David on 7/10/2008 at 17:50
Absolutely not.
Besides, it's actually extremely trivial to programatically 'screenshot' a webpage, even if it is never actually being displayed on a screen.
MorbusG on 7/10/2008 at 20:31
How about a white-list approach?
If your ip isn't on that list, you wouldn't get an answer at all from the server.
To get onto the list you'd need a static ip-address, send in drivers license scan, and donate 10EUR to TTLG. :joke:
Then there are services like akismeth but I dunno.
Jenesis on 7/10/2008 at 21:17
How about displaying several CAPTCHAs and adding some text that says 'write the letters from box (random number) in the text box below'? The scripts might well just grab the first CAPTCHA they see, so if the one you ask for is never the first one, the bots would all fail even with a human solving the CAPTCHAs for them, because they'd be solving the wrong ones.
EZ-52 on 7/10/2008 at 21:40
What about something along the lines of "the sum of the first three BLUE numbers", a sort of three way barrier?
Or is that to complicated for a simple CAPTCHA System?
On a similar note, my websites forum had an overdose of spammers recently, so much so that the hoster has disabled my site until I can get a more sophisticated CAPTCHA system in place (we used just a basic type the letters in the picture system).
raevol on 8/10/2008 at 01:56
You do have the .ru and .rus TLDs banned right? That'll stop most of your headache.
mOdEtWo on 9/10/2008 at 08:18
Quote Posted by MorbusG
Then there are services like akismeth but I dunno.
...akis
meth :joke:
I use (
http://akismet.com/) Akismet on the BCG Forums (vBulletin supports it in 3.7) and even tho we do have a problem with spambots signing up, most of the posted spam is classified as spam after a quick test by Akismet, and then put into the moderation queue. This means that the spam posts are never displayed publically, and it is easy for us to find the spam and delete it, and meanwhile ban the spambot. It doesn't catch all spam, but I guess close to 90% or so. I have yet to see a false positive.
I suggest you look into this, David. Akismet and vBulletin 3.7.x (which introduces the (
http://www.vbulletin.com/forum/showthread.php?p=1450764#post1450764) Delete as Spam feature) might help a lot. Akismet costs $5/month for a commercial license.
Avalon on 13/10/2008 at 00:27
A forum I run got overrun by spam at about the same time TTLG did.
For me, the problem was as easy to fix as bumping up the image used in image verification to "SEIZURE MODE." I checked all the options to add extra triangles and colors and all kinds of crazy shit to the image. I'm guessing a new script or whatever the hell was found that worked around the previous image setting.
Muzman on 30/10/2008 at 05:56
If the accounts are being manually verified how come these guys keep getting in?
Is it because the Man behind the 'ual-ing can't easily tell who's legit either? (I mean no slight. I wouldn't have any idea). Or are a lot of them sort of human assisted bots?
David on 30/10/2008 at 06:34
Sometimes I just can't tell. There are over 100 sign ups a day and I'm rejecting well over 90%
Sometimes the IP addresses or email addresses or usernames have never been used for spamming before and we get to be the guinea pig. There are several databases on t'Interweb for tracking this information and the spammers we are on the receiving end of get added to these databases.
Scots Taffer on 30/10/2008 at 06:45
I've noticed from one of my blogs that these guys often come in off referral pages that are just targets for potential spammers, can you track that as part of their sign-up process?
