Pan African glyphs list

Some of the issues I’ve noticed while cleaning up the Google Fonts Latin African glyphset are also present in Glyphs.app’s Pan African glyphs list. When comparing the two, the Pan African glyph list has Ǟ ǟ Ȫ ȫ Ƀ ɋ ɾ ʕ Ṑ ṑ Ṓ ṓ Ẁ ẁ Ẃ ẃ but GF Latin African doesn’t.

The glyphs for ɋ and ɾ /qhooktail/rfishhook do not belong in the Pan African glyphs list.
The letter Ɋ ɋ is used in Kate and Kube, and was used Numanggang, those are languages in Papua-New-Guinea. The capital Ɋ is not in the Glyphs.app Pan African glyph list anyway.
The ɾ is an IPA symbol, not a letter used in African orthographies.

The glyph for Ƀ should not be in the Pan African glyphs list, ƀ is not anyway.

ʕ probably shouldn’t be there either, it is used in transliteration of Arabic for example. Orthographies that use it are North American or South American as far as I know.

The glyph for ƒ, /florin, should be sorted as a letter and be called /fhook to avoid confusion.

The glyphs for ǞṒṐȪǟṓṑȫ probably shouldn’t be in the Pan African glyph list. They don’t appear to be used in current orthographies.

I think the only ones that may need to be in GF Latin African are Ẃ ẃ and maybe Ẁ ẁ but they are already in more prioritized smaller glyphsets.

As the GF Latin African set is generated from gflanguages language data based on references, I’d recommend updating the Pan African glyphs with the additional characters that are currently missing and that are not in Basic list (or the common Western European list).

2 Likes

Thank you for compiling the list of differences!

Regarding ƒ, the glyph name florin stems from the Adobe Glyph Library. I see it’s also used in the stub file GF_Latin_African.glyphs. I agree that it should be recategorized as a letter by default, but I am unsure whether the legacy of the name is strong enough to not change it.

Yes, it can be the glyph’s production name, uni0192 is also valid.

That can be fixed once the source it gets the name from is fixed.

What do you mean?

The current naming scheme and automatic code doesn’t work with Fhook, florin and florin.sc for example.

Since the glyph name has a long tradition, being part of the early Adobe Glyph Library, I was more hesitant to change the name, even in the source format. But you are probably right that keeping the production name in exported fonts should be enough and that most tooling dealing with source files nowadays relies exclusively on code points for the character identity of a glyph.

Thank you for this insight, Florian. As a font designer I see why it’s important to name glyphs properly either by their legacy names or, most importantly, by their encoded names or code points.

1 Like