Chris Lilley

@emilio now?

> fantasai: Seems like we should fix up spec to say what we want shipped and once that's ready we say go ahead > Rossen_: Do you want to go...

> Should we add some tests here to reflect that? That would be great

System colors are [already defined](http://w3c.github.io/csswg-drafts/css-color-4/#resolving-other-colors): > Each [``](http://w3c.github.io/csswg-drafts/css-color-4/#typedef-system-color) keyword computes to the corresponding color in its color space. However, such colors must not be altered by 'forced colors mode'.

> What is this in response to? It was in response to > Is currentColor the only issue? What about system colors that depend on the color-scheme computed value, e.g.:

> ```css div { color: color-mix(in lch, canvastext, blue) } div.dark { color-scheme: dark; } ``` > Depending on whether an element has the class "dark" or not, the computed...

> Another open question: what should we do for negative calc values? Hmm. I thought that was already covered: > Percentages are required to be in the range 0% to...

> If you have two negative percentages (or 0%), and they are clamped at computed value time, you will end up with two 0% values which is still invalid per...

> it's the same handling as any other property or function that only allows positive numbers, e.g. `padding` or `` in `inset()` True, although some examples demonstrating this wouldn't hurt.

I discussed a similar issue with @emilio at TPAC and the problem is that most, but not all, of [to serialize a CSS value](http://w3c.github.io/csswg-drafts/cssom-1/#serializing-css-values) for colors moved to CSS Color...