Submitted by murph on 7 October 2007 - 11:20am. computers
I've upsized my spam-blocking to ReCaptcha, which is what ArborWiki is using. ReCaptcha is a blurred-word test, but it has real utility. I at first thought the bit "stop spam. read books." was some snarky meaningless "get off yer computer" thing, but I knew something was up when I saw "Carnegie Mellon" attached to it.
Apparently, ReCaptcha is hooked into the Internet Archive's scanned text recognition engine: it gives you two words to decode, one of which is known, and used to see if you're a real human, and the other of which is unknown, and your answer combined with other recaptcha entries to provide the text recognition with an answer. Neat!
If you're a known human commenter on this site, you probably won't ever see it (sorry), but you can still appreciate that it's there. (And took maybe half an hour to get set up both on this site and the personal wiki.)