Re: Conditional IDness considered unnecessarily complex

On Oct 12, 2007, at 11:49, Henri Sivonen wrote:
> In the interest of practical, simple and high-performance  
> processing, I suggest making the IDness of id unconditional. If  
> this means that you can't duplicate the value in xml:id, the fix is  
> not to duplicate the value in xml:id.

The original idea was that both attributes would be unconditionally  
ID, and that it was recommended to duplicate the value. XML Core  
objected to that on the grounds that it makes SVG impossible to  
validate with DTDs. That, of course, is a bad reason given that it  
already isn't. However given XML Core's clout with all things XML,  
despite them being dead wrong on this, the SVG WG decided to go with  
their suggestion.

Feel free to take this up with XML Core, maybe they've since changed  
their view.

-- 
Robin Berjon - http://berjon.com/
------------------------------------------------------------------------
"RDF: codifying vagueness in the strictest possible way"
                         -- Kip Hampton

Received on Monday, 15 October 2007 10:33:02 UTC