Emilio Cobos Álvarez

I haven't seen any compat issues with that, fwiw. And I think making `calc()` quirky would be unfortunate unless we see them.

Per spec `calc(% + 0px)` behaves the same as `%`. So no, it's not quite the same than when you're mixing actual lengths.

And I intentionally changed the behavior of that particular test-case in https://bugzilla.mozilla.org/show_bug.cgi?id=1577139, per the resolution in https://github.com/w3c/csswg-drafts/issues/3482.

That is not always compatible with the above discussion, see http://w3c-test.org/css/css-tables/calc-percent-plus-0px-auto.html for a simple example where Chromium agrees with Firefox. I don't feel _too_ strongly about this, but I think...

I'm ok if we do this in the name of "shortest path to interop", but I think this is a bit sad :/

> @emilio Firefox isn't consistent with itself wrt. heights here - for example: > https://www.software.hixie.ch/utilities/js/live-dom-viewer/?saved=7546 > https://www.software.hixie.ch/utilities/js/live-dom-viewer/?saved=7547 Can you elaborate? What should I see? I know of differences with other...

Ah you're totally right. We use `HasPercent` in one case (the replaced elements quirk) and `ConvertsToPercent()` in another... It seems the `HasPercent()` call (which causes the discrepancy) was introduced all...

It's not terribly complex. we intentionally don't differentiate between a calc or not since https://bugzilla.mozilla.org/show_bug.cgi?id=1577139, though we could re-introduce it... That being said, I think given Firefox is inconsistent, and...

Yeah, seems reasonable to do this after looking a bit more into it. Still slightly weird but :)

> There’s a potential concern about using DOM attributes as part of the layout algorithm. I don’t quite understand this one fully - if someone (@emilio?) can help me understand,...