[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 #143 from Olli Pettay <bugs@pettay.fi> ---
(In reply to Koji Ishii from comment #141)
> With so much thanks to Olli, wchen, and Hayato, I've experimentally
> implemented the proposed algorithm, and now I'm seeing two different results
> in our tests.
> 
> The two are minor, but since they are changes in the behavior not discussed
> in this thread yet, any opinions/insights from anyone is also appreciated.
> 
> 1. The proposed algorithm stops event path for orphaned nodes
> When target nodes are not distributed, the event path in the current spec
> bubbles up to Document, but the proposed algorithm stops. Example:
> 
> A
>   SR
>   B
> 
> Note that SR does not have IP. When an event fires on B, the event path is:
> Current: B, SR, A, Document, Window
> Proposed: B


But shouldn't B's destination insertion points be empty and from B event
propagate to A.
Or do I misunderstand your example?
(B is a child node of A)


> 
> This also occurs when younger shadow roots do not have any shadow insertion
> points.
In which case exactly? If you dispatch event to a node in older shadow tree and
younger
doesn't have insertion point, event should propagate only to the older shadow
root.


> 2. Events that are Always Stopped at...where?
> The current spec of "Events that are Always Stopped" says "...stopped at the
> root node of the node tree"
> http://w3c.github.io/webcomponents/spec/shadow/#h-events-that-are-always-
> stopped
> 
> Current our implementation allows older and younger shadow roots to receive
> the event and stops before shadow hosts.
hosts? plural?

mousemove _within_ older shadow dom shouldn't propagate to the younger shadow
dom

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

Received on Thursday, 12 February 2015 12:49:00 UTC