[whatwg] IDness and warnings

[Henri Sivonen]

On Nov 6, 2006, at 11:06, Ian Hickson wrote:

> On Mon, 6 Nov 2006, Henri Sivonen wrote:
>>
>> Considering that xml:id support for Gecko has been blocked ( https://
>> bugzilla.mozilla.org/show_bug.cgi?id=275196 ), would it be a good  
>> idea
>> for a conformance checking service to emit a warning if an element in
>> the XHTML namespace has an attribute whose attribute type is ID  
>> but the
>> attribute does not have the local name "id" in no namespace?
>
> That's a judgement call and outside the scope of the WA1 spec -- as  
> with
> any warning, it would be helpful to authors who don't know what  
> they are
> doing, and annoying to those who do.

Agreed. I am on the fence with the judgement call, which is why I am  
interested in the opinion of the readers of this list. It has been  
suggested to me at least twice in private that I could consider not  
supporting xml:id at all (i.e. emit an error even). However,  
supporting xml:id is not that hard, since I already added an xml:id  
processor to the processing pipeline so anything keyed on IDness just  
works. (For referential integrity checking, I changed the Schematron  
assertions to use XPath id() and I am writing Java code that looks  
for IDness rather than the name "id".)

-- 
Henri Sivonen
hsivonen at iki.fi
http://hsivonen.iki.fi/