Re: Standardizing de-facto behavior built on an optional or undefined specification

James Hopkins wrote:
> I've recently come across several sections of the CSS 2.1 spec [1] which 
> describe behavior that is either intentionally undefined or optional 
> (per RFC2119), and have already commented on one of these issues in a 
> previous email [2].
> 
> After some testing, and in both examples, the four main browsers (FF, 
> Safari, IE, & Opera) have adopted comparable implementations, and appear 
> to exhibit identical behavior. In these cases, I believe it would be 
> beneficial to standardize this established behavior in order to prevent 
> possible interoperability issues for authors (who have become reliant on 
> it), which may arise from retaining the definitions found in the current 
> specification.
> 
> [1] Scaling of replaced elements 
> (http://www.w3.org/TR/CSS21/visudet.html#the-width-property), position 
> and tiling of 'background-image' applied to inline elements 
> (http://www.w3.org/TR/CSS21/colors.html#background-properties)
> 
> [2] http://lists.w3.org/Archives/Public/www-style/2009Nov/0324.html

Replied to the first issue...

For the second, this has been defined in CSS3 by giving it a control:
   http://www.w3.org/TR/css3-background/#the-box-decoration-break
I don't think it's necessary to backport the definition to CSS2.1.

~fantasai