On 15-12-01 06:39 PM, Adam Twardoch (List) wrote:
> I don't think anything special is needed there, but actually, I'd like to ask
> how GLIF/UFO should store the encoding information for glyphs that represent
> Unicode Variation Sequences (UVS), which need to end up being written in a
> format 14 "cmap" subtable as per
>
https://www.microsoft.com/typography/otspec/cmap.htm
>
> In effect, they are not nuch different from a ligature glyph, only that the
> "ligation" behavior is happening on the Unicode codepoint level, not the
> OpenType GSUB level.
>
>
http://unifiedfontobject.org/versions/ufo3/glif.html
> allows multiple "unicode" elements when one glyph can be accessed via
> different codepoints, but suggests that only a single codepoint can be stored
> in the element. With UVS, however, it needs to be a sequence of codepoints.
Allowing sequences in "unicode" element sounds about right to me.
b
> Best,
> Adam
>
> Sent from my mobile phone.
>
> On 02.12.2015, at 02:02, Tal Leming <
t...@typesupply.com
> <mailto:
t...@typesupply.com>> wrote:
>
>> Hm. This hasn't come up before and it isn't something that I'm familiar
>> with. What kinds of things do XML specifications need to do to handle them?
>>
>> Thanks,
>> Tal
>>
>>
>>> On Dec 1, 2015, at 7:59 PM, Bran Kom <
briankim...@gmail.com
>>> <mailto:
ufo-spec+u...@googlegroups.com>.
>> <mailto:
ufo-spec+u...@googlegroups.com>.
> <mailto:
ufo-spec+u...@googlegroups.com>.