[TazeTSchnitzel]

> Beyond that, the skin tones used are incredibly reductive. Human skin tones are not as simple as 6 different shades of brown.

Yes, they are a bit more complicated. But 6 choices that correspond to a widely-used system (Fitzpatrick) is far better than none.

[chimeracoder]

> correspond to a widely-used system (Fitzpatrick)

The fact that a system is widely used when classifying the impact of UV light on melanoma does not imply that it has relevance in another.

> 6 choices is far better than none.

Actually, no, sometimes the "solution" is worse than the problem. It's quite regressive to bake an outdated conception of the color theory of race into a standard that aims to "educate and engage academic and scientific communities, and the general public" (the stated mission of the Unicode Consortium).

As one of the "billions of people who aren't white" that you refer to in your original comment, I find this approach more problematic than the existing status quo (leaving it up to the font creators).

[TazeTSchnitzel]

> The fact that a system is widely used when classifying the impact of UV light on melanoma does not imply that it has relevance in another.

You may have a point there. But it is a classification of skin colour as well.

> It's quite regressive to bake an outdated conception of the color theory of race into a standard

Color theory of race? This isn't the Von Luschan chromatic scale, that was used to enforce racial segregation. It's a simple colour selector based on level of melanin.

It's not perfect, sure, but I think it's surely better to allow a choice of skin shades than to make everyone white (the de facto result otherwise). You might object that implementers could make everyone black, say, and that's also true. But either way, you're enforcing a "default" skin colour, and there is no such thing. Different human populations have different skin colours.