So, I got this page started as a preventive measure. Most of these ideas have already been used in the past, and are probably second nature to more experienced Wikitravellers. But I wanted to make sure we had an explanation of the ideas, and also a failsafe mechanism in place in case SoftSecurity fails us at some point.
I realize that other wikis, like Wikipedia, employ hard security measures like user bans more often than we do (that is, not at all -- yet). I'd like to continue to rely on friendly, open corrections as long as possible. We don't have enough contributors to start making enemies out of potential ones -- however slight the potential may be.
As usual, comments welcome. --Evan 01:43, 6 Feb 2004 (EST)
I'm wondering if by mentioning the fork option, if we're not making it more legitimate/likely. Eric Raymond argues that there is a great social pressure against forking open source projects, and so it is very rare. I'm not sure how well "open source psychology" transfers to a Wiki though. So I'm wondering if we shouldn't mention it at all? I know this reasoning is kind of flaky, but I thought I'd bring it up anyway. The other side of this is that maybe by pretending that certain issues don't exist (by not mentioning them) then we would jeopardize the strength of the Wikitravel community. --Dawnview 00:34, 9 Feb 2004 (EST)
I agree that forking is probably is a last resort, and finding some kind of accommodation is probably the best way to deal with policy problems. But personally I think it's worse to have catastrophic policy battles than to have forks. I'd rather see groups of editors go different ways based on policy differences than have all the editors just give up on Wikitravel in disgust at political infighting.
As to whether the pressure not to fork carries over to Wiki: I think some, but not as much as you'd think. There's a lot of examples of forks in wikis, like Wikinfo and Enciclopedia Libre from Wikipedia, or CommunityWiki and FermentWiki from MeatballWiki (although there's dispute whether these last cases are "true" forks).
As to whether or not to mention it: well, the main reason I set up this page was to discuss when and how we implement user bans (a feature of the MediaWiki software). It's something I never want us to have to do, but I'd rather we were ready for it when the time came than trying to scramble and come up with some procedure in the middle of a divisive period. I kinda feel the same way about forking: it's better to work out how we deal with it when it's not an issue than scramble to deal with it later. --Evan 01:52, 9 Feb 2004 (EST)
I notice that there is some (chinese) WikiSpam being posted. If you are not already aware of the site http://spammers.chongqed.org/ is documenting cases of WikiSpam so that the google pageranking of keywords is defeated. Rather than just deleting WikiSpam, persistant WikiSpam should be chongqed by being submitted and buried completely. Evan - good to see ROBOTS.TXT disallows search engines access to the page history. -- Huttite 06:48, 1 Oct 2004 (EDT)
As written, this policy puts all user-ban decisions in the hands of admins. That's contrary to the general spirit of policies here where we consider admin power to be merely a janitorial tool. I don't really have a better suggestion, but it's kinda odd. -- Colin 03:54, 5 Mar 2005 (EST)
Problems with this policy
This policy fails to adequately address childish but persistant vandalism. While the general procedure works, the step from "it's your problem to revert everything" to "full user ban" is too severe a jump. Wikipedia handles persistant vandals with warnings, and then a short 24 hour ban. The 24 hour ban just makes the vandal go away. Sure they can retry the next day, but that's not as annoying since they can be re-tempbanned. What is annoying is to follow a vandal around and fix the problems for no reason other that to preserve our reputation as "never banned nobody." Additionally, even if we have someone playing babysitter and following the vandal around, the recent changes page and all the history pages get so spammed as to be unusable. I propose the following:
1. Add more admins so that more of our users have single-click revert buttons to ease the task of revertion.
2. Allow 24-hour bans for a user who deliberately vandalize 10 or more pages within a 24 hour period, and who vandalizes 5 or more pages after being warned. By deliberately, I mean that they intend to cause damage, and are not merely users trying to help or even just someone trying to promote their site in extlinks.
Personally, I'm no longer interested in cleaning up after persistant vandals without more policy in my favor. This just increases the burden on others, which risks having them burn out too. -- Colin 15:20, 17 Mar 2005 (EST)
I made a patch against the version of Mediawiki which we run that allows an admin to revert everything that shows up on a given user contributions page. This allows restriction by time, so that contributions by an earlier user of the same IP would not be effected.
