Will see, but I expect this won’t be air-tight, since the spammers manage to register anyway, and there is already a captcha on registration. There are tools for solving captchas cheaply, but not for posting good comments.
“X can’t be that difficult to implement”: something one probably shouldn’t say about an open source project unless one is about to give it a go themselves.
There’s are at least two opportunities to suffer motivation drain in your comment :)
“X can’t be that difficult to implement”: something one probably shouldn’t say about an open source project unless one is about to give it a go themselves.
By writing that captcha was implemented because it was easy to implement, you suggest that easy-to-implement property has nontrivial explanatory power. I disagree, since I believe this property also holds for other alternatives, so can’t explain the choice, and shouldn’t be used as an argument. It’s just a matter of rationalist nitpicking.
(It’s more difficult for me, because I don’t know the project (or python, for that matter), but must be trivial for people who do. In any case, the intended meaning, as applied to Karma limit, is the same as with captcha in the sentence that originated the exchange, so it refers to how difficult it would be to implement the Karma limit (instead) for the person who implemented captcha.)
I accept your criticism and will stop nitpicking. What we did was particularly “easy” because we only turned that captcha on and styled it. Captcha-on-submit was already there in the Reddit code.
We’ll try something else to reduce the impact of spam soon.
A captcha is now presented when submitting an article with less than 1 karma.
see http://code.google.com/p/lesswrong/issues/detail?id=233
Will see, but I expect this won’t be air-tight, since the spammers manage to register anyway, and there is already a captcha on registration. There are tools for solving captchas cheaply, but not for posting good comments.
This was easy to do.
We figured that some spammers would sweatshop registration then hand over the accounts to bots. If true, this should increase the cost of spam to LW.
More spam today. Prohibiting posting by users with Karma<1 can’t be that difficult to implement—we even have the 20 Karma cutoff on the main site...
“X can’t be that difficult to implement”: something one probably shouldn’t say about an open source project unless one is about to give it a go themselves.
There’s are at least two opportunities to suffer motivation drain in your comment :)
By writing that captcha was implemented because it was easy to implement, you suggest that easy-to-implement property has nontrivial explanatory power. I disagree, since I believe this property also holds for other alternatives, so can’t explain the choice, and shouldn’t be used as an argument. It’s just a matter of rationalist nitpicking.
(It’s more difficult for me, because I don’t know the project (or python, for that matter), but must be trivial for people who do. In any case, the intended meaning, as applied to Karma limit, is the same as with captcha in the sentence that originated the exchange, so it refers to how difficult it would be to implement the Karma limit (instead) for the person who implemented captcha.)
I accept your criticism and will stop nitpicking. What we did was particularly “easy” because we only turned that captcha on and styled it. Captcha-on-submit was already there in the Reddit code.
We’ll try something else to reduce the impact of spam soon.
Just seconding this. Still more spam today, so CAPTCHA is ineffective.