[css-inline] HTML taking a dependency on ascent/descent metric

[annevk]

FYI: HTML is taking a dependency on https://drafts.csswg.org/css-inline/#ascent-metric (and descent) in whatwg/html#5429. It'd be appreciated to get a heads up if these break at some point in the future.

cc @jfkthame @yiyix

[fantasai]

@annevk That seems fine in theory, but I do want to point out that the ascent/descent metrics as used by CSS are non-interoperable across implementations, and we're not able to fix this. (There's some discussion of this in #2228 and some follow-up in #4792 ) I'm not sure what the state of the CanvasAPI is on that front.

[annevk]

What I'd personally like is for canvas to use the same underlying drawing primitives so "ideally" it would be equally non-interoperable until that is either sorted or accepted as implementation-defined behavior. @jfkthame can speak much more to all this though.

[fantasai]

@annevk Sounds good. You might also want to cross link into the definitions for “alphabetic baseline”, “hanging baseline”, and “ideographic baseline” in https://www.w3.org/TR/css-inline-3/#baseline-types

I filed #5312 on the emHeightAscent and emHeightDescent metrics.

[annevk]

Done in whatwg/html#5723. I think this can be closed then. And we can use #5312 to track the remainder.