Re: [css-regions] Changed @region rule to ::region() pseudo-element

On May 20, 2013, at 7:02 AM, François REMY <francois.remy.dev@outlook.com> wrote:

>>> ::region(>strong) {
>>> abc: def;
>>> abc: def;
>>> @where :hover {
>>> abc: def;
>>> abc: def;
>>> }
>>> }
>> 
>> I take this to mean '::region(>strong) :hover'. But how to do the embedding for '::region(>strong):hover' or '::region(>strong :hover)' or '::region(>strong:hover)'? The @rule didn't suffer this lack of clarity.
> 
> You're right. I noted this issue before, but didn't notice it impaired nesting.
> 
> So, if we want a pseudo element and justify this with nesting possibilities, we will need a pseudo-element without the parentheses (ie: a combinator-like pseudo-element), or we should use a true combinator or an at-rule instead.
> 
> 
> The advantage of the combinator is that you can use both the independent and grouped syntaxes:
> 
>   my-region ... region-content {}
>   my-region ... region-content-2 {}
> 
> and
> 
>   my-region { 
>      @then ... region-content {}
>      @then ... region-content-2 {}
>   }
> 
> and
> 
>   my-region {
>      @then {
>          region-content {}
>          region-content-2 {}
>      }
>   }
> 
> while with @region you are forced to use the second alternative
> 
>   @region my-region {
>     :scope region-content {}
>     :scope region-content-2 {}
>   }
> 
> (note that it may be seen as an advantage by some)

I agree that the combinator approach looks and works better than the pseudo-element approach. But it seems a bit overkill to create a whole new combinator, just for use with regions only. 

Also, of the examples you gave, the @region version had the best combination of compactness (once you remove the unnecessary ':scope' parts) and DRY-ness. And to my eye is the clearest.