change encoding to utf-8

[fbexiga]

In the cases where the strings have accentuated characters, each one of those is counted twice. For example "ä" would be two chars, which is not correct.

Changing the encoding to utf-8 fixes that. :)