Line length in Fido texts is based on bytes, not characters

[punktniklas]

When Fido texts are read from disk they are broken up into lines (since the Fido text format is based on paragraphs, not lines). The line length used is 79 when reading texts and 75 when quoting. This is counted in bytes, which was the same as the number of characters when everything was converted to ISO Latin 1. Now however it's possible that the text is in a multi byte format such as UTF-8. This means that if all characters in a line use two bytes the effective line length will be less than 40. And for characters that use even more bytes (such as many asian languages) it will be even shorter.

The reading logic should be updated to count characters (Unicode code points), not bytes.