Sometimes, I see some characters as the cp1252 and/or iso-latin-1 interpretation of their UTF-8 encoding.
Assuming it's not a client side bug, it looks like PerlMonks outputs UTF-8 while claiming to output cp1252 in some circumstances. I'll see if I can nail down the details the next time it happens to me.
If I view the same node from a different URL, I see the characters correctly. This could indicate a problem with one of the special nodes (e.g. node 3333)