IP editors, so maybe this discussion needs to be stalled and restarted after the fundraiser is over, i guess its the same route to consensus for placing it on the general header as for on the watchlist header? Mion (talk) 21:05, 24 December 2008 (UTC)

We could put it in the MediaWiki:Anonnotice, but there's this extremely bright contradiction between this just being a "feeler survey" and this being something that will cited in the future as consensus.... --MZMcBride (talk) 21:21, 24 December 2008 (UTC)

What he said. There's (so far) what seems like consensus for a small set test (maybe FAs? Maybe BLP FAs?) to see how it goes. Maybe we'll all hate it, maybe we'll all love it. But this is just to see if the idea has any merit, which it seems to clearly, but not for "lets do it German," and we're not deciding to turn on anything on some wide scale yet. rootology (C)(T) 21:50, 24 December 2008 (UTC)

The point so far is that in this consensus 1/3 of the editors are excluded and exactly the editors that are restricted by the measure, so I would like to turn it around that who created this survey first sets up a survey for the editors who are affected and from there we can start the discussion about it, MediaWiki:Anonnotice is a good way, but this survey is already filled in by the other 2/3 group, so maybe a new discussion in 2009 is a good idear ? Mion (talk) 22:02, 24 December 2008 (UTC)

Being that registering is trivial, this is really a moot argument. Those who are kept from editing and want to edit bad enough will register Timmccloud (talk) 17:31, 29 December 2008 (UTC)

This way lies Citizendium, which is a failure. Those who would prefer their method should attempt to salvage that project; it may well deserve to succeed if we fail. Septentrionalis PMAnderson 21:29, 30 December 2008 (UTC)

A poll announce could be placed on IPs' editing page. IPs would see the announce when they edit an article. Nicolas1981 (talk) 12:36, 3 January 2009 (UTC)

Cross-reference

A few days ago I posted some thoughts and points for this discussion on the talkpage. Just cross-referencing here in the hope they don't get overlooked (I didn't want to post them on this page which was just the "survey" at that time, but if anyone wants to transplant to over here I won't stop them). Newyorkbrad (talk) 17:47, 26 December 2008 (UTC)

As is usually the case, Brad's points on the talk page are well worth reading.

As is usually the case, he generates a whole lot of light without turning up the heat. David in DC (talk) 18:27, 26 December 2008 (UTC)

Time/edit based flagging

If there is some question about wither article development will suffer because anon users can't edit articles, why don't we only flag articles that have a certain number of edits or have existed for a certain length of time? This would mean that newer or less developed articles would be easy to contribute to, while older and more mature articles (that probably have content worth protecting) will require a users to get at least a little involved with the community before they get to contribute. Would this address some of the issues coming up in this discussion? --NickPenguin_(contribs) 04:24, 28 December 2008 (UTC)

FlaggedRevs doesn't mean people can't edit (otherwise it would just be semiprotection), it just means that their revision won't immediately be seen by everyone. In any case, the goal isn't to keep the good info in, its to keep the bad info out. A low traffic stub is just as susceptible to libel as any other page. Mr.Z-man 05:11, 28 December 2008 (UTC)

Where the period for which the edit not immediately be seen by everyone may be 3 weeks de:Spezial:Seiten_mit_ungesichteten_Versionen. Mion (talk) 09:18, 28 December 2008 (UTC)

Or not at all, you have to wait 3 weeks.... Mion (talk) 09:21, 28 December 2008 (UTC)

If you write for a print encyclopedia, you may have to wait more than two years. If the material is proper and germane to the article in WP, it will get used. If not, the wait prevents damage to the WP reputation. I consider a maximum wait of three weeks to be rational, though I am sure many would argue that 3 days would generally be sufficient. Collect (talk) 13:50, 28 December 2008 (UTC)

I think the bad edits typically would be spotted and rejected quickly, but the good edits would take a lot longer to look over and verify are good. As for "3 days would be sufficient", I think there may be a misunderstanding. I don't think it's a time limit one imposes. If I'm not mistaken, (though the link Mion provided is restricted, so I'm not entirely sure) they're saying that the backlog can take up to 3 weeks to sort through, not that there is an imposed delay of 3 weeks. Either way, waiting 3 days or 3 weeks for one's edits to be accepted would seriously cripple the fun, efficiency, and timeliness of Wikipedia (unless my crystal ball is broken). -kotra (talk) 03:52, 30 December 2008 (UTC)

There is another scenario for 3 days or 3 weeks for both IP and registered editors, editor A has a free afternoon and decides to extend an article with four sections, 2 days later editor B also able to find the edit button decides to spent 6 hours to improve the article by extending it with similar missing information, and similar to A references the contribution, now the Sighter has to make a choice, decides, and dumps the work of editor B. Now, both editors had no intention to vandalize, but for one of them its the last contribution to this project. Mion (talk) 06:57, 30 December 2008 (UTC)

← What choice does he have to make? If I understand correctly, edits are made on top of the last "live" revision, not the last flagged one, so why would one "dump the work of editor B"? -- Army1987 – Deeds, not words. 11:08, 30 December 2008 (UTC)

