- From: Tab Atkins Jr. <jackalmage@gmail.com>
- Date: Wed, 26 Feb 2014 15:26:38 -0800
- To: Cameron McCormack <cam@mcc.id.au>
- Cc: www-style list <www-style@w3.org>
On Wed, Feb 26, 2014 at 3:09 PM, Cameron McCormack <cam@mcc.id.au> wrote: > Tab Atkins Jr. wrote: >> Because it's harder and slower to do a tree-walk like that, where >> jumping to an ID-scoping root is cheap. > > For rules in the <style scoped> you don't need to search. You will have to search, unless you're saying that rules in *another* <style scoped> with the same scoping element can't refer to the scoped name. That would be inconsistent with allowing it in style='', and inconsistent with what makes sense in a shadow root. It's also just more difficult for us - we already have the data structures set up to handle scoping names to shadow roots, but not arbitrary elements. Then again, we also don't have scoped stylesheets right now. > For style="" > attributes on elements within a style scope, you are right that you won't > know what your scope element is (although in Gecko we do know that the > element is in a style scope) and will have to traverse up to find it. While we're talking about style='', this seems fragile. If this allows arbitrary descendants to opt in, then it means that you can accidentally co-opt it by inserting *another* scoped stylesheet between the intended stylesheet and the element reffing the name. Sticking to just ID-scoping roots avoids this, because those are very intentional boundaries that are visible across the whole document, not accidental ones meant to apply only to the contents of a single stylesheet. ~TJ
Received on Wednesday, 26 February 2014 23:27:25 UTC