Re: Issue 130339 in chromium: Chrome ignores 3rd parameter of -webkit-transform-origin

[chro...@googlecode.com]

Updates:
Labels: -Area-Undefined Area-WebKit WebKit-CSS WebKit-Rendering

Comment #2 on issue 130339 by tk...@chromium.org: Chrome ignores 3rd
parameter of -webkit-transform-origin
http://code.google.com/p/chromium/issues/detail?id=130339

(No comment was entered for this change.)

[chro...@googlecode.com]

Updates:
Cc: shawnsi...@chromium.org vang...@chromium.org wiltz...@chromium.org
Labels: Feature-GPU-Compositing

Comment #3 on issue 130339 by jam...@chromium.org: Chrome ignores 3rd

[chro...@googlecode.com]

Updates:
Status: Assigned
Owner: shawnsi...@chromium.org

Comment #4 on issue 130339 by shawnsi...@chromium.org: Chrome ignores 3rd

parameter of -webkit-transform-origin
http://code.google.com/p/chromium/issues/detail?id=130339

Assigning it to myself just so it doesn't get easily lost. I'll post an
update when I start working on it.

[chro...@googlecode.com]

Updates:
Owner: ccame...@chromium.org

Comment #5 on issue 130339 by ccame...@chromium.org: Chrome ignores 3rd

parameter of -webkit-transform-origin
http://code.google.com/p/chromium/issues/detail?id=130339

[chro...@googlecode.com]

Updates:
Status: WontFix

Comment #6 on issue 130339 by ccame...@chromium.org: Chrome ignores 3rd

parameter of -webkit-transform-origin
http://code.google.com/p/chromium/issues/detail?id=130339

The bug here appears to be with Safari, rather than Chrome.

In particular, if the point <x,y,z> is specified by
-webkit-transform-origin, and the transformation matrix T is specified by
-webkit-transform, then the overall transformation should be:
[1 0 0 x] [1 0 0 -x]
[0 1 0 y] * T * [0 1 0 -y]
[0 0 1 z] [0 0 0 -z]
[0 0 0 1] [0 0 0 1]
It appears that Safari is not performing the final left-multiplication to
re-center the content at the specified origin after the transform is
performed.

I've attached a stripped-down version of the originally provided file,
along with its rendering in Chrome and Safari. The top image shows that an
identity transformation with non-zero Z-origin has a non-identity effect on
Safari. The middle image shows an ordinary rotateY(180deg). The bottom
image verifies that a non-zero Z-origin correctly affects a rotateY(180deg)
transformation on Chrome.

Marking works-as-intended and will follow up with a bug report to Safari.

Attachments:
perspective-test-ex.html 1.9 KB
ex-chrome.png 75.5 KB
ex-safari.png 97.3 KB

[chro...@googlecode.com]

Comment #7 on issue 130339 by shawnsi...@chromium.org: Chrome ignores 3rd

parameter of -webkit-transform-origin
http://code.google.com/p/chromium/issues/detail?id=130339

Did you have a chance to double-check exactly why Safari and Chromium
behavior differ? When we left off our offline discussion, we forgot that
question was still in the back of our heads =)

Its probably worth digging into that, at least for a little while.
Personally I did not expect that their behavior would differ; I thought the
transform-origin was handled by WebCore for both Safari and Chromium.

[chro...@googlecode.com]

Comment #8 on issue 130339 by ccame...@chromium.org: Chrome ignores 3rd

parameter of -webkit-transform-origin
http://code.google.com/p/chromium/issues/detail?id=130339

Filed bug against WebKit for Safari follow-up
https://bugs.webkit.org/show_bug.cgi?id=88587

[chro...@googlecode.com]

Comment #9 on issue 130339 by ccame...@chromium.org: Chrome ignores 3rd

parameter of -webkit-transform-origin
http://code.google.com/p/chromium/issues/detail?id=130339

Thanks, I just double-checked and WebKit nightly indeed reproduces the
issue. From the discussion in the WebKit bug it looks like this is an
already-known issue.

[chro...@googlecode.com]

Comment #10 on issue 130339 by shawnsi...@chromium.org: Chrome ignores 3rd

parameter of -webkit-transform-origin
http://code.google.com/p/chromium/issues/detail?id=130339

Just to clarify - double-checking WebKit nightly and Comment #7 were two
different double-checks.

[chro...@googlecode.com]

Comment #11 on issue 130339 by pet...@google.com: Chrome ignores 3rd

parameter of -webkit-transform-origin
http://code.google.com/p/chromium/issues/detail?id=130339

I spoke to the dev who originally brought this bug to my attention, and we
both feel it is still a bug in webkit, but working in the 3D space makes it
a bit harder to figure out. He's going to create another example and we'll
get it attached to this this week.