The sighter is doing control of the edits of the IP editors (non trusted), registered editors (non trusted) and registered editors (trusted), if an edit is added to an article it stays hidden until a Sighter gives authorization for it, The second editor can't edit on the finished version of editor A, as that version is not published yet. So editor B starts 3 days later with the same version as editor A started with. Mion (talk) 11:29, 30 December 2008 (UTC)

Huh? I don't know how it was actually implemented on the German Wikipedia and I don't understand German enough to figure out myself, but IIRC the original proposal was to have the "article" tab at the top (the one beside "discussion", "edit this page", etc.) split in two, "stable" and "live" (except I'm not sure the proposed term was "live"), with the "stable" version being displayed by default. I don't think it was proposed that IP editors be precluded from seeing the "live" version. (If this is not the way it actually works, I might withdraw my support. How is it implemented on the German Wikipedia?) -- Army1987 – Deeds, not words. 11:55, 30 December 2008 (UTC)

..I was cheking it and your right Army1987, editor B has the option to start on top of the last contribution, still the end of the scenario is the same, if editor B dumps the contribution of editor A and adds a contribution similar in size, once the page gets sighted and goes public editor A sees a different page than he or she was working on. Mion (talk) 13:43, 30 December 2008 (UTC)

If someone tries to edit the "stable" version, we might display a big pink warning suggesting him to look at the "live" version first. Like the one when you get if you try to edit a revision other than the last one (e.g. this), but with different text, and possibly larger. -- Army1987 – Deeds, not words. 13:51, 30 December 2008 (UTC)

Further testing on DE shows some effects. The setup is as followed

1 Administrators

2 Sighter

3 Registered trusted users

4 Registered untrusted users

5 Unregistered untrusted users (IP)

I have 2000+ edits on DE, and so i'm automatically promoted to Sighter, but I only contribute content and corrections to DE, no patrolling, don't run a Cluebot, so I requested to have the Sighter status removed (de:Wikipedia_Diskussion:Gesichtete_Versionen#Sichter_abstellen , back to the former state. What happends is that my edits from now on are not shown to the public until a Sighter comes by, my edits end up in the hidden queue, you can see that on [[1]]. Mion (talk) 08:01, 1 January 2009 (UTC)

This sounds like the opposite of what we want. It leaves small, remote articles to languish without any oversight, and introduces the workload burdens on the large ones which everybody watches. — Werdna • talk 01:28, 29 January 2009 (UTC)

What's wrong with flagging

The expectations of the flagging advocates appear to be more or less the following:

1. That almost every statement in a BLP will be sourced (not only those "challenged or likely to be challenged", as WP:V says, but also such things as birthdays, which the subject is likely to care about).
2. That the sources actually exist.
3. That our text fairly represents what the source says.
4. That the source is (except in cases where we discuss both sides of a controversy) representative of consensus information on the subject. This requires reading other sources, to see if the information we have is the POV of the subject's friends - or his enemies.

This is a level of checking which FAC frequently fails to achieve. Not only should it be done once, it should be repeated for each substantive change.

Will flagging do this? How many editors will set out to check and flag articles? About 400 have bothered to !vote on the issue; counting those who have voted and will not flag, and those who will be brought in to flag, that seems a reasonable estimate of active flaggers. That's a thousand articles each; even checking and keeping up with a hundred would be a full-time job.

What will actually happen?

• Many articles will never be flagged.
• In many articles, an editor will look it over, see nothing glaringly wrong, and flag it on "Looks OK to me". If there is something wrong, from a typo to deliberate distortion, this will make it harder and slower to fix.
• Some articles will be edited from a Point of View, and then flagged. Quite often this will be sincere, since the editor is only telling the world the Truth about the rascal, or the hero. (When this is Dick Cheney or Sarah Palin, some other editor will catch it, I hope; but there may be thousands of cases where this is not caught.
• Some vandals will insert fake references and flag. They will be banned eventually, but in the meantime...
• And a few cases will actually get a super-FAC review.

Our readers will (until they learn better) assume that all articles have been thoroughly checked; but many of them won't have been.

Wikipedia is not a reliable source. Like many technical fixes, this will make us look more reliable, without actually improving our reliability significantly. This is no service to the encyclopedia. Septentrionalis PMAnderson 18:24, 2 January 2009 (UTC)

Very well said. I was kinda wavering about FR for a while, but this has convinced me that flagged is the wrong way to go. NuclearWarfare ^{contact me}_{My work} 22:53, 2 January 2009 (UTC)

This perspective entirely misinterprets what "sighting" revisions is supposed to counter. While I agree that there is always the risk that wikipedia's readers will also misinterpret what reading a "flagged" revision entails, we can't stop people believing that wikipedia is an academically-reliable source if they want to continue to do so. Remember that the absolute worst that can happen is the status quo: those "many articles" that "will never be flagged" will merely remain as they were before (not that there is much evidence that this assertion is even valid). There is no loss of reliability as a consequence, and the potential loss of the appearance of reliability is probably no bad thing.

