Re: [css3-namespace] what is a "no namespace"

fantasai wrote:
> Christof Hoeke wrote:
>>
>> hello,
>> a comment to the last call of the CSS Namespace spec but actually I 
>> always wondered a bit if it could be explained a bit more clearly what 
>> "no namespace" means?
>>
>> --- another thing first:
>>
>> While writing this comment I originally had the following thought:
>>
>>     In XML no namespace means
>>      1. elements with no prefix and no default namespace in the 
>> enclosing document/element
>>      2. elements resetting a default namespace defined in the 
>> enclosing document/element with::
>>         <x xmlns="">...
>>
>> Using this simple logic it would be allowed to define::
>>
>>     @namespace empty "";
>>
>> to select element x from above with::
>>
>>     empty|x
>>
>> But in XML the empty string is not allowed when defining a namespace 
>> with a prefix (http://www.w3.org/TR/xml-names/#iri-use), only when 
>> defining the default namespace (if I understand 
>> http://www.w3.org/TR/xml-names/#defaulting right).
>>
>> So the above should not be valid in CSS either I guess?
>> The syntax does not prevent this though and no comment is there either 
>> (maybe I overlooked it?).
> 
> Added to the issues list. Anne or I will get back to you on this one.
>   http://dev.w3.org/csswg/css3-namespace/issues-2
> 
>> Anyway, I also think the term "no namespace" should be more explicit. 
>> In an very old draft 
>> (http://www.w3.org/1999/06/25/WD-css3-namespace-19990625/#tag-selector) 
>> I found the sentence:
>>
>>     The namespace component may be left empty to indicate that the 
>> selector is only to match elements with no declared namespace
>>
>> Maybe the single word "declared" could be re-added to 
>> http://www.w3.org/TR/css3-namespace/#css-qnames to make this clearer?
> 
> I borrowed some wording from Namespaces in XML and clarified it as
> 
>     The prefix of a qualified name may be omitted to indicate that the name
>     belongs to no namespace, i.e. that the namespace name part of the 
> expanded
>     name has no value.
> 
> Does that work for you?

It is much clearer now, just one small point. I am not sure if it would 
clarify it a bit more if one could add something like (I guess there is 
a better wording though):

	... that the namespace name part of the expanded name
	has no value or is the empty string.

Somehow if I read "no value" I think it has to be ``null`` or ``None`` 
but not "".

thanks
christof