today while generating the instances needed for #301 i stumbled upon the following major bug: Kerning is utterly broken for all /l combinations with not apparent reson.
The glyph is not present in ANY kern class, nor has any kern values attached even as simple glyph to glyph nor glyph to class pairs. This puzzles me! Examined all the auto-generated features but to no luck. Finally started removing feature by feature from the font ... finally arrriving at the conclusion that removing the ordn feature fixes the problem?! Which is even more puzzling...
feature ordn {
# Ordinals
#> feature
sub [zero one two three four five six seven eight nine] [A a]' by ordfeminine;
sub [zero one two three four five six seven eight nine] [O o]' by ordmasculine;
sub N o period by uni2116;
#< feature
} ordn;
Removing the feature apparently fixes the bug?! so i will keep it that way for a while, just to be sure that the exported TTFs are usable!
today while generating the instances needed for #301 i stumbled upon the following major bug: Kerning is utterly broken for all /l combinations with not apparent reson.
The glyph is not present in ANY kern class, nor has any kern values attached even as simple glyph to glyph nor glyph to class pairs. This puzzles me! Examined all the auto-generated features but to no luck. Finally started removing feature by feature from the font ... finally arrriving at the conclusion that removing the
ordn
feature fixes the problem?! Which is even more puzzling...Removing the feature apparently fixes the bug?! so i will keep it that way for a while, just to be sure that the exported TTFs are usable!
To be investigated!