I, for one welcome my robot overlords. I'm just tired of bots spamming my site with pizza of cheese variety
sage is not a downvote. I'd say you're here for summer/just switched from 4chan.
no, no. I'm writing my own implementation for my own imageboard. I'm trying to think of tests the browser's Javascript engine (along with however many successful posts) can do to help me determine.
Le mao, counter tripfagging :^) it's secretly a ploy to get people to try 314chan out, don't tell anyone
yes that was the intent
sage
We got them to think milk and the okay hand sign were racist. Why couldn't we?
Don't you have TRS set to Mud & Ruts?
Didn't want to use a trip wtf
This is a bait thread, but what the hell. Maybe some newfags might benefit.
You're a faggot and you need to die for tripfagging, but basically bots can't into logic. I see anti-spam measures as a red queen race. Anything you do will only last for so long, so for it to be reasonably effective you'd have to change the measures every so often. So the question is, what can humans easily do that bots can't? I wanted to say, "follow instructions", but honestly most humans can't do that very well either. If you put too much of a barrier in their way, it will function the same way as a lock for them too. So it has to be effortless.
Bots = scripts = marginally more illiterate than most niggers. Scripts only do what they're written to do. They can't improvise. Decent AIs could maybe manage some limited interpretation of written instructions, but those aren't your targets. You're filtering the script kiddie tier shit. So you could do several things:
Finaly uuuu h4v3 a raisin to t4lk liek this, you ginormous faggot.
One example of what you could do that might work for awhile is use javascript to recolor the words that are part of the instructions after the user performs a certain action. If the script you write is obtuse enough, that would probably defeat most bots. "Hit these buttons in a particular order to show the instructions."
There's plenty of them out there you could replicate. Even with a half way decent logic puzzle game, I doubt most Unsolicited Communication Technicians™ could be bothered to waste the time working out the logic to bypass the unique spam barrier of a single site. Unless of course you get the traffic to merit that kind of special attention.
Here's an idea. Maybe you could initially instruct the user, in a weird way, to type an unspecified word that matches a category. The case-insensitive word could then be checked against a custom encoded json list containing words from that category to see if the input demonstrates conceptual grasp of the grammatically erroneous, but still comprehensible instruction. If it was me, I'd break the encoding key into several fragments and place it throughout the script in unintuitive ways for the sake of buying the encoding some shelf life.
If the input passes, then you could display a series of emojis. Some of the emojis could be rotated to various angles in some obscure fashion, possibly server-side. Then you'd ask the user to pick the emoji that matches the answer to a question that requires inference.
Obviously the user needs to select all of the smiling or laughing emojis. This has the added benefit of filtering out people who hate babies, like communists and marxist sympathizers.
Tell dddudeman to fix his site.
Just provision CloudFlare on your website.
onlineaspect.com
To prevent spam make sure you follow these points (quoted from sitepoint):