- From: Anne van Kesteren <fora@annevankesteren.nl>
- Date: Tue, 29 Nov 2005 09:36:49 +0100
- To: Andrew Shellshear <Andrew.Shellshear@cisra.canon.com.au>
- Cc: kono.masahiko@canon.co.jp, www-svg@w3.org
Quoting Andrew Shellshear <Andrew.Shellshear@cisra.canon.com.au>: >>> We put them on SVGSVGElement and not SVGDocument (or Document) >>> because there might not be an SVGDocument. This can happen when we >>> are in a multi-namespace XML application, and SVG is just one >>> component of it, such as in CDF scenarios - here, the root element >>> of the XML document isn't the svg element, and there is no >>> SVGDocument. >> >> How does SVGDocument depend on the root element of the document? It >> might very >> well be that the document object supports all specialized document >> objects. So >> the document object has both HTMLDocument and SVGDocument implemented. (I >> believe this is the case in some UAs, at least.) > > It *could* be the case that there is no SVGDocument at all. It's > allowed. So we can't put focus-related methods on it. (Sorry if I'm > missing the point, here.) What do mean with "It's allowed"? And since you're talking about CDF scenario's, what should actually happen when there are sibling svg:svg elements? Does this model still make sense? Wouldn't it be easier to require that when elements in the document appear to be in the SVG namespace SVGDocument must be implemented on the same object that implements the Document interface defined in DOM Level 3 Core? >> Also, this answer does not say if the second question, regarding >> nested svg:svg >> elements, is addressed. > > True. This case applies only to SVG Full, of course - an SVG > implementation inside a foreignObject will be a separate entity, and > won't interact with the "parent" svg. I suspect in the Full case, > any calls to setFocus or moveFocus would not cross the SVGSVGElement > boundary - they would only apply to descendants of this SVGSVGElement > (stopping at any child SVGSVGElement). SVG Full 1.2 is a work in > progress, and we will address this comment before the next published > draft. Ok, thanks. -- Anne van Kesteren <http://annevankesteren.nl/>
Received on Tuesday, 29 November 2005 08:37:10 UTC