joshua.zelinsky(a)yale.edu wrote: We remove all kinds of links, sources, and text that, some folks argue, help articles. For example, we remove links to resellers of recreational vehicles, which doesn't harm the articles greatly much less deface them. We remove links to websites that charge for access, we remove links to most foreign language sites, etc. Each of those removals may remove some value, but the articles are actually improved by making the remaining material stand out more, free from the clutter of spam links, poor sources, and fringe material. Just like pruning a fruit tree or a rose bush can make it healthier, more productive, and more attractive. Will

Will Beback wrote: It's not hard at all, Will. If you'd like to see how easy it is, go propose that we remove all external links that aren't sources. Or even just external links to sites where the subjects of articles self-publish. Or forget proposing. Just go do it. You will find an ocean of people willing to tell you those links do indeed belong in the articles. And once you have discovered that they are all in the encyclopedia for good encyclopedic reasons, then perhaps you'll have an easier time accepting that we are not willing to compromise the encyclopedia's content. Especially for something where even you admit the benefit is arguable, and where some of us argue that there is harm, not benefit at all. William -- William Pietri <william(a)scissor.com> http://en.wikipedia.org/wiki/User:William_Pietri

Quoting Will Beback <will.beback.1(a)gmail.com>om>: This is a strawman. The point is that we shouldn't be removing links from an article unless those links are somehow damaging to the content. The distinction between a random blog or a spam link to buy cars and Michael Moore's personal website should be obvious.

Quoting Will Beback <will.beback.1(a)gmail.com>om>: The problem with spam links at all is that even a few of them make people less likely to trust whether the external links provided are useful. And speaking frankly, one of the first things I try to do when learning about a controversial topic is see what external links Wikipedia has. I doubt I'm the only one who does that. As long as the page which happens to have an attack on the Wikipedian is a relevant external link it is better and doing less damage to the article than a link about buying cars or a random blog. And the vast majority, most likely all, the damage from harassing links will occur whether or not we link to the website. The end result of this also is to remove more and more to prevent harassement. For example, if someone keeps harassing an editor until the person's article is deleted, do we delete it? No, not any more than we would if the person in question had politely asked for their article to be removed as one of borderline notability. Nor do we make convenience alterations and remove pertinent information of notable people simply to stop their little campaigns on Wikipedia. We do remove information when the sourcing is questionable, but that's basically it. And there's no substantial difference changing that policy whether we change it for birthdates, external links, sourced criticism or anything else.

Quoting Will Beback <will.beback.1(a)gmail.com>om>: Yes, but letting our personal goals get in the way isn't a good thing. "Somewhat relevant" may be too low a standard here. If we have another link that could just as well go in the section that isn't there obviously we should replace it. For many articles we have many good links about the same issues and we can't include them all. In those cases, when we have no other way of deciding, I don't see anything wrong with deciding based on harrassement concerns. However, we're not talking about somewhat relevant links, we're talking about links to the official sites of the subjects. I'm not misinformed. I should have added something like " content that is well-sourced but the individuals are not major public figures and would likely get deleted if it went to AfD, or they are such major public figures that we need to bow to them, minor well-sourced details about public figures that were not widely reported and maybe a handful of other circumstances" I've not done OTRS work, but in the times I've been peripherally or incidentally involved with OTRS, the reasoning for removals has generally been good, and when the person was genuinely notable a replacement article was eventually created that was often very similar to the original content. And if we're really deleting much more than that we have much more serious problems than the external links policy.

Steve Summit wrote: I think we need a new example. The Michael Moore issue us barely even real harassment. Here's the story of "user:Joe". This is a true, recent story, but with the information changed because it was indeed harassment. Let's call the editor in good standing "Joe" and the the issue cockfighting, though it wasn't. It's a topic with an eager and engaged group of proponents. There's a website called "RoosterChat" and Wikipedia had an article on it (that was eventually deleted as non-notable). Joe took some actions and edits to correct the NPOV of articles on the topic and became the target of that community's ire. They posted messages on their chatboard with his personal information, including phone number, and a request that board members call him at work to convey their feelings. The immediate situation was handled through diplomacy, but in the long term the editor was forced to reduce his involvement in the topic and take other steps to restore his privacy. How should a policy deal with this situation? Should we maintain our link to the chatboard (which could only used because it was the subject of the article). Should we link to the harassment as an example of that community's activism? Should we tell valued editor that the link is more important than his privacy or well-being? Will