Again, the second point neglects that the position of a sighted article containing imperfections is somehow worse than the current position. No article is perfect, and every article is being worked on continuously. Errors and imperfections will eventually be found and corrected; doing so is just as easy as at present. If attention is drawn away from the sighted revision, it is because it is being attracted on to more obvious imperfections, but assuming that every editor on wikipedia will suddenly become pure RC patrollers and never review anything other than the latest changes is entirely fallacious.

From implying that there will be nowhere near enough reviewers in point one, to jump to assume that we will give out the 'reviewer' flag so widely as to routinely give it to vandals is a rather bizzarre step. Other than that, the point yet again neglects to consider that this is exactly what happens at the moment; it's bad, and it will be found and corrected, and such editors will have their 'reviewer' flag removed with the same ease as 'rollback' flags are removed for abuse of that permission. How does this imply the end of the universe?

• Of course having an imperfect sighted article is worse than an imperfect unsighted article. If the imperfection is noticed by an anon (or anyone else without sighting) it will take longer to fix the visible version; it will have to wait for a sighter. Septentrionalis PMAnderson 19:25, 3 January 2009 (UTC)

I would be interested to see where you got the assumption that the purpose of FlaggedRevisions is to directly improve reliability. You are quite right that this is not possible; any attempt to use FlaggedRevisions to mark particularly accurate or verifiable revisions will, in my opinion, fail. Fortunately we have not seen proposals of that nature for some time; modern FlaggedRevs proposals suggest to do exactly what is proposed above: use FlaggedRevisions as a shield against vandalism, spam, libel and general nonsense, the kind of crap that we have battled against for seven years. No bot, no extension, no team of dedicated but overstretched RC patrollers, can protect our articles against POV-pushing, fake referencing, and subtle vandalism. What these things can do is spare us from that deluge of drivel, and allow the humans to spend their time where it can do the most good: writing and improving articles. The wiki environment ensures that we can never fall back further than our starting point. Tools like these can ensure that we never need to fall back at all. Happy‑melon 23:20, 2 January 2009 (UTC)

• Then it has no business in this poll, which is about how to fix BLPs. We protect BLPs against random obvious vandalism now; their special nature requires protection inobvious vandalism, such as unsourced or tendentious edits. Septentrionalis PMAnderson 19:25, 3 January 2009 (UTC)

Flagged revisions are form over substance. Take a look at anabolic steroid, a FA that had a lot of problems (bias and downright misrepresentation of sources) even after is was supposedly reviewed by a more ambitions mechanism than flagged revisions. Even though it's in the top of WP:PHARM articles visited, there are only 1-3 active editors at any given time, and there's a POV conflict going on.

Also, I don't see what's going to prevent someone with an established reviewer account from creating other accounts to make edits and then flagging his own revisions? Even if this is prevented somehow, in articles with ethnic conflicts there usually are enough participants, so it's extremely easy to create a buddy system to circumvent any effect of flagged revisions. Xasodfuih (talk) 10:48, 3 January 2009 (UTC)

I agree that we can not check all the things you describe. The German implementation is much lighter: As I understand it, you can flag an article if it does not contains blatant vandalism. Nicolas1981 (talk) 12:44, 3 January 2009 (UTC)

If it is much lighter, then it does not address the problem of BLPs at all, and the promise it will is fraudulent. Introducing a new layer of impediments to editing for the sole purpose of keeping down obvious vandalism (at which our watchlists do quite well) is pointless. Septentrionalis PMAnderson 19:16, 3 January 2009 (UTC)

You're right, thanks for the link... maybe we should restrict its use to the 10.000 most often read BLPs or something similar ? So that the lag stays inferior to 24 hours or so. Nicolas1981 (talk) 12:52, 4 January 2009 (UTC)

Straw poll on 'trial' implementation of FlaggedRevisions

The discussion on the implementation of a 'trial' configuration of FlaggedRevisions on en.wiki has now reached the 'straw poll' stage. All editors are invited to read the proposal and discussion and to participate in the straw poll. (from Wikipedia:Village pump (technical)) . Mion (talk) 18:43, 2 January 2009 (UTC)

See also WT:Flagged revisions/Trial#Just a thought... for an idea that might be useful for BLPs. Geometry guy 20:03, 5 January 2009 (UTC)

Simple Solution

It continues to astound me that "anyone can edit" is interpreted as "anyone can vandalize". Requiring registration for all editing would put an end to the problem and save hours and hours of energy.....which could then, in turn, could be used for good. So many outsiders (non-editors and others who know very little about Wikipedia) have a low opinion of Wikipedia (much to my dismay) because all they know is the reports of vandalism that appear in the media. I try to explain that vandalism is almost always quickly reverted, but it doesn't help much in their understanding and TRUST of the information within. Requiring registration is a simple solution that would work. Kmzundel (talk) 18:43, 6 January 2009 (UTC)

Perhaps I'm missing something - your plan would certainly put an end to IP vandalism (along with all other IP edits...), but I don't see where it would reduce the actual amount of vandalism by one byte. It would change the way history pages look, but it wouldn't reduce the payoff for vandalism, and it wouldn't reduce the incentive for vandalism, so it wouldn't reduce vandalism. It would reduce the volume of casual typo fixes (and grammar and style fixes) because random readers wouldn't sign up (no matter how lightweight we make it) and come back to make those fixes; they would just skip them. I don't see how it would reduce vandalism. It would reduce our reputation for openness, since we wouldn't be very open - but it wouldn't reduce vandalism. It would just shuffle it around a little bit. I don't see this as a solution. — Gavia immer (talk) 20:00, 6 January 2009 (UTC)

