It's already on bugzilla:
http://bugzilla.wikipedia.org/show_bug.cgi?id=1247
Alfio
On Wed, 5 Jan 2005, Brion Vibber wrote:
A quick test on ang.wikipedia.org in IE 6.0 on Windows XP SP2 shows that some of the two-byte UTF-8 characters are incorrectly URL-decoded by IE and are inserted as two characters composed of their individual UTF-8 bytes interpreted as 8-bit Latin-1 characters.
This is pretty clearly an incorrect interpretation of the code by IE. It can probably be worked around by skipping the URL encoding, but I'll have to do more testing for scripting-attack safety and compatibility in other browsers, so I don't have time to do it right now while I'm at work. Would you mind filing a bug report on bugzilla.wikipedia.org to make sure I don't forget? Thanks.
-- brion vibber (brion @ pobox.com)