Re: Selectors Tests

Andrew Fedoniouk wrote:
> And what about fieldset:disabled selector then?

Shouldn't match, I would think.

> And how would you specify say this:
> 
>   legend:disabled { color:gray; }

What does it mean for a legend to be disabled?

> Ideally :disabled flag shall be "on" for all elements that
> have @disabled attribute by themselves or in one of their parents.

That seems like a discussion for the HTML working group, not this one.

>> In both cases :enabled basically means "can be submitted to the 
>> server" at the moment.
> 
> But <option> and <optgroup> are not submittable.

The former can have values sent to the server, though.

> In any case :enabled (as an UI state) is quite far from concept of 
> successful control [for submission] in HTML.

Actually, in HTML in particular they're more or less identical for most 
controls.  But in any case, that's an HTML issue, not a CSS one.

> It appears as set of :enabled elements is different from :disabled set
> so that ":enabled is what that could be :disabled" is not true all of 
> the time.

Uh...  It's true pretty much by definition at the moment.  Though again, 
the goal is to move to a different definition, one where the markup 
language will define what's in these states.

>>> Stays: "This [disabled] attribute is inherited but local declarations 
>>> override the inherited value."
>>
>> I have no idea what that could mean, and neither does anyone else as 
>> far as I can tell.
> 
> It means that "disabled" attribute is inheritable. Not more not less.

Yes, and since the idea of "inheritable attribute" is not defined in 
HTML it means absolutely nothing.  You could just as well say that the 
"disabled" attribute is grozflzsnork.  Again, take that up with the HTML WG.

>>> XTF... what about XBL then? Those <input>s in shadow trees... how 
>>> they interact with :disabled/:enabled thing?
>>
>> They match it depending on their state.
> 
> If :disabled is inheritable then yes.

Nothing to do with inheritability.  They just do what I said, based on 
code inspection.

Seriously, you need to get away from HTML here, especially because HTML 
doesn't actually define how this stuff should work for it (unlike, say, 
XForms, which does).

-Boris

Received on Sunday, 9 November 2008 17:20:12 UTC