Re: [xml-dev] Towards XML 2.0

[David Carlisle <davidc@nag.co.uk>]

On 06/12/2010 00:08, Michael Kay wrote:
> I was assuming that if <xml:comment> were chosen to represent comments,
> its main (perhaps only) difference from other elements would be that it
> would be ignored for validation purposes - i.e it's not under schema
> control.

You say "other elements" but wouldn't you want it to generate a comment 
node rather than an element node, so not be selected by xpath * for example?

If you have a schema that says every child of x is y, and a valid input, 
I'd have thought that you'd want to know in advance that everything 
selected by x/* was a sequence of y nodes. If xml:comment generates 
element nodes that are ignored on validation, this assumption would be 
false.

>
> I've been interested to read people's views that comments should be
> allowed to contain "non-well-formed" content. It's a worthwhile point,
> but I don't find it compelling. The editing tool that creates the
> comment can always escape any markup within the comment.

Commenting is the easy bit, but if the editing tool is to also able to 
uncomment regions, then it needs to fix up broken markup, or not allow 
you to uncomment or allow you to generate non well formed document or 
something.

David