This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
5.1 Events that are Always Stopped[1] states that: > The following events must always be stopped at the root node of the node tree While by following the algorithm in the event path calculation algorithm in 5.2 Event Paths[2], it includes "the destination insertion points of CURRENT". Either the wording of 5.1 or the algorithm need to change. [1] http://w3c.github.io/webcomponents/spec/shadow/#events-that-are-always-stopped [2] http://w3c.github.io/webcomponents/spec/shadow/#dfn-event-path-calculation-algorithm
The algorithm is correct and intentional: We should fix the wording like this: > The following events must always be stopped at the root node of *youngest shadow tree*.
Please see bug 20247 comment 10.
(In reply to Anne from comment #2) > Please see bug 20247 comment 10. Thanks, but I'm afraid bug 20247 is an orthogonal issue. Even when we can resolve bug 20247, this issue, *where* we should stop an event, still apply. I'm totally supportive for bug 20247 comment 10, but I've not started on working bug 20247 yet...
Fixed at https://github.com/w3c/webcomponents/commit/2a3b7ed34bdbfb8ab095c914d41bf22703a117f5
Let me add a note in more reader friendly way here: (1) "Events that are always stopped" are stopped at the boundary between the *youngest* shadow root and it's shadow host. (2) "Events that are always stopped" are *NOT* stopped at the boundary between the older shadow root and the shadow insertion points in the younger shadow root. This situation, (2), is very similar to the following situation: (3) "Events that are always stopped" are *NOT* stopped at the boundary between the child node, A, of the shadow host and A's destination insertion point. In other words, the sub-path, from the youngest shadow root to it's shadow host, is the only moment where an event is propagating from "Child Component" to "Parent Component". (2) and (3) are not. I might add this note to the spec itself as a non-normative note.