Re: Concern about -webkit-mask

(2012/04/18 14:50), Robert O'Callahan wrote:
> SVG masks have the unfortunate
> properties that they require an external mask element and they use
> luminance instead of alpha to define the mask operation, so I think
> something like -webkit-mask is also useful. Someone should look at the
> ways people are using -webkit-mask, then ideally we can extend SVG
> 'mask' with enough of the -webkit-mask syntax to cover that usage (and
> for that syntax, mask using alpha instead of luminance); if not I guess
> we'll have to define a new property with the functionality of -webkit-mask.

The SVG WG is already looking at using alpha for masks in SVG 2. See [1] 
and [2]. The result might look something like:

   mask: <funciri> | none | alpha(<iri>) | luminance(<iri>)

(Possibly also add 'child-alpha | child-luminance' to that list since we 
want to remove the dependence on IDs.[3])

I think the requirement for a <mask> element is also worth 
reconsidering. (Actually, the SVG spec is a bit confusing in this regard 
since at one point it actually says that a mask property simply refers 
to "another graphical object"[4])

Regards,

Brian

[1] http://www.w3.org/Graphics/SVG/WG/wiki/Proposals/Alpha_mask
[2] http://www.w3.org/2012/01/12-svg-minutes.html#item01
[3] http://lists.w3.org/Archives/Public/www-svg/2012Feb/0071.html
[4] http://www.w3.org/TR/SVG11/masking.html#MaskProperty

Received on Wednesday, 18 April 2012 06:41:37 UTC