Announcement

Collapse
No announcement yet.

Reporting spambots.

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Originally posted by droste View Post
    Then a human solves the captcha and the bot continues its work. So *no* captcha can stop this, because the task of the captcha is to identify humans and humans are the one who solve them.
    Already answered in previous post. But first step is to update to better captcha to verify its not automated. If its manual proceed to next step of making registration(and ban) cost higher than any profit.

    Comment


    • Originally posted by crazycheese View Post
      Already answered in previous post. But first step is to update to better captcha to verify its not automated. If its manual proceed to next step of making registration(and ban) cost higher than any profit.
      Yes, I didn't want to say that the switch to recaptcha is a bad idea. I like the idea behind it (digitize books) and it also has the audio option for visual impaired people. I just wanted to say, that it probably won't solve the problem :-) But I don't like the idea of having a entrance fee to post here.

      Comment


      • Originally posted by droste View Post
        Yes, I didn't want to say that the switch to recaptcha is a bad idea. I like the idea behind it (digitize books) and it also has the audio option for visual impaired people. I just wanted to say, that it probably won't solve the problem :-) But I don't like the idea of having a entrance fee to post here.
        But you do understand why you need fee?
        Another way is invitation. But thats about it, I have no other ideas...

        Perhaps,.. community verification with verification history is good way..

        For example, someone registers here and posts.
        If he is newbie and posts are ok, then existing members vote for his trust. This vote is recorded on those who vote as well.
        For example, newbie needs 25 points to acknowledge he is human. Members vote only on meaningful posts and only once on post. The members can also do opposite - mark him as spammer with one click.
        Once he gains needed amount - he is a member and limitations are lifted off.

        Lets assume he is reality is human spammer, automatic spammers are banned fast by method above.
        When he starts spamming, he is reported here in this thread as spammer to forum moderator. We already have much less spammers with this method.

        The moderator will also see those who voted him up and decide on their status, for example if they upvote same persons that turned around to be spammers. This decision however is best let to human with good logic.

        What do you think about it?

        Comment


        • Originally posted by crazycheese View Post
          But you do understand why you need fee?
          Another way is invitation. But thats about it, I have no other ideas...
          I understand that this would be a way to handle this, but I don't think it's a good way. But I also don't have other/better ideas.

          Originally posted by crazycheese View Post
          Perhaps,.. community verification with verification history is good way..
          [...]
          What do you think about it?
          All in all it comes to "need more moderators". There's already a feature to report posts, but someone has to handle these reports.

          Comment


          • Originally posted by droste View Post
            I understand that this would be a way to handle this, but I don't think it's a good way. But I also don't have other/better ideas.
            Me either...

            Originally posted by droste View Post
            All in all it comes to "need more moderators". There's already a feature to report posts, but someone has to handle these reports.
            It comes down to amount of spam.
            The hardest spam is human-made.
            Autoregistrations can be prevented with good captcha. Thats 99% of volume.
            %1 will have to rely on CAE to pass this barrier and this will cost spammer 0.01$ already for each bot.

            Now, if he is able to generate more profit than loss, we loose. Otherwise its not economical for him and we win.
            The money system will damage him by monetary side - it instantly charges him with 1$, total 1.01$.
            If he does not generate more than 1$ profit, he is at loss.
            Downside - yes, many people may not have the money for one-time registration, so it might be difficult.

            The trust system approaches from different side - by increasing the moderator attention hundredfold.
            It will require him to post something that makes sense. Otherwise his links will not work.
            The voting by already trusted members will resolve "need more moderators" problem. All established users are moderators.
            If some user misuses his right and downvotes the good post - he will be traceable. Same way, if group of already established users upvote a spammer, they will be traceable.
            This method will allow to dispatch whole wine grape of spammers instead fighting the singular cases.
            The downside is that implementing this method is complex investment.

            But the third option - have more moderators, makes no sense.
            Not many people are happy to generate useless posts and report him.
            Michael himself does not have so much time.
            And hiring an army of anti-spam moderators is also useless.

            Not discussing IP block here, because many tunnels are used by valid and spammer users alike.

            My two cents.

            Comment


            • Cynt55hia

              The usual crap...

              Comment


              • Jesus7Clark with images.

                Comment


                • Robert932: hidden images

                  Comment


                  • MashFrant: sports livestream link spam

                    Comment


                    • flake50

                      Obviously a bot and necro champion.

                      Comment

                      Working...
                      X