[MPEG-OTSPEC] WD for AMD 2 and COLR v1 enhancements

[Laurence Penney]

Thanks for your thoughts, Peter. I think the problem is unsolvable without the introduction of a “rendering layer” identifier for each glyph layer.

As a kind of hack, one could assert that (with a certain flag set), palette indexes identify these layout-wide layers. Then, the font would instruct a renderer to render all the palette 0 glyph layers, then all the palette 1 glyph layers, then the palette 2 layers, and so on. Presumably this would ignore layer order in the COLR table. The problem of defining how far the scope of this “rendering layer” extends, in terms of the actual layout on the canvas, is perhaps a blocker, and non-solid colours present blend mode issues too.

- L