This page is for suggesting entries to add to the list on BannedContent.
Note that when despamming runs, your suggestion is reverted, since it is now considered to be spam.
Do you want to volunteer as an administrator with the right to edit the BannedContent page? Alternatively, consider joining the Banned Content Network with a page I can feed back into our own list.
I suggest we put on the regex "adult\S?friend" (i.e. "adultfriend" and "adult-friend" are caught, but not "adult friend"). this would take care of things like http://adultfriendfinder.adultfriendfindernow*com.
I'm posting it here first before adding it because I'm proposing banning a speech pattern, rather than just a specific domain.
I also propose adding "freewebpage\.net", "pathfinder\.gr", and "adult.?cams".
Remember that people can easily defeat the regex if they just want to talk about the site (although maybe only the nerds will know that this is possible); the regex only really blocks linking.
I am greatly concerned by the inclusiveness of some of the regular expressions used to ban domains. For example, if sex.com were banned like most of the banned domains, it probably would appear as sex\.com on the list. However such an expression would match content like sussex.com, sex.comprehensive-doctors-advice.org or sussex.company-maps.co.uk. I recommend that when banning domains the regular expressions should look something like \b([\w\-.]+\.)?example\.com\b instead of example\.com.
Along those lines, I've compiled a list of 5488 domains used by wiki spammers. Of these domains, 4063 are not matched by any of the current regular expressions on the BannedContent list. Are you interested in me submitting a few thousand regular expressions that match these additional domains used by wiki spammers?
I may be missing something obvious, but why not just this?
Chris, since a period counts as a word boundary I definitely agree that your suggestion would work. The only reason I suggested the slightly more complex expression is because it produces what I think is nicer "edit denied" feedback in anti-spam systems that identify the offending text. However, I wouldn't argue with someone who preferred your simpler expression. The most important part of my suggestion is the part that you've noted, namely that expressions intended to match domain names specify word boundaries at both ends.
The only thing I worry about is that I also merge lists form the Moin Master Bad Content list, and from the MT Blacklist. This means the old regular expressions would reappear automatically, unless I automatically process them (much like I do right now when escaping dots and adding a leading space. Maybe it would make more sense to keep things simple and unban the occasional false hit. I also remember at least one occasion where I had "stupid spam" – people just added domains to a page without linking them. Somebody told me that perhaps Google took those into account as well for pagerank. I don't really care, but I felt that we should ban this kind of crap, too. This is why I changed back from the more complex regular expressions I used back then: Back then I prefixed every domain with
As for submitting the domains you have, do you have them available on the web somewhere? If so, I could just merge them in on a regular basis. This might make more sense if you intend to maintain this list on your own site.
There's a shared banned-content page used by Wikimedia at http://meta.wikimedia.org/wiki/Spam_blacklist . It'd probably be a good idea to work out some kind of synchronization with them. I'm working on a banned-content extension for MediaWiki based on the code we already use for Wikitravel.
Suggestion: secretporn69\.com . Just got hit with this one today; see http://wikitravel.org/wiki/en/index.php?title=New_Zealand&diff=74870&oldid=74756 .
Define external redirect: RichardParker