David Gerard wrote: I asked first. What kind of policy or procedure do we have or should we have to deal with a situation like this? Will

RLS wrote: Then what do you propose we do in these situations. While this is a specific example, this is the more common problem then the MM/THF example. Will

Will Beback wrote: My view: We should not alter article content one iota in response to external badgering. The only thing that has changed is our view of them, not the NPOV view of them. Ergo, we behave exactly as if they were harassing Editors in good standing should be able to link to the harassment to the extent that they believe it serves some legitimate purpose in furtherance of Wikipedia's mission. If they are linking for some other reason (e.g., or gossip or furthering the harassment) they should be dealt with through our usual mechanisms for miscreants and the clue-deficient. We should tell the editor that we hope they understand that articles should not be affected in any way by internet drama. To the extent that the editor wants to take defensive legal action (like getting a restraining order) we should support them. And we should encourage them to ask for a fellow editor to take over maintaining the article(s) in question, hopefully from a pool of people with thick skins and diplomatic skills. And in my opinion, we should always reach out to the offended party or community to at least make nice, and hopefully to work with them to the extent possible. A vast amount of this drama comes when all sides believe that we should have a great encyclopedia, and they just have a different idea of what that means than us. Often that's just because we've thought about it for years and they've thought about it for minutes. William -- William Pietri <william(a)scissor.com> http://en.wikipedia.org/wiki/User:William_Pietri

William Pietri wrote: Why would we support an editor in getting a restraining order, but not support them by doing any restraining ourselves? What form would this support take? So you're saying that if an editor is harassed by an outside group then the editor (and Wikipedia) should give in to that harassment. And you think that won't affect the POV of a topic? If a group succeeds in driving off one editor after another, how many thick-skinned editors are there willing to take their places? Let's think about it long enough to come up with a formal policy or procedure to handle it. You make some good suggestions, but mail-list posting don't write policies. Will

RLS wrote: Claiming the moral high ground is not part of the mission of Wikipedia. If someone else wants to claim it then fine,. let them. The purpose of Wikipedia is to create a free, neutral encyclopedia. Removing links does indeed make harassment harder to find. Retaining links pushes the harassment into the harassed editor's face. I'd hope that the policy we draft doesn't draw attention to the removal process. Since we remove hundreds or thousands of links a day that shouldn't be a problem. I agree that all means of eliminating or minimizing harassment should be used, including contacting the webmaster/harasser, calling the police, and writing the carrier. That suite of tools should also include taking down the link from Wikipedia, a top-10 internet site. Official channels can have a high threshold for action, can take a long time to act, and can even cost money. While they should be pursued if necessary, there are simpler ways of reducing the impact of harassment and they start with us. W.

joshua.zelinsky(a)yale.edu wrote: If Wikipedia becomes known as a volunteer job that leads to off-site harassment that the community will do nothing to stop then that might tend to reduce the appeal of editing too. I again dispute that removing poor sources harms NPOV. Will

David Gerard wrote: I believe that when an editor on Wikipedia gets into a dispute and uses their blog as a weapon in that dispute, then that the blog is no longer a suitable source for Wikipedia. We have plenty of good ways to resolve personal or content disputes between editors. We should do everything we can to discourage disgruntled editors from using outside websites to affect Wikipedia content and POV. This thread is concerned with future proposals, not past errors. If you want to discuss the past we can do so in a private exchange. Will

When considering a proposal we should always look at how the proposal would effect past actions. While Dave may be making this come across as a bit too personal the Making Lights example is an excellent test case. The fact that if your proposal had been in play all the links would have been removed should give us pause.

Jimmy Wales wrote: I don't see this being a reason even in extreme cases. If the person in question posts a death threat or libel in some other completely unrelated blog posting, how does that in any way affect the fact that he "vigorously disputed the allegations put forward by the New York Times" in the original blog posting being used as a reference? That's already covered by existing guidelines. The procedure when a reference URL changes to something other than the referenced text is to try to find the old version in archive.org, or otherwise "fix" it. If it can't be found anywhere, though, we still don't remove the reference but rather record the date that the original link was found to be inactive; even inactive, it still records the sources that were used, and it is possible hard copies of such references may exist, or alternatively that the page will turn up in the near future in the Internet Archive, which deliberately lags by six months or more. <http://en.wikipedia.org/wiki/WP:CITE#What_to_do_when_a_reference_link_.22goes_dead.22> I suppose if the "dead" link goes to something actively misleading instead of just a standard 404 page one should deactivate the link, though, to make sure readers notice the fact that it's not the material that was originally referenced.