Requiring registration would slow down both vandalism and typo fixes. The effectiveness of this can be debated. Leon math (talk) 03:40, 7 January 2009 (UTC)

It would mean we would be able to ban vandal accounts, and if the same vandal made multiple accounts we could CU them and ban the IP; and the need to register would deter some impulse vandalism. But it would cost us much of the 90% of anon edits which are helpful, most of them minor. Septentrionalis PMAnderson 18:08, 7 January 2009 (UTC)

I have participated in an on-going similar discussion at Wikipedia talk:Main Page featured article protection and the endless debate just wears me out. The work of vandals is protected, while my good works aren't. Again I ask, why must "anyone can edit?" be interpreted to mean "anyone can valdalize"? Someone who genuinely wants to make worthwhile and positive edits does not mind registering. And, yes, registration would eliminate IP edits. Kmzundel (talk) 19:49, 7 January 2009 (UTC)

Your insistence on interpreting "anyone can edit" as "anyone can valdalize" would appear to run counter to our WP:AGF. -- The Red Pen of Doom 01:01, 8 January 2009 (UTC)

Good point Red Pen. Following that line of thinking (which I agree with), both these "solutions" to vandalism on BLPs violate WP:AGF to varying extents. Neither FRs nor semi-protection shout "we think you're trying to help" to new users. Killiondude (talk) 01:55, 8 January 2009 (UTC)

A society which relentlessly assumed AGF would have no need for police, or for that matter any sort of criminal justice system. However, it would also be out of contact with reality. On WP we have a blocking policy to deal with the reality of vandalism. (Nevermind that it also gets perverted into shutting down political dissent, when WP's foolish idea of the inevitability of consensus between reasonable people regularly fails, as it must in all human endevors). AGF was never meant to assume a priori that all human acts would be benign, anyway. It started out as a general principle that when an already completed act or communication can reasonably be interpreted in more than one way, that the way which assumes best intensions should be chosen. That's it. Don't try to make it more, because it can't be made into more. Some nastiness simply speaks for itself, res ipsa loquitur. And the past also predicts the future. Nastiness will happen. SBHarris 12:25, 10 January 2009 (UTC)

Unhelpful comment: Requiring registration? Aw, why take the fun away from huggle? :) Seriously though, we have so many editors and so many tools working to remove vandalism, it's just not that big of a deal anymore. It's a tempting thought, but there's really no need to limit the anon vandalism if it will limit the number of anon helpful edits. Vandalism just isn't that big of a threat anymore. Okiefromokla ^questions? 06:20, 11 January 2009 (UTC)

You are ill-informed, sir. But I suggest that if you think it's so small a problem as not to require much of anybody's time, you volunteer at WP:OTRS and tell them to just forward all the persistant IP BLP vandalism problems straight to you. That will make us all feel better. Let me know when you get tired of it. SBHarris 21:54, 12 January 2009 (UTC)

I'm an anti-vandalism editor (with 10,000+ anti-vandalism edits...I'll just mention again), and I'd agree with Okiefromokla, more than disagree. It's great the community is having this discussion, but unfortunately, it isn't all that productive, in the sense that there are easier, more targeted ways to address vandalism. Considerable improvements could be made by IMPROVING THE TOOLS. Specifically, I would like my tool edit 100% faster. (This would not require much effort.) That additional speed would improve my overall editing speed by 25-50%. I would like a second, new tool that automatically calls up all the other edits made by a vandalizing editor -- but only those which have not already been reverted. This second tool would address a large amount of anonymous IP vandalism that "slips through the cracks".

No change of overall Wiki policy should happen until after the anti-vandalism tools are further optimized. There's no reason to turn the anonymous IP editing community on its head. Piano non troppo (talk) 03:56, 28 January 2009 (UTC)

Why?

I don't see why biographies of living persons being vandalized is any worse than the vandalism of any other article. Overall, it's not more likely either. Too many pages are protected; there are pages that have been semi-protected just for a few cases of vandalism over the course of months. ^Tezkag₇₂ 18:43, 11 January 2009 (UTC)

If you were the subject of a Wikipedia article, would you appreciate that your biography contain vandalism or libel, knowing that it is widely visible on the net and propagates rapidly through mirrors et al ? This type of vandalism has an outstanding potential to negatively affect 'real life', which is why it is particularly harmful. Additionally, it is a constant legal threat for the project. That being said, there are indeed cases where semi-protection has been applied inappropriately or for too long a time, if you find one, request unprotection. Cenarium (Talk) 00:22, 12 January 2009 (UTC)

If I was important enough for there was a page about me, I'd watch it daily or pay someone to and any rubbish would be flushed. --Gibnews (talk) 08:54, 12 January 2009 (UTC)

