External links policy/Archive1

This page serves as a place to discuss the current Spam blacklist policies and guidelines, and for the communities to work together on improving the usefulness and use of this tool.

cs.wpEdit

A recent addition of the freehosts wz.cz, mujweb.cz, sweb.cz and xf.cz to the spam blacklist by Amgine raised much debate on the Czech Wikipedia, as the pages hosted by these websites are frequently linked in the Czech articles. According to this list, more than 800 external links on the Czech Wikipedia has been blacklisted. This makes editing of the influenced articles significantly harder, as the user must first check what's wrong, delete the (in most of the cases useful) link and then make the contribution. This confused several users who were not aware of the spam blacklist.

The users have been informed about the purpose of blacklisting the website, which brings positive results for the small Wikipedias other than Czech (as they were being spammed with links to subdomains of aforementioned freehosts), but is quite bad for the cs.wiki: about seven hundred articles, 2,5% of the total, can't be now edited without removing useful links, new links can't be added and some essential tasks (eg. marking pages as copy vio or adding image sources to their descriptions) are made significantly harder.

Some suggestions of solving this problem have been made:

  1. Not include the http:// part in the links - this would make them "unclickable", but not blocked by the blacklist; there have been some complaints about this suggestion on cs.wiki, arguing it is not "clean" and it would make the links harder to use for the reader; however, implementing this suggestion is better than not keeping the links at all and is probably the only one not needing the developers' intervention
  2. Switch off the spambot extension on the Czech wiki (as a temporary solution, until one of the other ones is used; arguments for this solution are based on ability of Czech sysops to deal with possible spambots)
  3. Local whitelists (would be the best solution possible; however, it may take some time to implement this feature)
  4. Less restrictive regular expressions - that is, limiting the spam blacklist to exploited subdomains of the freehost instead of blocking the entire site; this has been discussed with Amgine and considered not very effective, because one of the spamming bots used dozens of generated subdomains, all of which redirected to the desired page

After some discussion on the Czech Wikipedia IRC channel, switching off the spambot extension until local whitelists are made was decided to be the best possible solution. However, no attempts to contact the developers on this issue have been made so far. --Timichal 21:10, 7 June 2006 (UTC)

OpinionEdit

  1. That is a temporary workaround, not a solution.
  2. The question is, how many bots are there bouncing on Wikimedia servers?
  3. It depends, see bugzilla:6321.
  4. That is a „solution“ that seems so blatantly correct to me that I just cannot understand why it has not been already done. But since it hasn't, it obviously has not been accepted, so that there is no point in raising it again.

--Mormegil 19:56, 19 June 2006 (UTC)

WhitelistEdit

A local whitelist is now available. Patterns can be added to MediaWiki:Spam-whitelist on each wiki by it own admins; matching URLs will be excluded from the blacklist matching. --brion 20:15, 22 June 2006 (UTC)

Maybe someone can explain why first blacklisting and then whitlisting again should be better than just do nothing? -- 88.76.209.112 20:12, 10 January 2007 (UTC)
The blacklist is over all wikis using it... the whitelists are specific to each wiki. Cheers! Eagle 101 22:52, 2 February 2007 (UTC)

VirusEdit

Is the Spam-blacklist the appropriate mechanism to block a site that attempts to infect computers that view the site with computer viruses? --Gerry Ashton 14:35, 8 May 2007 (UTC)

Absolutely. Please report this site along with evidence of spamming ASAP. MaxSem 14:41, 8 May 2007 (UTC)
I have already done so; see the diff of my report. Note that I was not logged in at the time I made the report. --Gerry Ashton 15:11, 8 May 2007 (UTC)

questions about "unlisted guidelines"Edit

It is being argued here[[1]] that a site should be added/remain on the Spam Blacklist not because it sends spam, but that "it doesn't add anything useful." The same calim is made elseware throughout the talk[2] as well. The single guideline spicifically says:

  • Only blacklist for widespread, unmanageable spam.

So I, looking to remove racetotheright.com from the blacklist (reasoning listed here[3]) was wondering if this could be clairified. Is the spam blacklist, which only blacklists for widespread unmanageable spam, also blacklist things that passes opinion wise as "not useful"?--Zeeboid 19:45, 30 May 2007 (UTC)

Referring linksEdit

Some sites offer a "referring" program, in which they give certain advantages to users who redirect other users there. These links have no place at all in Wikipedia, as there is always a non-referring link that can be used, one that would not exploit the good faith of casual readers.

It is common in the English Wikipedia to see links to Play-Asia, an importer business which gives a small percentage of an operation to referral. These links have the format *.play-asia.com/SOap*. Whenever a casual reader clicks on the link, he is directed to the product page, and if he buys it, the owner of the referral link will get a percentage of the sale. I propose banning any referral link. Right now, the Play-Asia is the most obvious one. Amazon links would be interesting too, however they don't have a fixed format. Thoughts? -- ReyBrujo 00:19, 9 July 2007 (UTC)

The Spam blacklist is used by a multitude of wikis, both Wikimedia projects (like Wikisource) and external wikis (like Wikia). I suspect that many communities using the blacklist would oppose prohibiting referral links, which are useful tools for generating funds. However, The English Wikipedia can easily blacklist links violating local policy with w:MediaWiki:Spam-blacklist. —{admin} Pathoschild 03:14:40, 10 July 2007 (UTC)