@davidbaron Came across a Firefox inconsistency codepen.io/thebabydino/pen/Qโ€ฆ
2
recently @brianskold was, I think, working on a patch to update to spec change that says having an animation of transform...
1
... acts as though it's set, even if it's none. I think that update would change the behavior here?
1
Yes, I'd expect so.
1
The change we made only affects the creation on stacking contexts, not computed style
1
1
It probably should affect the computed style that way, assuming that's what Chrome/Edge do. Maybe will-change too?
1
The change was based on a spec change: if you have a delayed animation for prop x acts as if will-change:x was set
1
Replying to @brianskold @anatudor
Yes, and maybe will-change: transform should make computed transform be matrix() rather than none?

Oct 22, 2016 ยท 10:36 PM UTC

1
Replying to @davidbaron @anatudor
either that or interpolating between "none" and any other transform (incl. none) produces identity matrix at endpoint
1