W3C home > Mailing lists > Public > public-cdf@w3.org > January 2006

CDR: which elements should support the ReferencingElement interface?

From: Maciej Stachowiak <mjs@apple.com>
Date: Mon, 2 Jan 2006 01:46:20 -0800
Message-Id: <172DFE4C-D2B2-4FF6-84EA-359364245AE1@apple.com>
To: public-cdf@w3.org


The spec completely shies away from defining what inclusion-related  
elements should define the ReferencedElement interface. Instead it  
seemingly leaves this up to compound document profiles and individual  
language specs. Where the individual language has already specified,  
this seems reasonable. However, where it hasn't, this seems like a  
bad idea. For example, a hypothetical HTML+SVG profile should support  
the interface on the same HTML elements as plain HTML (for HTML-in- 
HTML inclusion) or HTML+MathML. It would present significant  
difficulties if different profiles specified this differently.  
Therefore, I recommend that the specification should determine how  
and whether the interface applies to current inclusion mechanisms in  
w3c interaction domain languages.

Off the top of my head, I can think of the following ways that a  
document may include another:

HTML4 / XHTML1:

- object
- frame
- iframe
- img (if the included content is SVG - this is mandated by SVGT 1.2  
for UAs that support both SVG and HTML)

SVG:

- animation (only other SVG)
- use (only other SVG)
- foreignObject
- image (inclusion of another SVG is disallowed by the current draft  
but it does say that "the resource referenced by the 'image' element  
- represents a separate document which generates its own parse tree  
and document object model
(if the resource is XML).")

Any language that supports CSS styling:

- any elemenent with a CSS background image that is SVG (mandated by  
SVGT 1.2)

This is not necessarily an exhaustive list, just the ones I thought of.

Please address this in the CDR spec, because it is not appropriate  
for different profiles to make different requirements here. In some  
cases individual language specs may sufficiently specify, in which  
case just normatively reference them.

Regards,
Maciej
Received on Monday, 2 January 2006 09:46:31 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 8 January 2008 14:10:40 GMT