[Mediawiki-l] spam attack avoids captcha

Platonides Platonides at gmail.com
Thu Mar 17 21:06:52 UTC 2011


Daniel Friesen wrote:
> The spammers haven't coded the bots to handle QuestyCaptcha yet, but if 
> people start using it to stop them, then they will code it into the bot. 

Questy captcha is a free form question. You can't learn how to bypass
any instance of it. Not even a human would be able to bypass it using
just generic knowledge (eg. a foreign speaker).
The most that bots could do is to try with common answers such as a the
wiki name, or domain. No bot would be able to solve all by itself "Who
is the current companion of Doctor Who?"*
OTOH it's highly annoying for those users which don't know about Doctor
Who, or that for some reason get their answer rejected. I was once
denied access by one captcha of that kind (naming the character of the
image) despite having found the correct answer, I don't remember exactly
what was the issue. So you may also want to provide some email address
for appealing.


*Once a human instructs it, they are able to use it multiple times at
that wiki, though.




More information about the MediaWiki-l mailing list