fontbakery icon indicating copy to clipboard operation
fontbakery copied to clipboard

diacritic check summary

Open RosaWagner opened this issue 3 years ago • 6 comments

I try to reformulate what was said in the meeting yesterday for the diacritic check. I'll update this to-do with more details

  • [ ] #3959
  • [x] #3490
  • [x] FAIL if missing combining mark (and mark feature) -> this is already done in gf-profile with the glyphset check since combining marks are in the defined GF glyphsets.
  • [ ] #3328 and #3322
  • [ ] #3296
  • [ ] #3293
  • [x] #3308
  • [ ] #4210

ccmp related checks:

  • [ ] #4068
  • [x] #4059

RosaWagner avatar Jun 02 '21 14:06 RosaWagner

it is OK to have a summary issue like this to track progress on a set of things. But it's important that the details are discussed in the individual issues for each of these new check proposals. Please post an issue for "WARN top-comb accents have _top and top anchors to allow multiple combining accents"

felipesanches avatar Jun 02 '21 16:06 felipesanches

[ ] WARN Diacritics are used consistently (#3296)

I would propose a FAIL status for this one.

Added a new item to the above list.

vv-monsalve avatar Jun 04 '21 17:06 vv-monsalve

@RosaWagner mentioned on chat "For the problem with GDEF table and confusion between legacy accents and comb accents, I can confirm that fixing Radio Canada is fixing this display issue on the API. It seems therefore more ans more important to implement [this issue]"

davelab6 avatar May 06 '22 16:05 davelab6

FYI, all the Tiro Indic families have a GDEF table warning;

⚠ WARN: Check mark characters are in GDEF mark glyph class. (com.google.fonts/check/gdef_mark_chars) ⚠ WARN The following mark characters could be in the GDEF mark glyph class: acutecomb (U+0301), brevebelowcomb (U+032E), brevecomb (U+0306), candrabinducomb (U+0310), caroncomb (U+030C), cedillacomb (U+0327), circumflexcomb (U+0302), commaaccentcomb (U+0326), dieresisbelowcomb (U+0324), dieresiscomb (U+0308) and 9 more.

This is the consequences of such issue in GF: Capture d’écran 2022-05-18 à 10 59 04

@davelab6 I don't know if the warn should be a fail, or if GF should stop displaying glyphs that are actually not encoded in the font.

cc @vv-monsalve @m4rc1e

RosaWagner avatar May 18 '22 09:05 RosaWagner

This seems important enough to increase priority, so I moved it now to the immediately next planned release.

felipesanches avatar Mar 17 '23 10:03 felipesanches

Adding this issue to bump up priority on this list of checks, this is the example of a font that shouldn't have passed QA: https://github.com/google/fonts/issues/6226

RosaWagner avatar Apr 26 '23 15:04 RosaWagner