With this patch in place there is absolutely no reason whatever to block an entire group of users sitting behind a given IP for any length of time. Think about it: Blocking is what the vandal wants us to do. Heck this particular one spent most of his first (and longest) vandalism run asking me to block the IP, probably so that he could send a letter to the editor of his local student paper about what weeners we were for blocking the whole school. Well, I didn't do it. -- Mark 15:46, 17 Mar 2005 (EST)
I can already get the effect of your patch by going to the user contributions page, finding all (top) pages, and middle clicking the revert next to them. Don't get me wrong -- your patch is useful for undoing large scale foo (though I hope you limit the reversion so it doesn't kill contribs by previous users of the ip address). But the problem I'm trying to address is the continuing and ongoing vandalism which happens after you hit revert-all.
The script you wrote did handle the ongoing vandalism problem. But I have concerns with it: 1. your script was in effect no different from a user ban (except the reverts could be reverted if they were good contribs) 2. your script doesn't solve the messy-history problem and 3. if your script was distributed openly and used by vandals, we would have no choice but to do ip-blocks. -- Colin 16:00, 17 Mar 2005 (EST)
So after the recent User:OIdPine episode, should we add a case allowing blocks for deceptive user names? — Ravikiran 02:45, 26 July 2006 (EDT)
While I didn't think so when I first started editing here, I now tend to agree that blocks are counter-productive. I don't buy into the idea that a vandal will eventually become a useful contributor, but it does seem that anyone vandalizing the site gets bored and leaves when no one pays any attention to them. For example, if I immediately revert someone's vandalism they really are more likely to stick around, but when it goes unnoticed for a while they seem to wander off quickly. Given that, I'd say blocks aren't going to be productive - it's attention, and if anything encourages someone to be more creative in how they abuse the site.
Long story short, I don't think any policy that calls for blocks will be effective, and that even a deceptive user name isn't a worthwhile reason to block someone. -- Ryan 02:52, 26 July 2006 (EDT)
During today's reverting fun Colin asked if someone could start a discussion about handling trolls. Trolls are annoying and it's hard to ignore them, but I think it would be helpful if we could follow some of the advice outlined in WikiPedia:Wikipedia:What is a troll#Not feeding the trolls, particularly the section on "the value of slow reverts" - I don't think we need to leave offending content for more than a few minutes, but an immediate revert simply encourages further vandalism. The benefit of taking a "don't feed the trolls" approach is:
Someone who is trolling is looking for attention, and by immediately reverting any changes they make we are giving them something to do.
It doesn't really hurt us to have random gibberish displayed on a few articles for a while, and for important pages such the Main Page we can always add temporary protection to make sure that visitors to the site aren't confused.
I don't think it's worthwhile trying to enforce any kind of policy for handling trolls, but it might be good if we could discuss the best way to deal with them when they show up. -- Ryan 01:56, 6 February 2007 (EST)
I don't like the approach. I wonder if having your hard work reverted in seconds by a team of 20+ users is a bit of a downer and causes them to wander off. Anywho, if you look at User:Sapphire/Sandbox#IP/Vandals I've figured out the vandal's IP address and we could prevent it from creating more user names, which gives the vandal the power to move all these pages around. Otherwise send an email to the ISP asking for their help. -- Andrew H. (Sapphire) 13:28, 6 February 2007 (EST)
While I think most trolls quickly wander away after half an hour of being quickly reverted, I do think that some are not put off by that. Usually the pattern with an actual vandal (not just an unhappy editor) is that they fight the revert awhile, then they start attacking your talk page, then they give up. But not every vandal fits this profile. -- Colin 13:40, 6 February 2007 (EST)
Congrats on tracking down the IP address, Andrew; the only long-term solution to all this is to hit the idiots where they live. This said, I don't think it's particularly useful to try to get inside the heads of the people that do this vandalism. The real question is: what policy is best at meeting our goals? Given that none of us (except Evan and Maj) do this for a living, and we have other demands on our time (I just happened to be home sick today -- no big deal, thanks -- and wouldn't have been able to help parry the attack otherwise), it seems to me that we just have to do what we can, when we can. Simply adding some of the pages most frequently targeted by "real" troublemakers -- Talk:Main Page, etc. -- to our watchlists should let us catch the more serious vandalism when it occurs, without trying to implement a policy that doesn't meet the goal, won't necessarily work, and is incompatible with the way most of us do things here. -- Bill-on-the-Hill 13:42, 6 February 2007 (EST)
In response to Andrew, bear in mind that whether it's on a wiki or on the local street corner, someone causing vandalism is generally doing so solely to get some sort of response from the people whose work is being vandalized. If there's no response, there's no payoff - it's more fun to play games when you're playing against someone. -- Ryan 15:24, 6 February 2007 (EST)
Well, here's a chance to do the experiment. Guess who's back. -- Bill-on-the-Hill 19:25, 6 February 2007 (EST)
Just registered to this specific wiki, I had never seen such huge scale of 'edit attack' on other wikis. I must say I do not understand why someone would sit in front of his computer to spend his time to do such thing repeatedly, life can be boring, the weather can be bad but how does someone arrives at such low life level?
