- From: Ian Hickson <ian@hixie.ch>
- Date: Tue, 25 Oct 2005 20:10:44 +0000 (UTC)
- To: Andrew Shellshear <Andrew.Shellshear@cisra.canon.com.au>
- Cc: www-svg@w3.org
On Tue, 25 Oct 2005, Andrew Shellshear wrote: > > On Fri, 20 May 2005 13:32:15 +0000 (UTC), Ian wrote: > > According to section C.4: > > > > # The outermost <svg> element must be defined in the SVG namespace (e.g., > > # <svg xmlns="http://www.w3.org/2000/svg">); otherwise the document is > > # in error. > > > > This seems like a bogus requirement. If the outermost <svg> element > > _isn't_ in the SVG namespace, then it isn't an <svg> element, and is > > out of the scope of this specification, surely. > > Because of the history, the SVG WG feels it is necessary to explicitly > say that the namespace declaration must be provided. If it were not for > this history, we would not include the sentence you dislike. I agree with the sentiment, but I feel the current text is just as misleading, because it is technically incorrect ("bogus", as I said earlier). Could you rephrase it to something like: User agents must only consider elements explicitly placed in the SVG namespace by XML Namespace declarations in the document (e.g., <svg xmlns="http://www.w3.org/2000/svg">) as being SVG elements. The current text is wrong on many counts: * The document might not be in error, it might be another language, unrelated to SVG. * All the SVG elements must be in the SVG namespace, not just the outermost element. * The <svg> element must be defined to be in the SVG namespace to be treated as SVG even if it isn't the root element, e.g. in compound documents. (I am not satisfied by the WG's current response.) Cheers, -- Ian Hickson U+1047E )\._.,--....,'``. fL http://ln.hixie.ch/ U+263A /, _.. \ _\ ;`._ ,. Things that are impossible just take longer. `._.-(,_..'--(,_..'`-.;.'
Received on Tuesday, 25 October 2005 20:10:51 UTC