- From: Tab Atkins Jr. <jackalmage@gmail.com>
- Date: Mon, 29 Jun 2009 07:53:34 -0500
- To: Shira Gutgold <gutgold@yahoo.com>
- Cc: www-style@w3.org
On Fri, Jun 26, 2009 at 2:58 AM, Shira Gutgold<gutgold@yahoo.com> wrote: > Hello all, > I've been trying to ascertain whether the :parent selector and previous > siblings selector will be included in CSS3. > Reading the mailing list correspondence, I got the impression that the > :parent selector will be implemented but that the implementation of a > previous siblings selector would be too costly and therefore probably > abandoned. > So first of all, could someone please confirm whether my impression is > indeed correct? I can't absolutely confirm, as I'm not an implementer, but yeah, that was the impression I got from Boris. Parent selector wouldn't be *too* bad performance wise, and *maybe* previous adjacent sibling, but ancestor and previous sibling are right out at the moment. > Assuming the :parent selector will be implemented, I suppose a possible > workaround for targeting previous siblings would be > to style all the children of that parent like so: > a.classname :parent a{ > } > and then remove the style for the following selectors like so: > a.classname ~ a{ > } Except for the fact that the selector can't work quite like that (you're targeting a descendant of a.classname which is a parent of a.classname), yeah, that'd be perfect. It's just like the workaround you have to do to simulate :first-child. The actual parent selector would be something like: :parent(a.classname) a Though in this case it seems more appropriate to call it :child() or :has-child(), as the general pattern of functional pseudoclasses is to name them after what they're selecting off of. ~TJ
Received on Monday, 29 June 2009 12:54:39 UTC