Re: [filter-effects] feGaussianBlur and stdDeviation = "0"

On Tue, 26 Feb 2013 19:25:47 +0100, Tab Atkins Jr. <jackalmage@gmail.com>  
wrote:

> On Tue, Feb 26, 2013 at 9:13 AM, Dirk Schulze <dschulze@adobe.com> wrote:
>> Hi FXTF,
>>
>> I would like to change the behavior of feGaussianBlur filter primitive  
>> if stdDeviationX = stdDeviationY = 0. If both radii are 0, UAs must  
>> handle the filter as error and refuse drawing. This seems to be the  
>> behavior on all browsers at the moment.
>>
>> For authors it often makes more sense if feGaussianBlur just returns  
>> its input - unmodified. Note that this is the case for the shorthand  
>> filter functions already (blur() and drop-shadow()).
>>
>> This affects feDropShadow as well - a shorthand for different filter  
>> primates including feGaussianBlur. It is not possible to draw a  
>> non-blurred drop shadow without this change.
>>
>> If there is no objection then I modify the spec.
>
> This sounds like a silly restriction, and your change is completely
> reasonable.  As long as no other impls object, I definitely support
> this.

This is already the case AFAIK, see [1]. Or am I missing something? To me  
it makes perfect sense to make the feGaussianBlur primitive a no-op for  
when stdDeviationX = stdDeviationY = 0, but it doesn't make sense to  
ignore the whole filter chain if a filter happened to contain such a  
primitive.

We should however consider clarifying that feMorphology with its radius  
attribute[2] set to 0 should also be treated as a no-op.


[1]  
https://dvcs.w3.org/hg/FXTF/raw-file/tip/filters/index.html#feGaussianBlurStdDeviationAttribute
[2]  
https://dvcs.w3.org/hg/FXTF/raw-file/tip/filters/index.html#feMorphologyRadiusAttribute

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

Received on Thursday, 28 February 2013 16:12:01 UTC