Wednesday, October 17, 2007
Mediawiki spam and abuses
I am running a website using Mediawiki engine. All was going well, until I started to get my pages spammed with links about pills. They were not too many so I was removing them manually (using rollback) and then ban the IP. But after some weeks, I started to get too many to do this manually. They were probably some bots. I looked into config files and found $wgSpamRegex variable. I simply added some keywords into this and it prevented further spam from those bots. But later, some other bots came with other keywords. They had random like keywords so I could no longer use the $wgSpamRegex technique. Then I learned about Spam Blacklist Extension and I found this helpful forum http://www.mwusers.com/forums. I also found this: Blocking Spam in Meediawiki. With these resources, now it should be much easier to maintain a wiki site. Something I simply don't understand, it is why some of the spam bots were simply adding some random characters in my pages. No links, no advertising, nothing else. Just a few random characters like: coauier or afoilta. Are they some kind of probes and then after they finish the scanning the real spam will arrive?