There are 250,000 people you think should hire somebody to do this for them, then? Because that's about how many BLPs we have. And you don't mind that Wikipedia policy forces them to do it, if they don't want the consequences? I will AGF and assume you're not trolling us. But see [2] SBHarris 21:32, 12 January 2009 (UTC)

I know I watch pages for people I know in real life who have articles, and some have even "signed up" to a watchlist-changes-service, where I email them for any major or semi-major changes to their pages. However, beyond removing rubbish/vandalism, I don't actively edit them. I also know that many agents (ie. paid employees of famous people) watch the Wikipedia articles, on the basis that OTRS gets a stackload of agent-initiated emails. Daniel (talk) 09:01, 13 January 2009 (UTC)

Ignoring the ethical issues for the moment (which is what people like Doc & Lar appear to base their opinions on), the very real problem is that a biography of a living person brings up legal questions of libel. If someone puts defamatory material in an article about a living person who is not a celebrity, the contributor may be sued. Defamatory material has been added to articles, & the Foundation has been sued over at least one biography. Setting forth a policy on WP:BPL is an attempt to fix the problem with rules; implementing flagged versions or semi-protection is an attempt to fix it with software. This is a discussion to find the best mix of these two solutions to limit -- if not eliminate -- the problem. -- llywrch (talk) 18:05, 14 January 2009 (UTC)

If you think it's bad now, it can get worse. —Admiral Norton ^(talk) 18:08, 25 January 2009 (UTC)

Being sued is one thing, being sued successfully is another. Lawyers should make the call about whether one method or another leaves Wikipedia more vulnerable to suits. Not editors.

I agree with Tezkag72's point, that vandalism of any Wiki article makes the encyclopedia look bad. As someone who's just spent a few months on anti-vandalism patrol, quite a lot of anonymous vandalism doesn't have to do with living people, or people at all -- but, for example, popular TV shows, the towns people live in, and article topics that are school assignments.

After thousands of edits, I can discern my own editing habits -- where a snap judgment can safely be made, and where it cannot. This overall debate is too focused on a one-size-fits-all solution. A good many problems could be solved relatively simply by improving the anti-vandalism tools. That's a discussion that should happen between anti-vandalism editors, and systems programmers. When they come up with some viable alternatives, *then* is the time to present to the community. Otherwise there are far too many design details to deal with. Piano non troppo (talk) 00:26, 27 January 2009 (UTC)

Oppose

Provided that a BLP article is being supervised properly, none of this should be necessary. It would be unfair to discriminate against IP users, or to turn regular editors into full-time libel lawyers.--♦IanMacM♦ ^{(talk to me)} 10:36, 14 January 2009 (UTC)

What are you opposing, please? It is unclear to me from your post. Thanks - KillerChihuahua^?!? 00:53, 15 January 2009 (UTC)

The current BLP guidelines are sufficient as long as the article is supervised properly. Semi-protection is necessary only when vandalism is a problem. Flagged revisions would make a big deal out of BLP articles, and separate them from the rest of Wikipedia. What really matters is that nonsense and vandalism are reverted quickly, and more rules would not necessarily help.--♦IanMacM♦ ^{(talk to me)} 11:40, 15 January 2009 (UTC)

We don't have enough eyes to "properly supervise" 250,000 articles in realtime. I suppose that similarly, 3 and 4 year-olds should be allowed into deep public pools without their parents anywhere around, "so long as we have enough lifeguards." We would have no unsolved crime if we had enough police everywhere. So? SBHarris 04:03, 17 January 2009 (UTC)

As is pointed out elsewhere on this page, banning IP edits is not a magic bullet. This rarely stops the hardened vandals. Flagged revisions open up a can of worms for the user who flags the revision as "clean", including the possibility of legal liability if the article was deemed to be libellous. The current system - while admittedly not ideal - should be kept.--♦IanMacM♦ ^{(talk to me)} 11:14, 17 January 2009 (UTC)

Participation table

The Protecting BLP articles feeler survey for the implementation of Flagged revisions on the English Wikipedia started on 19 December 2008 the results, after 22 days on January 10 and after 29 days on January 17:

Feeler question	January 10	January 10	January 17	January 17
Implement Flagged Revisions for blps and Flagged revisions with expiration for all non-blp articles/content pages			support 20	oppose 8
Implement Flagged Revisions for all articles / content pages support	support 53	oppose 112	support 54	oppose 116
Implement both Semi-Protection and Flagged Revisions for all BLPs	support 5	opposed 12	support 7	opposed 14
Implement Flagged Revisions for all BLPs	support 153	opposed 87	support 161	opposed 94
Implement Semi-protection for all BLPs	support 164	opposed 171	support 187	opposed 187
I don't like any of these options but something has to be done for BLPs	support 22		support 23
Leave BLPs exactly as they are	support 44		support 52
548 unique users voted

CNET article

This CNET article from 24 January refers to the ongoing debate on this issue. The issue had been discussed before the latest BLP nonsense with Senators Edward Kennedy and Robert Byrd referred to in this Washington Post article. The tide may be turning in favour of flagged revisions to prevent this sort of nonsense, but it has been pointed out that even an experienced editor cannot guarantee that a given revision does not contain subtle nonsense or vandalism.--♦IanMacM♦ ^{(talk to me)} 17:07, 25 January 2009 (UTC)

