1322189 - Transform interpolation of rotate functions don't match on the correct primitive type

Status: NEW

(Core :: DOM: Animation, defect, P3)

Description

[alancutter]

User Agent: Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/55.0.2883.59 Safari/537.36

Steps to reproduce:

Spec: https://drafts.csswg.org/css-transforms-1/#interpolation-of-transforms
Test case: http://jsbin.com/malabicozu/edit?html,output



Actual results:

The following tests fail:
Animating transform from rotate(10deg) to rotate(90deg) should not use post multiplication
Animating transform from rotateX(90deg) to rotateY(90deg) should not use post multiplication
Animating transform from rotate(10deg) to rotate3d(1, 2, 1, 90deg) should not use post multiplication
Animating transform from rotate(10deg) to rotateZ(90deg) should not use post multiplication
Animating transform from rotate(10deg) to rotate3d(0, 0, 1, 90deg) should not use post multiplication
Animating transform from rotate(10deg) to rotate3d(0, 0, 2, 90deg) should not use post multiplication


Expected results:

All tests should pass.

Comment 1

[Brian Birtles (:birtles)]

Simplified test case: http://jsbin.com/sugafebevi/edit?html,css,js,console,output

Comment 2

[Emilio Cobos Álvarez [:emilio]]

http://jsbin.com/rowiyihano/1/edit?html,output is the original test-case with the URLs fixed.

Alan, do you know if this is tested in WPT? If not, would it be helpful to upstream this? Thanks!

Comment 3

[Emilio Cobos Álvarez [:emilio]]

Also, this regressed at some point, since now 17 tests are failing.

Comment 4

[alancutter]

This is not tested in WPT. For added context this test came from a fix in Blink: https://chromium.googlesource.com/chromium/src/+/a5df6c6cf9b46d62121f92e5e4161058877c7051%5E%21/
Upstreaming this to WPT would be useful (though I'm still not happy with how hackily the expectations are checked).

Comment 5

[Brian Birtles (:birtles)]

Still 17 tests are failing even after bug 1464615 landed.