Will Moz-Xform be able to support some form of deep copy?

Dear All,

I am trying to write an Xforms app which allows a user to add & change 
annotations written in XHTML where the annotation is contained within a 
textarea

If I use a repeat e.g.:

<xf:repeat id="r2" style="" nodeset="mif:class/mif:attribute">
						<tr>
							<td>
	
<xf:output ref="@name">
							</xf:output>
							</td>
							<td
style="width: 50%; text-align: left; vertical-align: top;">
	
<xf:textarea id="annotation" ref="mif:annotation/*" />
							</td>
						</tr>
					</xf:repeat>

Then I just get the first bit of text etc but the moment it rus into an 
xhtml element (e.g. <p> </p> then that's the end.


IF there is no scope for this in XForms now then there it will need to 
be added unless you are expecting all the users of forms to use plain 
text for writing in textareas.

At the moment if an annotation contains

"This is very <B>important<B/>"

They will only see "This is very".

It is not just doing XHTML editiing either. Imagine you have an SVG 
tree/structure within a model. How do you simply pass the text between 
<SVG> & </SVG> to a svg rendering tool with out having to use Javascript 
to examine the actual model & do all the owrk (which rather destroys the 
point of XForms).

i.e. any form of "multi format editor/form" will need it much like:

http://www.alphaworks.ibm.com/tech/cxde

So long as I can supply an external to XForms editor/viewer (JScript or 
Object/plugin) with the relevant fragment as a string /deep copy then it 
should be able to handle the display/editing of that fragement & pass it 
back for storage in the model.


Adam