The problem is that flagged revisions cannot prevent cases such as Kennedy's and Byrd's. Usually common sense directs us to revert the unsourced death report, but I'm not sure whether I would do it in this case. Also, the vandal might have added a source citing Kennedy's seizure and claimed it was about his sudden death. As reviewers only check for obvious vandalism and not for sources, this would be promptly sighted and create an even worse embarassment for Wikipedia. Admiral Norton ^(talk) 18:15, 25 January 2009 (UTC)

No. I would personally remove the sighting ability from anyone that doesn't check revisions. The whole point of flagging is to prevent errors from creeping in that currently pass RC patrol. If someone doesn't want to actually check the revision, that person should not flag it. Cool Hand Luke 18:32, 25 January 2009 (UTC)

Senator Edward Kennedy was declared dead in this edit by User talk:Gfdjklsdgiojksdkf who is now blocked indefinitely. Captain Obvious would have no difficulty in pointing out that a ban on IP edits would not necessarily have prevented this incident. As for flagged revisions, there is still no substitute for the normal rules of WP:V and article supervision. Although the Washington Post has knocked Wikipedia over this incident, it was in many ways a routine piece of vandalism, and was fixed in three minutes with this edit. Hardly Seigenthaler incident Part 2, it has to be said.--♦IanMacM♦ ^{(talk to me)} 18:45, 25 January 2009 (UTC)

That's true. My main concern is with lower-profile BLPs where the errors can persist for longer. Cool Hand Luke 18:49, 25 January 2009 (UTC)

Well, I've heard some very different opinions about it here and read some very different instructions about it on de.wiki. Admiral Norton ^(talk) 17:05, 26 January 2009 (UTC)

1. Oppose. IP edits are a nice colored flag for casual vandalism, making it easier to spot. I actually think that casual vandalism is a common start point for a user that eventually becomes a positive contributor (one of the reasons that the {{test}} templates are worded the way they are) and that locking users out flies against one of the fundamentally memorable and charming things about Wikipedia. Sometimes this is a price we pay for the sanity of our vandal fighters, but it should be applied as a reaction to actual vandalism, not in anticipation of it. I think the legal fears are way overblown. — brighterorange (talk) 20:30, 25 January 2009 (UTC)

It is not out of the question that Wikipedia could one day face a successful libel action over one of its BLPs. The main risk comes from the lower profile BLPs, since any nonsense about a US Senator or President will be taken off within a few minutes. What a libel court would be looking for is not what was said, but whether it could reasonably have been prevented or removed more quickly. The other risk for Wikipedia over BLPs is bad publicity, because the media loves to be able to print "Wikipedia screws it up/ is less reliable than Britannica" stories.--♦IanMacM♦ ^{(talk to me)} 20:50, 25 January 2009 (UTC)

I think everyone agrees on here that common sense dictates Wikipedia/media is not committing libel based on a user's edit. As far as I know though, this has not been tested much legally. If it has not, I'd rather this be tested legally by the most well-known user editable resource with a fairly large budget than by some little web site. If that little web site loses, Wikipedia becomes responsible for all edits and we're totally screwed. If this case happens, it needs to be Wikipedia involved. I know this sounds like we're just asking for trouble but in reality we're already under a great deal of risk as long as there isn't case law on this.--Hammy (talk) 05:16, 26 January 2009 (UTC)

It is very plausible that we could be found not liable for user's contributions, but contributory if it's found that we could've taken reasonable steps to prevent forseeable harm but didn't. WilyD 05:57, 26 January 2009 (UTC)

Disclaimer: IANAL (I am not a lawyer). Wikipedia's content is hosted under US law, so US libel law would apply if anyone got sufficiently heated to sue Wikipedia for libel. There is an excellent summary of US libel law here, and it gives these examples of what could be considered to be libel:

• A communication that imputes a serious crime involving moral turpitude or a felony.
• A communication that exposes a plaintiff to hatred.
• A communication that reflects negatively on the plaintiff's character, morality, or integrity.
• A communication that impairs the plaintiff's financial well-being.
• A communication that suggests that the plaintiff suffers from a physical or mental defect that would cause others to refrain from associating with the plaintiff.

Most importantly, the plaintiff would need to show negligence in allowing the publication of the defamatory material. This has never received a test relating to a Wikipedia article, so it is still uncharted waters. John Seigenthaler could have claimed negligence by Wikipedia in allowing defamatory material about him to remain unchecked on the site for over four months, but since he ruled out legal action we shall never know what the result might have been. Hopefully this is an area where Wikipedia never has to go to court, but everyone is agreed that great care is needed when supervising BLP articles.--♦IanMacM♦ ^{(talk to me)} 08:13, 26 January 2009 (UTC)

