That moment when spammers start pushing language research forward by generating functions from natural language specifications.


On Tue, Apr 8, 2014 at 1:08 PM, Artyom Kazak <yom@artyom.me> wrote:
Look what I've found: http://codecha.org/ . It might be an easy solution to the Haskell-specific CAPTCHA problem.

On 04/07/2014 06:53 PM, Daniel Trstenjak wrote:
On Mon, Apr 07, 2014 at 09:27:32PM +0700, Kim-Ee Yeoh wrote:
What if we replace captcha with a short, static question, the web form
equivalent of a secret handshake? And give it enough weighting to override
akismet?

E.g.

* What is Haskell's middle name?
* What is SPJ's middle name?

Yeah, I thought about something similar like: what's the result of 'map (+1) [1,2]'.

The main drawback to this is that it'll only be a matter of time before
spammers wise up. But that interval might be long enough for something better
on the horizon, e.g. akismet gets a lot smarter, better blog posts on tracspam,
etc.

I don't think that the ghc wiki is of particular interest for spammers
or that they gain a lot by understanding Haskell specifics. Most likely
they will never notice it.


Greetings,
Daniel
_______________________________________________
ghc-devs mailing list
ghc-devs@haskell.org
http://www.haskell.org/mailman/listinfo/ghc-devs

_______________________________________________
ghc-devs mailing list
ghc-devs@haskell.org
http://www.haskell.org/mailman/listinfo/ghc-devs