Behdad Esfahbod comments

Results 1486 comments of


                                            Behdad Esfahbod

differences in str(float) between Python 2 and 3

I think we can close this now?

[varLib] Incompatible glyphs should be an error not a warning

> Currently varLib will happily merge fonts with incompatible glyphs and issue a warning: > > > glyph XXXX has incompatible masters; skipping It used to be that way but...

[varLib] Incompatible glyphs should be an error not a warning

> (because fontmake pipeline is too noisy and issues warnings left and right, but that is for a different day) Let's fix that first...

Pass BASE table through subsetter by default.

I'm happy to pass it through by default, since that does produce a correct font. @punchcutter can you rebase this plese?

Test that passing a markFilterSet sets the appropriate lookup flag

@simoncozens can you address review please?

[merge] MarkAttachClassDef

Yeah I can give you some starting points for that. That said, I'm also thinking about a new modules (possibly called `processor` or `dispatch`) that will improve on and centralize...

[merge] MarkAttachClassDef

> Give me a couple of days. To write out my thoughts, not implement it. :))

[merge] MarkAttachClassDef

At least MarkFilteringSets were done in b8615f9de653a170b9e4e2ba7ec1d0fd3539fd65 MarkAttachClassDefs are bigger piece of work.

[merge] MarkAttachClassDef

> MarkAttachClassDefs are bigger piece of work. And can very easily overflow the allowed 255... In fact, that's the only place that is currently broken in, eg. a merged Noto.

[instancer] _instantiateFeatureVariations() limitation when a range rule becomes applicable

> I'm not sure I understand what you mean by "default-applicable because it starts at 0".. Because the rule includes 0, it should apply by the default instance.