Actually court cases in numerous jurisdictions have shown the location of the WMF servers is only of minor relevance. People can sue the WMF for libel in their home country in a number of jurisdictions including Australia and a few European countries. Of course the fact that the WMF is incorporated in the US means that they would probably have little ability to collect against the WMF however this could also potentially prevent members of the WMF from travelling to the countries were the WMF has been found libel and potentially make problems for any local WMF chapters as well as for local contributors to the WMF. See also Dow Jones & Co. Inc. v Gutnick and Libel#Australia where some of the jurisdictions are mentioned (albeit mostly unsourced). Bear in mind there may be more where it is either unmentioned or untested. In any case, I always regard libel arguments as mostly pointless. Yes we should avoid libel but our goal, as reflected by BLP policy should be to go much further then that. Nil Einne (talk) 12:40, 27 January 2009 (UTC)

Kennedy's death was also twice inserted by long-established named accounts. I suspect that one or both of them was editing in haste, and thought that he was removing what he was inserting; but these things will happen under FR too. Septentrionalis PMAnderson 14:50, 26 January 2009 (UTC)

Media coverage

This link shows how the BBC News website is covering the debate. BTW, was Jimbo Wales appearing in a production of The Mikado when the photo in this article was taken?

The BBC article does not point out that the proposed change applies only to BLP articles, and may give the impression that the brief nonsense over the "deaths" of Edward Kennedy and Robert Byrd was a major issue, which it was not. Today, Wikipedia told its readers (for all of one minute) that Barack Obama was a member of the Nazi party in this edit. The user responsible, User talk:Labz, is now blocked indefinitely. Just another day at the office.--♦IanMacM♦ ^{(talk to me)} 17:09, 26 January 2009 (UTC)

Sweden's largest morning paper Dagens Nyheter also writes about it. —Krm500 ^{(Communicate!)} 00:00, 27 January 2009 (UTC)

There is plenty of coverage in the media today, including the Times, the Guardian and the Daily Mail.--♦IanMacM♦ ^{(talk to me)} 09:01, 27 January 2009 (UTC)

Share of Edits

Much has been made about the relative scale of the entries to be 'protected' from vandalism - both one way or the other (supporters of the FR proposal that BLP are a tiny fraction of the total EN:WP; detractors arguing that BLP is a huge fraction). Given the proclivity of EN:WP (and I suspect all wikiproject) editors to edit 'current events' or more historically recent events (again, I have no statistics on this but would encourage anyone with the data to come forward to make me look like a savant/gibbering idiot) I'd be interested to see what share of the edits this change could impact. Certainly, a large share of the current volume is Vandalism of one sort or another, but that aside I think it's universally understood that one of EN:WP's greatest strengths is the volume of edits (that is, the number of eyes reviewing information, and then actively contributing). We could be disproportionately impacting the most-reviewed articles (That is, Current Event or BLP) articles - if the share of edits is disproportionately high here, as I theorize. Sahrin (talk) 12:32, 27 January 2009 (UTC)

There are a lot of BLP which get very few edit and watchers, which as a number of BLP regulars have pointed out, is one of our chief concerns. Most current events don't (or shouldn't) involve BLPs (meaning actually biographies of living persons) anyway (rather sub articles etc) since it's usually the case that it's impossible to get an idea of the scale of something in a persons life when the event is current. When they do, the number of edits to the BLP should be kept down anyway for the aforementioned reason. I say this as a semi-regular of both ITN and BLP. A lot of our highest traffic BLPs are already semi-protected too BTW, e.g. Barack Obama and this also tends to happen a lot during an extreme surge of interest (indeed sometimes full protection is needed when disputes arise). Nil Einne (talk) 13:24, 27 January 2009 (UTC)

As someone enjoying a new Wiki anti-vandalism role, with a thousands of edits in the last months, I could produce exact stats on the changes I've made, but vandalism (by anonymous IP's) that has some motivation regarding a particular person happens frequently in: 1) That person's Wiki page, 2) that person's school page, 3) a "random" Wiki article (usually the sort that's subject to other types of vandalism), 4) that person's hometown page, 5) the page for the team that person plays with, 5) the page for the rock band that person plays with, 6) a page about the person's first or last name (in that order of frequency). All are common. The percentage that represents of overall anonymous vandalism...that's trickier...partly because there's often a question about whether edits are, for example, tests or intentional vandalism...but it's very roughly 1/3. Piano non troppo (talk) 03:06, 28 January 2009 (UTC)

This isn't going to work

This page is too big. The poll is already too convoluted to give closure to this matter, let alone the comments that have no numbering. I'm beginning to seriously question Wikipedia's ability to rationally and controllably discuss the problems that face it. The community seems to not have an effective system for managing the increasing levels of participation we've been seeing in these kinds of forums and discussions; by any standard, this page exemplifies massive over-participation. This is something that will become a much bigger problem in future, and it's something we should work towards solving. —Anonymous Dissident^Talk 08:03, 28 January 2009 (UTC)

Agree with your every word. — Realist² 08:30, 28 January 2009 (UTC)

Seconded. I'm concerned that changes to biographies are one of WP's greatest problems - they certainly undermine it in the public mind - but now I'm here and can't find where best to comment, I think Anonymous Dissident has it bang on, and that a more useful activity must be improving the way the whole massive community works together. The near-incomprehensible mess which is this page is testament to the problem (yes, I see the irony...) Peeper (talk) 13:05, 28 January 2009 (UTC)

