Adding data for votf/vuid
dberlow opened this issue · 3 comments
Dave asks if we can provide prototype glyphs for testing development of these two axes, first described on variationsguide.typenetwork.com
Not yet given a complete feature list, much of the axis creation for this is complete. Figures ranging from regular to numerators, and small caps, can be tested today if blending, fencing and nesting this in their target style works. What we do not yet include is shifting glyphs that are not base aligned.
The question lingering for both of those axes, is whether they can use instructions similar to the existing TT component definitions, to eventually include facility for positioning the nested blends of glyphs that way, or whether positioning is implemented by adding axes for that as part of the blending (and fencing).
Hopefully the VARC format can make these attainable.
https://github.com/harfbuzz/boring-expansion-spec/blob/main/VARC.md
If FB can demonstrate Roboto Flex numeral sets and the moon phase characters as proposed in 2016 as vuid/votf, that will set stage for using VARC for fencing, I think :)
Fontra; it's used to make varcs for locl features for localization of CJK characters from Simplified Chinese to Traditional Chinese, Kanji for Japan and Hanja for Korea; and does it at two levels, although the format supports any number of levels or even recursion. So there are 4 glyphs for one unicode character, 1 is the default encoded glyph and the other 3 glyphs are accessed via locl feature, but those 3 glyphs contain the same components as the character glyph, yet with different locations for the components in their own 'local' design spaces.