One of the ways in which I prepare for an interview is to research the person I'm going to be interviewing. The first step is to toss their name into one (or more) search engines and see what falls out. Not getting any useful results isn't a negative (many people work on internal stuff with no concern for fame fortune and glory, and they're very good at it), but if I happen to learn a little about them from a blog entry or a profile on a social networking site, then I can direct my questions in a way that more precisely fits the candidate.
So today, I sat down to read through the resumé of an upcoming candidate and punched his name into the best damned search engine ever. What's relatively unique about this candidate is that his name contains an accented character. The search results handled this accented character cleanly of course, and when I followed the first result for "José Smith" (Not his real name -- I'm respecting privacy here), I found myself looking at his Facebook profile. All well and great of course, but then I noticed the URL in the address bar: http://www.facebook.com/Jos_Smith/123456789.
Jos??? The only Jos I know is a director with a penchant for vampires.
Now, I get not putting the UTF8 or latin1 literal in the URL, browsers don't agree nearly enough on URL encoding to be able to expect it to work reliably. And putting the urlencoded version would just be silly since /Jos%C3%A9_Smith/ looks even uglier. I get that. But what if his name were entirely non-ascii? Perhaps a chinesse name? Would it be so hard to do a transliteration? I seem to recall Derick sharing a few ideas on the subject a couple years ago...