Too big indeed, won't read it all. Just a general comment: I'm a "moderate" user since 2006 (7000+ edits, of them 3000+ in article space) currently having only a few (some 70-80) pages on the watchlist, most of them in very special categories. Some hardly get vandalized at all, others almost daily. These take up most of the time I spend editing Wikipedia. This is frustrating, and may lead to my leaving the project entirely. A few more "scandals" in the news, and Wikipedia will lose its credibility. Thus, I support measures that will prevent edits from unregistered users or new accounts becoming immediately visible, whatever the method of implementation. (end-of-rant ;-) --Janke | Talk 15:54, 28 January 2009 (UTC)

+1 zazpot (talk) 19:57, 29 January 2009 (UTC)

This page, like many of its type on WP, is indeed too big. Makes me begin to wonder just how many people there are out there who have read every single word of it? Any at all? (And in discussing this question, let's not lose sight of the fact that we're making the page even longer). ;-) Vitaminman (talk) 20:00, 29 January 2009 (UTC)

While I think you have a point (I could hear my internet connection groaning while this page downloaded); I disagree that wikipedia is unable to 'rationally and controllably discuss' problems. I think a lot of work need be done to improve the methodology, but most of that simply requires applying the 'wikipedia philosophy' to discussion. It's more than a little ironic that in attempting to correct what amounts to a few bad apples 'shouting during the play' editors duplicate the same behaviour. Wikipedia as a system is designed to organize and democratize the effort of presenting and preserving information - and yet as of yet we don't possess that ability to discuss the issues related to our ultimate designs. I definitely think the means of settling disputed demands an overhaul; but consider that the English Edition is just not emerging from it's 'pioneer' stage of development - the independent and ruthless lawman (the admin) who has the ability and power to serve and judge and jury in any dispute is no longer and effective solution to our problems. But again, I don't think we lack the ability to organize our discussions - the template is before our very eyes; we need only excercise it. Sahrin (talk) 05:17, 30 January 2009 (UTC)

I have a suggestion that might help organise debates such as this with lots of participation: We create three pages "/arguments for", "/arguments against" and "/other information". Instead of everyone putting in their views in forum type order, one could try and summarise the various arguments in the best possible/most logical way. This presentation could be discussed on the respective talk pages. I think this would help people see what is happening, and what the issues are and how they relate to another more easily. Also the strength of the arguments would be what you got from the page, not a head count.

It would have the added advantage here that one could arrange arguments to a range of related proposals at the same time: General arguments could be given for and against flagging, and then a small summary about how these arguments relate to specific proposals for implementation. Thehalfone (talk) 10:10, 30 January 2009 (UTC)

• My thoughts are that if we are going to resort to a straight poll, we may as well take advantage of the software and ask people to vote through their preferences or something for major discussions like this. A lot less mess and lot clearer results for poll situations. Whether everybody is entirely happy with the usage of polls is, unfortunately, a matter that the otherwise brilliant MediaWiki interface cannot help us with. —Anonymous Dissident^Talk 15:17, 31 January 2009 (UTC)

"Mass Over-Participation" is a problem most other websites and projects would KILL to have. There is an ever-stronger elitist trend here which worries me. Let people participate, and lets celebrate the fact that they care enough to do so.Riversider (talk) 17:57, 31 January 2009 (UTC)

Celebrating the fact that we have 500 people joining in isn't, unfortunately, going to solve the BLP mess. —Anonymous Dissident^Talk 23:42, 31 January 2009 (UTC)

1. Oppose This is a BAD idea, Wikipedia will stop working if we start doing things like this RP459 (talk) 18:03, 31 January 2009 (UTC)

Agree, at 665 kilobytes the page is now too long and would have been archived several times if it were an article talk page. The argument of overparticipation is perhaps not the best way of criticising this debate. Participation is a good thing, although most of the important pro and anti points have been raised already, and there is little need to repeat them. The proposed changes to BLP articles would be the biggest change in Wikipedia's history, so a consultation exercise was necessary.--♦IanMacM♦ ^{(talk to me)} 15:57, 1 February 2009 (UTC)

If this survey was closed after a week, two tops, it would have been somewhat manageable and we could move on to the next step (whatever that would be, even doing nothing). Instead, it's gone on for almost 5 weeks and the "This is a simple survey" in the intro has become a joke. I think it's good that we've gotten so much participation from the community, but if we don't draw any formal conclusions from it and move forward, it's a waste of everyone's time. Though the chance of anyone doing that now (or at least, doing it properly), is probably nil. -kotra (talk) 08:10, 2 February 2009 (UTC)

Y'know what? There's no consensus on this issue whatsoever, so how are we going to make this work? Why not just keep the status quo? — Rickyrab | Talk 18:46, 6 February 2009 (UTC)

Maybe that's the only conclusion anyone can objectively make: no consensus. However, I suspect that an extremely limited flagged revision trial might garner a rough consensus. Some have been proposed at Wikipedia:Flagged revisions/Trial/Proposed trials, but as far as I know, none have been formally put to the community at large for discussion/vote. -kotra (talk) 19:11, 6 February 2009 (UTC)

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.