Re: color-interpolation

Hi Cameron,

I've reviewed the test and it looks fine to me.

Attached to this email is the test with a minor add on which has the word "FAIL" 
in fill correct colour of each sub tests (except for auto). If a colour in one 
of the subtests is wrong, the word "FAIL" becomes visible.

Feel free to add this to the current test.

Cheers,

Anthony

Chris Lilley wrote:
> On Thursday, August 6, 2009, 7:41:04 AM, Anthony wrote:
> 
> AG> Hi Cameron,
> 
> AG> Cameron McCormack wrote:
>>> Consider the following document:
> 
>>>   <svg xmlns="http://www.w3.org/2000/svg">
>>>     <rect width='200' height='100'/>
>>>     <g fill='white' fill-opacity='0.5'>
>>>       <rect width='100' height='100'/>
>>>       <rect x='100' width='100' height='100' color-interpolation='linearRGB'/>
>>>     </g>
>>>   </svg>
> 
>>> Should the right half of the rectangular region be a different colour
>>> from the left half?  IIUC, the use of ‘color-interpolation’ should
>>> affect how the right white <rect> is composited into the black <rect> in
>>> the background.
> 
> 
> AG> Yes. The should be different colours because the 'color-interpolation' affects
> AG> how the right <rect> is composited on to the background.
> 
> Right. (it also affects things like gradients).
> 
>>> Batik, Opera and Firefox all show a uniform #808080 colour in the
>>> rectangular region, which makes me wonder if I’ve used the
>>> ‘color-interpolation’ correctly.
> 
> 
> AG> Opera 10 shows a uniform #7E7E7E for me.
> 
> AG> To my understanding:
> AG>   - the left <rect> should be #7F7F7F
> AG>   - the right <rect> should be #BBBBBB
> 
> Yes.
> 
> So, lets have that slapped into a test.
> I added all the other values for color-interpolation to the attached.
> 
> 
> 
>