Closed davelab6 closed 6 years ago
I tried this last night too with the same results. Is this a proposed spec somewhere?
I was also wondering about this. I don't think the spec supports this yet (@davelab6 correct me if I am wrong)
I think the values should come from the font postscriptUnderlinePosition
and postscriptUnderlineThickness
. Not sure if there's a var table for this...
I see the MVAR should support this: https://github.com/fonttools/fonttools/blob/master/Lib/fontTools/varLib/mvar.py
'unds': ('post', 'underlineThickness'), # underline size
'undo': ('post', 'underlinePosition'), # underline offset
But testing on Chrome and Illustrator CC doesn't change anything... Seems to be an application issue?
I hope @dberlow can comment
Oh?
Roboto delta underline axis
Executive summary?
The target use of the axis - In the middle of the design space as far as weight and width are concerned and from the smallest optical size up to Perhaps all but the largest headlines, Roboto Delta will have a flexible underline matches wght, wdth an op size, and has an additional flexibility of being located for use with specific character groups at specific alignments.
The list of target variables - ? 1.the Width of each glyph being underlined, 2.the distance from the baseline to the top of the default underline of an instance,
The options for underalignments, a. Default is at the descent? b. UDLN instance Under Uppercase and/or lining figures C. UDLN instance Under mixed case, (bisecting descenders?) D. UDLN instance Under all glyphs. And/or e. Slider that goes back-and-forth, user picks visually.
What the app does - ?
1.Figures out the string length to be undl-ed 2.gets the instance’s distance, baseline to top of stroke from udln
Issues - I don’t think this is how it’s done today. I think the MVAR contains those two values, weight, and depth of underline. For the same function, and production tools must give a delta range for the weight and alignment of the underline so that the mvar can interpolate to values for those two things for any given instance produced from the var.
So, the plumbing is there in the mvar, but training applications to parse the MVAR? Or Using the UDLN axis for variable alignment, the glyph for the weight, and the app for the width of the string, seems okay too?
So we're looking for a visual demo of the desired effect. I can put together a quick experiment using a bottom border, which would be useful to see if the browsers can even do subpixel rendering of the line.
Should this be moved to: https://github.com/TypeNetwork/tn-variable-fonts-brochure
moved and closing https://github.com/TypeNetwork/tn-variable-fonts-brochure/issues/84
All the options I know for Latin, are to underline anywhere from just below the baseline, to just below the em.
Dependancies are Case of line, for all caps vs mixed, Mid-Descender depth, for some tastes? Descender depth. for other tastes? Linespacing,
Moving this to TypeNetwork/tn-variable-fonts-brochure#84
When I activate
text-decoriation: underline
with the browser's Inspector, theUDLN
axis doesn't do anything... How to check out its functionality? :)