When importing a fnt file into Caps, comments and Unicode code points (U+xxxxx) are interpreted as kerning pairs. When importing something like this:
-- this is a comment, which should be ignored
space 10
� 9
A 10
U+42 10
C 10
D 10
E 10
...
pU+20 -1
a kerning pair for two dashes is created for the comment instead of ignoring it. For U+42 a kerning pair is created for ‘U’ and ‘+’ The character ‘B’ (U+42) does not take a glyph from the image table. Instead, the character set is thrown off, like so:
Kerning pairs defined in the fnt file are also not parsed correctly either, for instance pU+20 results in a pair for ‘p’ and ‘U’ instead of ‘p’ and ‘space’
This issue seems similar to this thread Oddities/contradictions re: fnt file format edge cases & fnt documentation which I think solved this for the compiler?