Those who make unwanted edits come in all shapes and sizes I guess... Special:Contributions/126.96.36.199 / Special:Contributions/188.8.131.52 has been a particularly persistent case, checking in several times a day to add apparently bogus info. Given a few weeks I expect he/she will go away, but in the mean time we've got well over a hundred edit/reverts in the article history. I somehow doubt this will be the only time we see this kind of behavior, so does anyone have any suggestions on how best to handle it in the future? The options I can see are:
Try to engage the user. This user doesn't seem to care about policies or consensus, and based on the talk pages for this IP on Wikipedia I suspect discussion probably won't work.
Use the spam filter. This has been tried, and the user simply changed the edit text to find ways around the patterns being used.
Ban the user. Not a great precedent to set, and bans are easy to get around.
Protect the article. Again, not a great solution as it prevents others from making edits.
Use slow reverts, and simply wait this person out. Not a great solution as it makes a mess of the article history.
Ideas or comments? -- Ryan 22:54, 24 February 2007 (EST)
I'll considered comment removed. -- Sapphire • (Talk) • 00:19, 25 February 2007 (EST)
If you take that approach I'd recommend staying away from the Myer Centre, as I hear it can be dangerous... -- Ryan 23:17, 24 February 2007 (EST)
I'm concerned this contributor may actually be mentally ill. His persistence rules out being a troll. Being poisoned three times implies that the person went back to Myer Centre after being poisoned twice, which does not sound rational. So seriously, I think we need to go gently on this one -- either slow reverts or a temp ban to redirect his energies elsewhere. -- Colin 00:11, 25 February 2007 (EST)
When you say temp ban do you mean blocking the IP for a week or two or protecting the page? -- Sapphire • (Talk) • 00:15, 25 February 2007 (EST)
Have a look at the edit history on Wikipedia (here and here). Granted, it's a shared IP address, but the random edits to high school articles fits the pattern of a bored kid making stuff up. In any case, no matter what this particular user's story is, do we have any better options than slow reverts or a temp ban? What if we just leave the offending content on the article for a few days/weeks since someone will obviously remove it in the future? If the consensus is to use a temp ban I wouldn't be opposed, although I really don't think bans are effective, and often have the opposite effect of making successful and notable vandalism into a challenge for someone to pursue. -- Ryan 01:22, 25 February 2007 (EST)
I was drink spiked, so I stayed away from where I was drink spiked, I was then gassed, so I stayed away from where I was gassed, then I was food spiked, so I eventually stopped going to the centre since it is crime ridden. I was spiked twice with GHB and gassed with hydrogen sulfide. I consider it disgusting that there are cameras all around and security did nothing about these incidents! So I'd rather warn people to stay away then go through what I have!
I'm having trouble figuring out how to cough the word psycho on a keyboard, so forgive my bluntness... – cacahuatetalk 02:05, 30 March 2007 (EDT)
Why wonder how to cough it? I agree the people who spiked and gassed me ARE psychos! I cannot understand how this centre is allowed to operate while it is so crime ridden! If they cannot guarantee the safety of visitors it should be shut down permanently! Also, your wondering about my mental state of health? I'm wondering about why security thought it was okay to let someone be attacked three times, I was actually POISONED twice and gassed once. It is a huge place and was hoping it was just bad luck, three times confirmed it wasn't. I am definately not mentally ill.
Did you file a police report and if so what's the case number (or whatever it's called)? I'll send an email off to the Brisbane PD and check to see if your story can actually be collaborated. Then all this will be settled. -- Sapphire • (Talk) • 10:51, 1 April 2007 (EDT)
I told them but was not given any case numbers,
What happened to you after these alleged incidents? Be very specific, as if you were telling the Queensland Police. -- Sapphire • (Talk) • 21:22, 1 April 2007 (EDT)
Also, when did this occur? Specific dates, please. -- Sapphire • (Talk) • 21:32, 1 April 2007 (EDT)
I became nauseous, dizzy and sick. I also don't have dates of the incidents.
I searched through the Queensland Police Department's records and I can't find anything similar to the story you keep telling. There has only been one recorded incident of a poisoning at the ball, but the person who did the reported the poisoning was the same person who placed the poison in the food. She's been busted by the PD. I have to say, I'm not buying the story and will work with the others to revert your addition of the story, unless you come up with clear proof it happened. -- Sapphire • (Talk) • 00:46, 24 May 2007 (EDT)
I don't care if you buy it or not, also, who cares about someone being poisoned at a ball, I was poisoned in the myer centre. Clear proof it happened? Not needed, due to the negligence of the myer centre security staff testing wasn't carried out. It did occur and a complaint was made, that's all that matters.
