[Bug 23887] [Shadow] Put only the final destination insertion point to the event path

https://www.w3.org/Bugs/Public/show_bug.cgi?id=23887

--- Comment #161 from Hayato Ito <hayato@chromium.org> ---
(In reply to Olli Pettay from comment #158)
> (In reply to Olli Pettay from comment #157)
> > Ah, you mean that. It is indeed then possible that the algorithm needs some
> > tweaking, since it was written
> > during the time when older shadow trees not distributed to any shadow
> > insertion point were thought to be not-in-document.
> (and that is in fact still the behavior in Gecko, since it kind of makes
> sense that one can sort of replace the older shadow tree. Going through bug
> 26365 to see why the spec has different definition atm.)


Yeah, that's a important (and confusing) point.

Recently, to makes things simple, I've tried to think as follows:

- Insertion points don't have any effect on *DOM connectivity*.
- Insertion points have an effect on how style inheritance is calculated. As a
result, some nodes would became 'display: none', however, they are still
considered as *connected*.
- From this point of view, we shouldn't call the older shadow tree *orphaned*
even when the younger shadow tree doesn't have a shadow insertion point.
  It's something like: 'display: none' is set to the *pseudo parent element* of
(child nodes of the older shadow root).


I've tried to make the current spec consistent with this concept, recently.
However, I agree we should say it more explicitly (and in a friendly way),
maybe in a non-normative section.

-- 
You are receiving this mail because:
You are the QA Contact for the bug.

Received on Tuesday, 17 February 2015 07:56:09 UTC