lucy24 - 12:53 am on Aug 14, 2012 (gmt 0)

I think I know what you're talking about, but let me double-check. I've sometimes found visitors from Japan whose queries looked at first sight like gibberish-- strings of percent-encoded stuff, meaning non-ASCII. But when I go to the trouble of disencoding, they turn out to be something from the FF range (decimal 65280etc, utf-8 EFhhhh), "Half-width and Full-width Forms". They look like Latin letters but are classified as Hiragana and Katakana.

Sometimes just to confuse me they'll have regular Latin script but wonky punctuation. This may be from a Japanese OS that uses its own spaces and punctuation, but brings in "real" Latin letters-- the opposite of what you'd see if your default font is Latin but you're inserting a few Japanese words. (It's not technically the same-- they are different characters, not just a different font-- but they look the same.)

Cyrillic is different. It's a pretty big alphabet so you could randomly make words that happen to look Roman. (Same goes for Greek if you stick with capital letters: ABEZHIKMNOPTYX.)

Oops. Uhm. Favorite subject, there. What was the question?