If I paste
Unicode character 2019 into my reply and hit preview, the server sees the UTF-8 encoding e2 80 99 and (mis)translates each byte into a separate HTML entity: ’
If it was working correctly, it would send back something like ’
The POST is being sent as UTF-8, and the result is coming back in UTF-8, so the scripts on the backend are probably just using the wrong charset (iso-8859-1?) to decode the form data.
This may help: http://ahinea.com/en/tech/perl-unicode-struggle.html