On 26/10/2007, Charlotte Webb &lt;charlottethewebb(a)gmail.com&gt; wrote: Mmm. We get quite a few of these (and if we're unlucky, it's squatted by a pornsite). Is there any practical way of spidering through our links to check for these? -- - Andrew Gray andrew.gray(a)dunelm.org.uk

On 10/27/07, William Pietri &lt;william(a)scissor.com&gt; wrote: Well, I do know that many "whois" sites will deny service to the IP of anything that looks or smells like an automated tool. Of course, there are ways around that, as we all know. —C.W.

Quoting Charlotte Webb &lt;charlottethewebb(a)gmail.com&gt;om>: Yes, I think we remove that sort of thing routinely anyways. I've certainly never hesitated to remove them.

Quoting Bryan Derksen &lt;bryan.derksen(a)shaw.ca&gt;ca>: Er yes. Bad phrasing on my part.

On 10/19/07, Will Beback &lt;will.beback.1(a)gmail.com&gt; wrote: Dude. Nobody's arguing against removing a poor source. We're arguing against removing valid, useful sources just because the same site contains harassment of an editor. THAT will cause us to violate NPOV. How you're coming across is that you want us to pull all links to http://www.example.com/aaaa.html, http://www.example.com/bbbb.html, http://www.example.com/cccc.html, and http://www.example.com/dddd.html, just because http://www.example.com/zzzz.html contains some verbal harassment of an editor. There is no reason to remove the links to aaaa through dddd if those pages haven't changed and they're still relevant to the articles they're linked from. We just don't need to link to zzzz.html, and it's hardly likely to be a useful source for an article anyway, unless we're writing a neutral encyclopedic article *about* the harassment, and that isn't very likely either as it probably won't pass WP:NOTE. When someone at example.com posts their harassment at zzzz, we need to NOT freak out and remove all the other good links to example.com. That's what will indeed draw undue attention, and compromise the articles about aaaa, bbbb, cccc, and dddd unnecessarily. And yes, I'd like to know where you get the idea that removing links to aaaa, bbbb, cccc, and dddd will make the harassment at zzzz harder to find, or accomplishes anything else useful. If example.com changes their pages at aaaa, bbbb, cccc, and dddd to compromise their content and replace those pages with harassing content, then yes, the links to those pages need to be removed. The content is no longer a reliable source, editorially speaking. In a case where the actual content remains valid, but they add verbal harassment to the bottom of those pages or something like that, then that bridge needs to be crossed ***IF WE EVER GET THERE.*** It is *not* something that we can make a value judgment on, or write a policy on, here and now using hypothetical examples. It's something that we as a community need to look at if it ever happens. The actual harassment, its severity, its location and placement on the pages, and the value of the sources to the articles all need to be taken into consideration. We cannot write a black and white policy on this, without agreeing in advance to compromise our editorial principles. I don't know about you, but I'm not willing to do that. --Darkwind

On 10/20/07, Jimmy Wales &lt;jwales(a)wikia.com&gt; wrote: I'm not sure why it has to be a discussion about websites at all. If someone is going around adding links to ED, then I have to wonder what the .. they are doing on Wikipedia in the first place. It seems much more promising to me to look at the substance of someone's actions, rather than at the patterns of URLs. -- Toward Peace, Love & Progress: Erik DISCLAIMER: This message does not represent an official position of the Wikimedia Foundation or its Board of Trustees.

