[whatwg] [wf2] 8.1. Relation to the CSS3 User Interface module

[Ian Hickson]

On Tue, 22 Mar 2005, Anne van Kesteren wrote:
> 
> :checked - according to Selectors this applies to HTML4 elements which 
> could have a SELECTED attribute, OPTION, a descendant from the SELECT or 
> DATALIST elements, is one of them, but is not listed.

This is a bug in the Selectors spec. I'll raise it with the CSSWG.


> :indeterminate - I suggest that the wording is changed to say that it 
> does not match any Web Forms 2 form controls as Selectors seems to 
> suggest that it does match certain HTML controls in a certain state.

Selectors doesn't mention HTML controls in its :indeterminate definition. 
There seems to be no problem here.


> :default - this psuedo-class should match more than just the inital 
> submit button. (I assume it also matches when that button is disabled?) 
> For example:
> 
>  <select>
>   <option selected>foo
>   <option>bar
> 
> The OPTION element with value "foo" should match :default as it is the default
> option in that form control. (When there is a <select multiple> more options
> could match :default.)

Since there is no way to determine (e.g. from the DOM) which elements were 
originally selected, I don't see any way to define this. I also question 
the usefulness of that -- GUIs, at least those I have seen, don't 
highlight the original selection in that way.


> Also, how does :default work with non form controls. Would 
> 'html:not(:default){background:lime}' really give a green background in 
> HTML documents?

Yes. :not() simply negates the result -- if :default doesn't match, 
:not(:default) does.

Note that :enabled is not the same as :not(:disabled), for this very 
reason.

-- 
Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'