W3C home > Mailing lists > Public > www-style@w3.org > April 2005

Re: Dynamic pseudo class cascade

From: Staffan Mhln <staffan.mahlen@comhem.se>
Date: Wed, 20 Apr 2005 20:45:13 +0200
To: Boris Zbarsky <bzbarsky@mit.edu>
CC: www-style@w3.org
Message-ID: <4266BF59.368.378F53@localhost>

On 19 Apr 2005 at 14:44, Boris Zbarsky wrote:
> > Yes, but an author writing:
> > <body link="blue">
> > probably will not consider that.
> 
> Quite frankly, there's nothing you can really do about pages like that.  Given 
> that markup, the links will also fail to show up if the UA happens to have a 
> default blue background, for example.  No need for dynamic anything.

Yes, the example might have been farfetched. However, such pages will 
proabably need to work anyway. The point was of course that even if 
you had an author writing:
a {color: black; background-color: white}
it would still override the :focus colors, so while colors are 
particularly hard to get useful, the way the cascade works might 
interfear with UAs possibilities for using any :focus settings.

> > Due to the dynamic nature of theese selectors i don't think it's
> > totally unreasonable to consider such a change. 
> 
> So what would you propose the change look like, exactly?

I'm not sure i'd say i am actually proposing it, but i saw no 
solution other than changing: 
http://www.w3.org/TR/CSS21/cascade.html#cascading-order

Possibly something like adding a dimension where any property that 
had dynamic pseudo-classes in its selector would be sorted to the end 
of:

"Sort by importance (normal or important) and origin (author, user, 
or user agent). In ascending order:

   1. user agent style sheets
   2. user normal style sheets
   3. author normal style sheets
   4. author important style sheets
   5. user important style sheets "
   6. Dynamic user agent
   7. Dynamic user normal
   8. Dynamic author normal 
   9. Dynamic author important 
   10. Dynamic user important 

 /Staffan
Received on Wednesday, 20 April 2005 18:45:12 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 27 April 2009 13:54:36 GMT