RE: Property inheritance and the 'clipPath' element

We looked into this and we agree that the CSS specification is ambiguous here.  We are working to provide a fix.

-----Original Message-----
From: Erik Dahlstrom [mailto:ed@opera.com] 
Sent: Tuesday, August 17, 2010 6:22 AM
To: www-svg@w3.org; Patrick Dengler; Robert Longson
Subject: Re: Property inheritance and the 'clipPath' element

On Sun, 08 Aug 2010 14:13:43 +0200, Robert Longson <longsonr@gmail.com>
wrote:

> Patrick,
>
> The reason that
>
> http://samples.msdn.microsoft.com/ietestcenter/svg112/svg_harness.htm?

> url=./svg112/svg/chapter_14.2.svg
>
> only works in IE is that the colour that is returned is "rgb(0, 0, 255)"  
> by
> firefox rather than "blue". Clearly this is the same rgb colour and as 
> far as I am aware getComputedStyle is allowed to return colours in 
> either fashion.
>
> Can you fix the test so that it works when getComputedStyle returns an 
> rgb colour? Or alternatively point me to where it is specified that 
> getComputedStyle cannot return an rgb colour in this case.

For the record the above testcase has been checked into the SVGWG repository as:

     http://dev.w3.org/SVG/profiles/1.1F2/test/svg/masking-path-geometry-01-f.svg


I agree with Robert that the testcase should allow the various serializations that may occur, unless there's something in DOM 2 Style that says otherwise. A hex-color (#0000FF) could also be returned, which is what Opera currently does. I hope that a future DOM Style spec will require a specific serialization.

Cheers
/Erik

--
Erik Dahlstrom, Core Technology Developer, Opera Software Co-Chair, W3C SVG Working Group Personal blog: http://my.opera.com/macdev_ed

Received on Tuesday, 17 August 2010 20:24:03 UTC