I have a CGI script. It uses CDBI + TT. Mostly all text outputs alright but some words have \x.. instead of some letters. Generally, there are 1-2 broken letters on a page.
The script has "use encoding 'utf8'" and "use utf8" in all my modules. All scripts and files are in UTF-8 (without BOM). I set utf-8 flag for CDBI data and decode all CGI parameters.
Where could be the problem?
Thanks.