User talk:XLinkBot/RevertList/archives/October 2012

Archives

• Started the archives w/instruct ect.--Hu12 (talk) 13:56, 8 February 2008 (UTC)

Searchable archive

Please use the archive template found on most/all noticeboards, which has the nice benefit of being searchable. Belorn (talk) 01:08, 21 October 2012 (UTC)

Entries with no log entry and no discussion

• onion: Linksearch en (insource) - meta - de - fr - simple - wikt:en - wikt:fr • Spamcheck • MER-C X-wiki • gs • Reports: Links on en - COIBot - COIBot-Local • Discussions: tracked - advanced - RSN • COIBot-Link, Local, & XWiki Reports - Wikipedia: en - fr - de • Google: search • meta • Domain: domaintools • AboutUs.com

Went here to find what argument/reasoning there was behind the entry of .onion, but sadly, neither a discussion or a log entry seems to exist. This also makes it half impossible to see who added it (only option is going through the history and narrow down the edit). For transparency, would it not be good if discussion + log entry was an requirement and not just an suggestion when new entries are added? Belorn (talk) 01:08, 21 October 2012 (UTC)

Curious, there should be log entries for everything (at the very, very least who added it so you can ask). I'll have a look if I can find that.

Regarding .onion - I think I added that at some point; .onion sites fail important parts of our guidelines, namely that they are not accessible by everyone, they need a special piece of software installed. Moreover, they were at some point inappropriately pushed to certain pages, and some of them are unquestionably questionable in nature (hey, CNN is not going to be behind a tor-site ..). --Dirk Beetstra ^{T C} 03:55, 21 October 2012 (UTC)

I've indeed added it during this discussion: MediaWiki_talk:Spam-blacklist/archives/March_2011#Hidden_wiki - it was at that time removed over and over, and only added by hit-and-run IPs. I hope this explains. If you have further questions, please don't hesitate to contact me. --Dirk Beetstra ^{T C} 04:00, 21 October 2012 (UTC)

Thank you. For those that are used to scripting, it might be useful to search through the current list and see if there are others without an log entry. Anyway, back to the subject of .onion, I was wondering if other sites beyond the hidden wiki site had been used in a unappropriated manner. Are there any method to see how many times xlinkbot reverted on a particular rule? COIBot link report is nice, but last entry is 2010. (just to mention, this is not an request for removal. I want to figure out how the situation looks before making up my mind). Again, thanks for a quick response and fixed log entry. Belorn (talk) 09:51, 21 October 2012 (UTC)

The XLinkBot logs are just his reverts, I don't keep any other logs (except system logs, which I wipe if 'nothing has gone wrong' (technically wrong)). The database from COIBot is 'new', it restarted beginning of this year, the old data is gone (I had to change server). So there will not be much there. Something may come out of 'user talkpages which have a .onion domain linked on them' (XLinkBot leaves the reverted link in a warning, mainly for tracking purposes): Special:LinkSearch/*.onion (though that misses the '.onion.to' which was pushed.

For .onion, I remain that it fails WP:EL, but not bad enough for a blacklist (though, just about ..). XLinkBot only reverts new editors, and IPs. And XLinkBot will not revert on a undo-type action by the same editor (other editors may however revert again). So it does not exclude links, it just warns that the link that was added is generally unsuitable as an external link and reverts the edit once. Also, it ignores older editors so they can be inserted by them without the bot worrying them. So if an IP gets reverted twice by the bot on the same page, that must mean that at least the editor added the link 3 times, and that it was once reverted by the bot and once by a human editor. It is then almost sure that it is not properly discussed on the talkpage. The bot will of course revert every addition if they are to different pages, but by the time that an editor on .onion comes then to three warnings in a short time, it certainly looks like he is spamming the link.

What I am worried about, is that you don't know what is behind .onion. Surely there is good material there, but the encoded link does not show you what you are linked to. It is in a way a redirect site, and the code can very well point to child porn, as well as to pictures of rose fields. In that way, the risk is the same as linking to flickr images (which are also on the revertlist). I think therefore that it is fair to have it on the revertlist, so that the user adding it is warned, and that it is also brought to the attention of other editors, and to the attention of recent edit patrollers. Quite some are linked in mainspace, which does suggest that consensus has pointed that way (even when I can find editors who have been reverted for adding the .onion links, some have also been blocked after one addition, which suggests that that addition was abusive). --Dirk Beetstra ^{T C} 10:51, 21 October 2012 (UTC)

By the way, I have meta-blacklisted the .onion.to, which is not the official site, but a redirect to it (or better, a link to a portal for it). I first blacklisted '.onion', but I reverted that just now). If anything there should be linked, it is the .onion, and I suggest that an established editor does that (maybe after discussion, since at the moment there is a tag that the site should not be linked). --Dirk Beetstra ^{T C} 10:55, 21 October 2012 (UTC)

I dont know if one need to worry that much about the url encoding. All url's has the ability to link to child porn as well as to pictures of rose fields. An URL is only meant to provide a method of human memorization, where a string of numbers are harder to memorize. Redirect site on other hand is a different matter, since it proxy for the request. Its the difference between a address written in a foreign language which one can not read, and getting directions from a English guide. In theory, both could lead one to the wrong place, as the person following either the sign or the guide has no power to "know" if the place is safe. Thus in the end, all one can do is to memorize and associate a bad address with a bad experience, like one can do with the hidden wiki. Anyway, my goal was to find out the reason why, and what events if any caused of the listing. I am still not sure if there has been any other site than the hidden wiki which has caused problems, so please see this as a small request for a feature in xlinkbot that provides data on problematic domains :). Given that the hidden wiki has been spammed by IP's, (but not often, say 3-5 times in the last 3 years if my memory serves me right), a revert entry to resolve that issue sound appropriate. Its a bit broad for my taste, but it does resolve the issue at hand, and one can always narrow the definition in the future if needed. Belorn (talk) 14:00, 21 October 2012 (UTC)