[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 #151 from Hayato Ito <hayato@chromium.org> ---
(In reply to Olli Pettay from comment #149)
> (In reply to Hayato Ito from comment #147)
> > (In reply to Koji Ishii from comment #145)
> > > The case where younger shadow roots do not have shadow insertion points:
> > > 
> > > A
> > >   SR1
> > >     B
> > >   SR2
> > > 
> > > When an event fired on B:
> > > Current: B, SR1, A, Document, Window
> > > Proposed: B, SR1
> > 
> > I can't agree this proposal because this violates the design principle of
> > Shadow DOM event dispatching of the following:
> No it doesn't
> 
> > 
> > "Attaching a Shadow Root to an element shouldn't have any effect for an
> > event dispatching in the node tree which the shadow host participates in,
> > basically"
> B is a child of SR1 in this case, not part of the node tree of A.
> 
> > 
> > In the proposal, attaching the younger shadow root would affect the event
> > dispatching in the document tree. That would violated the principle.
> 
> No, B is a child of SR1.
> (or I misunderstand where B is.)


Let me  make it clear what I said as follows:

- Attaching a shadow root to any node in any node tree doesn't affect an event
dispatching which will happen on any node, in the every node trees, except for
the newly attached shadow tree, basically.



In the example case, we can interpret this as follows:

- Attaching a shadow root (SR1) to any node (A) in any node tree (document
tree) doesn't affect an event dispatching which will happen on any node (B), in
the every node trees (both document tree and SR1 tree), except for the newly
attached shadow tree (SR2 tree), basically.


"Basically" here means "as long as we don't call event.stopPropagation() in the
newly attached shadow tree".

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

Received on Monday, 16 February 2015 01:50:54 UTC