The following is listed as policy, but seems to be ignored by most admins, myself included:
However, if a script is badly-behaved -- due to programming error or malicious intent -- an administrator can and should put a user ban on the IP address and/or user account the script is using. Again, the administrator should note the ban on the user ban nominations page, and the same procedure applies as for other bans.
I don't see the value in listing every IP for every script that hits the site on the ban page. If someone wants to ban a user it should go through the normal process, but do we really need an additional level of bureaucracy when fifty articles get hit with links to sex-with-ponies.com and an admin blocks the offending IP? Can this sentence be removed? -- Ryan(talk) 01:21, 31 March 2007 (EDT)
Removing SPAM from the history?
Is it worthwhile to delete then immediately restore the last version of a page to avoid leaving the newly added spam in the article's history? I'm kind of for this idea, but if others are too, it would be nice to tweek the restore process a little to make it easier...
As of now you have to "check" all the boxes for the revisions you want to restore. If there's a lot of history, then that means a lot of clicking on each and every check box. So, can we somehow get them all to be checked by default, or add a "select all" button, so that we can then just "uncheck" the most recent (spam) edit, and thus quickly delete/restore pages? – cacahuatetalk 00:30, 24 May 2007 (EDT)
There's no need to delete the histories for pages that have been spammed because Robots.txt tells search engines not to look at the histories or older versions, which effectively prevents spam urls from being indexed by sites like Google.
I was going to tell you, especially since you had to restore over 500 (600?) versions of the "Europe" article, but I wasn't 100% sure you were trying to remove it from the history. -- Sapphire • (Talk) • 00:43, 24 May 2007 (EDT)
Oh, I guess I haven't paid enough attention, but I thought that was their whole point was so that it would up their google rating. Well good news if what you say is true. And yes, I was trying to get it out of the history... and stupidly I didn't actually check the history first, cause if I had I would have noticed that the last several edits were the same thing, so I only deleted the very last one, leaving several. Because sometimes I'm lame. So I guess I can't thing of another reason we would ever need to delete and restore. But if we do think of something, my request above would help a lot. I'm gonna stop talking now. ok. bye! – cacahuatetalk 02:44, 24 May 2007 (EDT)
Challenges to policy?
I am uncomfortable with listing "challenges to policy" under "unwanted edits." It is my understanding that challenges to policy are beneficial to the sited and are very much "wanted edits," unless we are just feeling lazy. I do, however, think it should remain on the page under a different subheading—in order to indicate explicitly that these types of edits are welcome on Wikitravel. --PeterTalk 15:50, 28 June 2007 (EDT)
Perhaps a better way to describe that section is "Defiance of policy". Speaking for myself, I don't want someone who disagrees with an established policy making edits that defy it; I want them to first change that policy (if they can), so that their subsequent edits are in agreement with policy. - Todd VerBeek 20:35, 28 June 2007 (EDT)
I like that thought. I just tried to tweak the section to highlight the difference between defiance of and challenges to policy, but am not very confident I did a great job with it. --PeterTalk 21:29, 28 June 2007 (EDT)
Being one of frequent challengers of the policy, I believe the "policy on changing policies" should be more accessible and found more easily than it's now ("How to handle unwanted edits" is the last thing I would think of when I'm looking for how the policy is set up and changed). Things like "when I have something not fitting the current policy, should I discuss first, or make a practical precedent, or go ahead and change the policy itself first?" are very non-obvious now, and I would vote for having them explicitly defined, along with what's currently described in "Defiance of policy" section. --DenisYurkin 08:03, 1 July 2007 (EDT)
I'm not sure advertising is necessary. The way it usually works is that a new user makes an edit X, it's rolled back, the user complains, and he's told that edit X goes against policy Y and he should raise the issue on Y's talk page. Jpatokal 13:20, 12 July 2007 (EDT)
Making something implicit to be described explicitly in a place we can easy refer to is not necessary to advertise it ;-) --DenisYurkin 02:48, 13 July 2007 (EDT)
handling repeated adding of agencies and extlinks
We frequently revert adding of agencies and extlinks; as WT grows, more and more people revert. I think it's time to prepare a template that anyone can use on user's talk page to address anonymous or registered contributor who insists on adding his agency or extlink. Who can share your examples of such messages so we compile them in something anyone can use instantly? --DenisYurkin 16:20, 23 September 2007 (EDT)