Jimmy Wales schreef: They may be a minority, but most of the recent discussion was about borderline cases. The Nielsen Hayden blog, the WikipediaReview Signpost article, the Michael Moore site. Perhaps you don't hink these are borderline, but in each of these cases I've seen people arguing on both sides. Eugene

On 10/19/07, Jimmy Wales &lt;jwales(a)wikia.com&gt; wrote: I agree with your position that these actions were overboard; the problem is, they were both done, they were both actively supported by a noticable fraction of active admins/senior editors at the time they were removed, and it was not clear to anyone (objecting to the removals or neutral) what policy actually had to say about it. During the runup and early bits of the Moore removal, I objected to removing the links, but I couldn't tell if restoring them and warning the removers was the right thing to do, or a blockable offense I'd be committing. So I made some ANI comments and sat on my hands. -- -george william herbert george.herbert(a)gmail.com

On 19/10/2007, Eugene van der Pijll &lt;eugene(a)vanderpijll.nl&gt; wrote: And whether naming antisocialmedia.net in [[Judd Bagley]] should count as a personal attack on the people attacked by that site, even though the site itself is named openly in the NYT etc. as relevant. There was an arbitration case about this. The problem is the cases in the middle. What overrides NPOV? - d.

On 10/19/07, Jimmy Wales &lt;jwales(a)wikia.com&gt; wrote: That's basically what I've been trying to say throughout my several messages on this matter. The clear-cut cases are already well-covered by existing policy and guidelines--link the material if it's a good source, which typically won't include sites lacking enough in editorial control to become harassing; and don't link to the sites which aren't good sources. With that point assumed, therefore (logically) the editors who continue to bring up this matter must be arguing for a policy to deal with the borderline cases; I keep trying to point out that we can't do that in advance as it will really depend on the value to the encyclopedia of the good material linked, the harm any alleged harassment will/has cause(d), the actual placement of harassing material on the pages, etc etc. Regarding your numbered points, you're definitely right on #1. However, what I percieve from Will Beback and others who keep bringing this up is that they're trying to lean towards a position far closer to #2 than I (and many others) are comfortable with. --Darkwind

On Fri, 19 Oct 2007, Jimmy Wales wrote: The cases of interest that have turned up are such things as Michael Moore's page, which isn't a harassment site through and through. But one of the big problems happens when the harassment site *isn't being used as a reference* (for anything except its own content). One of the most notorious examples is that in the original BADSITES proposal, people couldn't even point to any examples of why attack site links might be useful because the attack site links got deleted as per the very policy that was being discussed. That's an incredible catch-22. There was also a case where a user was criticized for posting on an attack site and tried to give links to the attack site to demonstrate that his posts were innocuous. They were deleted.

encyclopedia" to remove don't only There is a (?unrecognized?) blurring of poor quality source vs. undesirable source, here, that's part of the reason this thread is so messy. Distinguishing the two will help a lot. A * poor quality * source is one that makes statements that are unlikely to be reliable (or where there is no clear case they are reliable), and hence cannot meet [[WP:RS]]. Self publication, blogs, and the like, are archetypical "poor quality sources". (By contrast in simple terms, a good quality source would be one that is reliable, or does credibly testify to certain matters or their genuine standing.) An * undesirable * source is one that we do not wish to promote, link to, or encourage, usually because it actively is engaged in harming us, or an editor, or contains text which pointedly attacks or harasses members of the community, etc. (Be aware the decision to "not wish to promote or link" is inherently non neutral - it always serves an agenda and perspective of one side.) Now... there is usually no problem with poor quality sources. We have no use for them so we never have to link to them. Also, so long as an undesirable sources is also a poor quality source, we have no problem or conflict either. Same reason, there is no good reason to link to such a source. The concern comes, what happens when a site hosts both good quality material, and * also * undesirable material. That is the heart of this issue, and this precision is needed to examine it. In some cases we can replace the source with another, so we don't have an issue. This would usually be the preferred answer anyway. So ultimately, there are precisely 2 cases left to consider: 1/ The entire site is basically an "undesirable source", and really, the entire site, or its aim, is undesirable to link to. But for some matters, or in some articles, it is also a good quality or valuable source, and has especial reference value in those limited areas. 2/ The site has one (or a limited few) pages which are "undesirable", but the rest of its pages, or the aim of the site in general, is not really a problem per se. But for some matters in some articles it is also a good quality or valuable source, and has especial reference value in those limited areas. These are the specific two cases we need a decision on -- on the basis "if this situation did exist for some site, how would we handle it?" I think looking at it this way is far more likely to be more constructive, and lead to a more helpful analysis, than just arguing the question, is it an "attack site", whatever that may mean. FT2.

William Pietri wrote: At the risk of prolonging a discussion that's already rather egregiously long in the tooth, what do people think about those? One of the larger objections to BADSITES and its resurrection in WP:NPA#EL concerned the prohibition of links to attack sites even in talk and project spaces, even as part of sober discussion of the alleged personal attacks. Is there still any sentiment to keep trying to ban those kinds of links?

