Re: [css3-background] 'background-size: auto auto' and images with intrinsic size in one dimension

On 9/08/2011 8:04 AM, Tab Atkins Jr. wrote:
> On Mon, Aug 8, 2011 at 1:36 PM, Brian Manthos<brianman@microsoft.com>  wrote:
>> Please submit some test cases on this.  Because it's gone well beyond simple at this point.
>
> The CSS2.1 testsuite somewhat tests this already in
> <http://test.csswg.org/suites/css2.1/20110323/html4/background-intrinsic-006.htm>,
> though it's limited by the lack of background-size, which wasn't
> introduced until B&B.  For example, Chrome fails this for silly
> reasons - we first size the viewport to be 40%/60% of the background
> canvas, then size the<svg>  to be 40%/60% of the viewport bounds.
>
> This will be tested thoroughly by the B&B testsuite, when it's
> written.  Would you like a few impromptu testcases now?
>
> ~TJ


I can provide some of these test or test that can be rewritten for the 
test suite. I was intending of demonstrating the bug in WebKit (not just 
Chrome) about sizing the background-image svg to the viewport instead of 
the background canvas since it breaks when certain background properties 
are transitioned. This is why WebKit handles the below test wrong.


http://css-class.com/test/css/3/backgrounds/svg-intrisic-size-transition.htm


The sizing works correctly in Gecko with transitioning. The sizing works 
correctly in Opera 11.50 (Opera 11.11 was buggy) and IE10 but with no 
transitioning.


I have further test with various heights or widths given in both pixel 
and percentages for the SVGs.




-- 
Alan Gresley
http://css-3d.org/
http://css-class.com/

Received on Tuesday, 9 August 2011 00:37:29 UTC