Re: [selectors4] :not(a, b) vs. :not(a):not(b)

On Mar 19, 2013 9:17 AM, "Peter Moulder" <peter.moulder@monash.edu> wrote:
>
> On Tue, Mar 19, 2013 at 10:27:32AM +0100, Simon Sapin wrote:
>
> > Selectors 4 extends :not() so its argument is a selector list. As I
> > read it, it’s just syntaxic sugar for multiple :not()
> > pseudo-classes. In other words, :not(a, b) is equivalent to
> > :not(a):not(b), just like not(a or b) is not(a) and not(b) in
> > boolean logic. Is this correct?
>
> They differ in specificity: :not(a):not(b) has the sum of the
specificities of
> a and b, while that of :not(a, b) is currently defined as the max of the
two.
>
> There's an issue open as to whether the specificity of :matches should
change
> from max specificity to something else, though that issue was raised
before
> :not was changed to take a selector list, so there isn't yet a
corresponding
> proposal as to how or whether the specificity of :not(a, b) might change
if
> that proposal for :matches(a, b) were to be adopted.
>
> > If so, it might be worth pointing out in the spec. Maybe in the
> > html|*:not(:link):not(:visited) example, by saying that it can also
> > be written html|*:not(:link, :visited)
>
> Nevertheless, that might still be a good idea, while noting the
difference in
> specificity.
>
> pjrm.
>

I think i've noted this in the past but that seems like a false choice
(specificity based on greatest).  It seems the choices should be between
simple pseudo specificity and multiply registering the same rule with
multiple specificities.