(Again, not making policy here, just engaging as an equal in a hopefully constructive discussion...) Steve Summit wrote: Yes, but we have to be careful and thoughtful. One thing I have seen recently is violations of No Personal Attacks through the posting of links to abusive websites. I will make a hypothetical example. A problematic editor who has a history of trouble with others and the Arbcom responds to something perfectly appropriate said by a well known Wikipedia admin, in the following fashion. "Anyone who wants to understand what is going on here, should simply read [http://www.someattacksite.com/Admin_name.html this link]." I think statements like that are an "incorporation by reference" of the material contained on the attack site, and should be removed immediately (and probably the problematic editor blocked at that point). --Jimbo

Actually, no. It's not off topic. Harassment policy, to the extent it is valid, is based on No Personal Attacks. Fred

Quoting Fred Bauder &lt;fredbaud(a)fairpoint.net&gt;et>: Fred, I think you are misintepreting what he said. The point was that even without any BADSITES policy this sort of thing is so blatant that anyone would consider it a personal attack. It is a blatant example of using the link to make the attack.

Quoting Will Beback &lt;will.beback.1(a)gmail.com&gt;om>: Will, as many editors have pointed out to you, we aren't saying that we shouldn't do anything. The objection is solely to the removal of links that would otherwise definitely be in an article. Again, we've already explained how these aren't "poor sources" and multiple editors have already discussed how this harms NPOV. Simply saying "I again dispute X" doesn't make the arguments for X any weaker.

joshua.zelinsky(a)yale.edu wrote: People won't want to join if it becomes known that participating in Wikipedia can lead to serious harassment at home and work, and that the community will do nothing to support its volunteers when that happens. I dispute the contention that removing poor sources harms NPOV. Will

Will Beback wrote: Citation needed. Citation needed.

On 10/19/07, David Goodman &lt;dgoodmanny(a)gmail.com&gt; wrote: They probably realized, as we do, that the notorious Everyking thread on WR didn't contain any forms of harassment, and reached the (not unreasonable) conclusion that there is sometimes a good reason to link to an "attack site". Either that or none of these considerations even came to mind as the committee announced its unprecedentedly urgent decision (and of course nobody saw any ethical dilemma in poring through the discussion threads of said attack site to collect specific information, whether or not they considered the remainder of it to be morally filthy-dirty). —C.W.

Quoting Will Beback &lt;will.beback.1(a)gmail.com&gt;om>: Um, he just said. Getting a restaining order. And yes, it isn't our job to do that, especially when removing a valid link in an article hardly adds restraint. No, but the editor may wish to have others help out or take the flack temporarily. I've done this before for editors who have not as thick skin as I do, and generally after things die down the editor goes back to that article. This isn't giving in at all. Giving in would be leaving and never coming back. It is merely a strategic switching of roles. *rest snipped*

joshua.zelinsky(a)yale.edu wrote: I take issue with what I believe your saying. What we're talking here about are disputes that are due to encyclopedia editing. Yes, there are some folks with pre-existing battles who come to Wikipedia to settle scores, which never should have come to WP to begin with. That's nt what we're talking about. The cases that are the most shocking are those that originate with editors enforcing Wikipedia rules. When folks are doing good they shouldn't be targetted for harassment by POV pushers. When they are we should stand by them. We lose the moral highground and compromise NPOV when we say to folks that are trying to change Wikipedia content that they are free to harass those who enforce Wikipedia policies, and editors who don't like it need to go edit elsewhere so the POV pushers can have their way. W.

Will Beback wrote: I never suggested otherwise. You said removing external links from articles was not a big deal. I suggested you go try doing that with a bunch of articles. Just removing ones like the link to the article subject's website. You will quickly have a host of people willing to explain in detail why it is a big deal -- right after they have reverted you. Ergo, the links do have value. I can't tell if you really have the notion that I would suggest that. If you don't, then it's a rhetorical device that tells me that we won't make any more progress here. If you do, then I have obviously failed to convey to you pretty much anything that matters to me. In which case I also think future discussion is unlikely to be fruitful. Sorry. William -- William Pietri &lt;william(a)scissor.com&gt; http://en.wikipedia.org/wiki/User:William_Pietri