I hate Captcha challenges, I wasted five minutes trying to sign up to a website to try out a piece of software because the damn ReCaptcha system was spitting out gibberish. Human-powered OCR only works because we don’t need to see every letter when reading words; we treat the words as “lumps” of meaning rather than collections of letters, and the odd scrambled or badly scanned letter won’t throw us. This only works when recognising words in our vocabularies.
The following lumps of gibberish are not in my vocabulary…
I’m sorry, I live in a country where our characters are based on the Roman/Latin character sets, not Greek. I don’t know where my captial-theta button is and I’m not about to faff trying to work out what Alt-xxx code I need either.
Ah now wait… we can do upside-down characters using Unicode, maybe that one isn’t so bad.
And this one should be really easy to solve… providing it’s not actually white text…
 – I’m trying to see if Tweetdeck works better than my current Twitter setup
I’m also trying footnotes instead of inline links to see if it makes things easier to read.