- From: Andrew Shellshear <Andrew.Shellshear@cisra.canon.com.au>
- Date: Thu, 01 Dec 2005 21:27:50 +1100
- To: www-svg@w3.org
On Fri, 20 May 2005 12:17:33 +0000, Ian Hickson wrote:
> The SVG 1.2 spec currently uses the terms "outermost svg element",
> "root-most svg element", and other such language, without, as far as I can
> tell, ever defining those terms.
>
> In light of other requirements in the specification and articulated in
> response to other last-call comments, it seems that there in fact an <svg>
> element can never have another such element as an ancestor. However, it is
> not entirely clear if that is the case, for example, one could imagine an
> SVG fragment in an XHTML fragment in a <foreignObject> in an SVG fragment,
> where one would see an <svg> element that is a descendant of another.
>
> Please replace all references to the "outermost svg element", "root-most
> svg element", and other such language, with a single, well-defined term,
> that handles these cases and clearly explains how a user agent must handle
> <svg> elements that are descendants of other <svg> elements.
The term we will use is "rootmost svg element".
Each instance of the term (which replaces all uses of "outermost svg element",
"root-most svg element", "root svg", etc.) now points to the Definitions section,
which defines it as:
rootmost svg element
The furthest 'svg' <http://www.w3.org/Graphics/SVG/Group/repository/spec/mobile/1.2/1.2NG/master/struct.html#SVGElement> ancestor element that does not exit an SVG context.
Note that this definition has been carefully chosen to be applicable not
only to SVG Tiny 1.2 (where the rootmost svg element is the only svg element,
except when there is an svg element inside a foreignObject) but also for SVG
Full 1.2 and SVG that uses sXBL.
See also SVG document fragment <http://www.w3.org/Graphics/SVG/Group/repository/spec/mobile/1.2/1.2NG/master/intro.html#TermSVGDocumentFragment>.
Thank you for your feedback. Please let us know if this does not address
your concerns.
Andrew.
Received on Thursday, 1 December 2005 10:29:58 UTC