Behdad Esfahbod
Behdad Esfahbod
Closing in favor of https://github.com/fonttools/fonttools/issues/3521
I pushed some fixes to CFF2ToCFF in: https://github.com/fonttools/fonttools/pull/3886 Note that you still need to instantiate font WITHOUT `--downgrade-cff2`, and then do a `fonttools cffLib.CFF2ToCFF` separately. But if you do that,...
> [@behdad](https://github.com/behdad) It works perfectly, thanks a lot! Nice! Thanks for testing.
@liZe Note that I also fixed, or tried to, the `--downgrade-cff2` argument.
Currently: ``` fonttools pens.svgPathPen AdobeVFPrototype.otf A ``` generates output using integers only: ``` ``` As soon as I set any `--variations`, even to the default, it switches to float because...
> * everything is decomposed in this CFF2, while the TTF still has components. Hence the CFF reporting on X, Chi, and Ha (XΧХ) while the TTF only reports on...
I have no clue why the test is failing on Python 3.8 (not finding the kink).
Now all of the new code is tested. I think this is ready; just the test failure remains.
> > * curve conversion may be responsible for new kinking possibilities being introduced > > The "J" for example, is kinky in both TTF and OTF. But only reported...
I don't know why Python 3.8 is failing. Otherwise, this is ready I think.