Intent to Ship: CSS Transforms unprefixed

512 views
Skip to first unread message

Douglas Stockwell

unread,
Apr 2, 2014, 1:41:05 AM4/2/14
to blink-dev

Spec

WD: http://www.w3.org/TR/css3-transforms/

ED: http://dev.w3.org/csswg/css-transforms/


Summary

CSS Transforms were announced in WebKit on the Surfin’ Safari blog in 2007. Usage of -webkit-transform is now approaching 50% of page loads. Given the prevalence of -webkit-transform there was even a time when Opera experimented with -webkit-transform in Presto. Now transforms are shipping unprefixed in IE and Firefox. The CSSWG also gives their blessing.


Link to “Intent to Implement” blink-dev discussion

No formal “Intent to Implement”, however there was discussion on this thread.


Is this feature supported on all five Blink platforms (Windows, Mac, Linux, Chrome OS and Android)?

Yes.


Compatibility Risk

Internet Explorer and Firefox ship roughly interoperable implementations. Manual testing of the 928 tests in the W3C test suite revealed some differences between our implementation and Firefox.


As noted earlier on blink-dev, there are some compatibility issues related to matrix decomposition and transform-style that are still under discussion by the CSSWG.


OWP launch tracking bug?

https://code.google.com/p/chromium/issues/detail?id=154772


Link to entry on the feature dashboard

http://www.chromestatus.com/features/6437640580628480

Eric Seidel

unread,
Apr 2, 2014, 1:50:43 AM4/2/14
to Douglas Stockwell, blink-dev
LGTM!

Darin Fisher

unread,
Apr 2, 2014, 1:52:33 AM4/2/14
to Eric Seidel, Douglas Stockwell, blink-dev
LGTM2

Adam Barth

unread,
Apr 2, 2014, 1:57:11 AM4/2/14
to Darin Fisher, Eric Seidel, Douglas Stockwell, blink-dev
That speadsheet is amazing.  For reference, here are the bugs linked from the speadsheet:

https://code.google.com/p/chromium/issues/detail?id=282226

We should fix those bugs and converge with other browsers, but the bugs don't seem serious enough to block unprefixing.  LGTM.

Adam

Dirk Schulze

unread,
Apr 2, 2014, 4:12:38 AM4/2/14
to Douglas Stockwell, blink-dev
928 tests seem to be many. In reality these tests just test a specific subset of CSS Transforms: 2D transforms on SVG content.  Which makes less than a quarter of the spec. Contributing tests is more than welcome!

Many of the spec requirements are simply not implemented for SVG in Firefox, Chrome or IE.

This is why you see so many “matching” behaviors between Firefox and Blink. Therefore, to color the column green is highly misleading. The implementations are just behind the spec.

Greetings,
Dirk

Adam Barth

unread,
Apr 2, 2014, 1:03:38 PM4/2/14
to dsch...@chromium.org, dstoc...@chromium.org, blin...@chromium.org
On Wed Apr 02 2014 at 1:12:41 AM, Dirk Schulze <dsch...@chromium.org> wrote:
928 tests seem to be many. In reality these tests just test a specific subset of CSS Transforms: 2D transforms on SVG content.  Which makes less than a quarter of the spec. Contributing tests is more than welcome!

Many of the spec requirements are simply not implemented for SVG in Firefox, Chrome or IE.

This is why you see so many “matching” behaviors between Firefox and Blink. Therefore, to color the column green is highly misleading. The implementations are just behind the spec.

That's fine though.  As long as browsers behave the same way, we can all make progress together.

Adam

Alex Russell

unread,
Apr 2, 2014, 2:25:06 PM4/2/14
to Darin Fisher, Eric Seidel, Douglas Stockwell, blink-dev
LGTM3


LGTM2
To unsubscribe from this group and stop receiving emails from it, send an email to blink-dev+...@chromium.org.

Reply all
Reply to author
Forward
0 new messages