Spam on amarok.kde.org

Miguel Angel Alvarez maacruz at gmail.com
Wed Jul 9 21:35:19 UTC 2008


El Miércoles 09 Julio 2008, David C. Manuelda escribió:
> I think this is not enough, as most of spammers are real humans. I.E. I
> have a website protected by a temporal number image, and 3 or 4 people
> registered to spam. So the only one explanation is that they are humans!
A captcha is designed to tell apart humans from machines, not spammers from 
more decent people. To better defeat spammers you should play with patience 
or effort, for example, waiting a minute to send the message wouldn't be a 
problem for anybody, except for the spammer who wants to send the most in the 
least time.

> On Wednesday 09 July 2008 15:47:06 Amara Emerson wrote:
> > We need to find a system which asks the user philosophical questions
> > that human beings can answer, but botmakers cannot. There may be a
> > final solution yet.
> >
> > Amara
That would not be possible, because then only a human could interpret if the 
answer is right or not, and the captcha must be processed by the server.
A good captcha is something which is very hard to machines to solve, but 
really easy for humans. Visual captchas are an example, but unfortunately 
many of them have been restrited to just read characters, which is a problem 
simple enough nowadays to be solved with machines.
A better aproach could be, for example, presenting 3 photos with 2 having some 
object in common, and ask the user which object it is, or even better, 
highlight by painting the object in the images.


-- 
Don't see the world through a window, be open{source}minded, and be free :-)



More information about the Amarok mailing list