~silnrsi/smith/+git/ufo2ft-master:ufo3-only-use-marked-groups-as-kerning-classes

Branch merges

Related source package recipes

No recipes using this branch. (?)

Create packaging recipe

Related snap packages

Related charm recipes

782f334... by Nikolaus Waxweiler <email address hidden> on 2018-04-11

UFO3: Only use groups starting with public.kern[12]. as kerning classes

5c13d7f... by Denis Moyogo Jacquerye on 2018-03-30

Merge pull request #239 from moyogo/postscriptWeightName

CFF topDict Weight is optional, fallback to None

6a979a9... by Denis Moyogo Jacquerye on 2018-03-30

[tests] info.postscriptWeightName fallbacks to None

99f3415... by Denis Moyogo Jacquerye on 2018-03-30

CFF topDict Weight is optional

postscriptWeightName can fallback to None

8ba91ce... by Denis Moyogo Jacquerye on 2018-03-24

Merge pull request #236 from moyogo/propagate-anchors

[propagateAnchors] Don't crash on anchor propagation on inexistent components

a4c8b2b... by Denis Moyogo Jacquerye on 2018-03-23

[tests] unused variable and whitespace

1c03520... by Denis Moyogo Jacquerye on 2018-03-23

[propagateAnchors] warn on missing baseGlyph

40dec09... by Cosimo Lupo <email address hidden> on 2018-03-16

Merge pull request #234 from anthrotype/cid-fontname

[outlineCompiler] don't set CID-only FontName operator in CFF TopDict

4d38dac... by Cosimo Lupo <email address hidden> on 2018-03-15

adjust expected test results after dropping FontName

6ab5022... by Cosimo Lupo <email address hidden> on 2018-03-15

[outlineCompiler] don't set CID-only FontName operator in CFF TopDict

The 'FontName' operator is only valid in CID-keyed CFF fonts

Since we emit a name-keyed CFF table, this field is redundant at best.

Even worse, on latest macOS 10.13 the presence of this operator in the
CFF breaks the QuickLook preview (space-bar from Finder), and the font
appears all blank; maybe they use it to detect whether CFF is name- or
CID-keyed, and then fail to render because in fact it isn't CID...

Besides, MakeOTF silently strips this when it is passed an OTF produced
by fontmake.

So we can safely remove it.

Note that the postscriptName from the fontinfo.plist is still encoded
in the Name INDEX structure (ttx dumps this as the 'name' attribute of
the CFFFont element).
It is just the dict operator that should not be there.