W3C home > Mailing lists > Public > www-dom@w3.org > July to September 2009

D3E draft, comments, ACTION-373

From: Olli Pettay <Olli.Pettay@helsinki.fi>
Date: Wed, 08 Jul 2009 23:48:06 +0300
Message-ID: <4A550606.3080209@helsinki.fi>
To: www-dom@w3.org
Hi all,

some comments about the D3E draft.

1.2
The graph should probably have DefaultView, not Window. Otherwise the 
spec needs to refer to html5 or wherever the Window object is defined.


"What if an element is removed from the document altogether? Is there 
different behavior if the event target is removed versus one of its 
ancestors?"
As said earlier
"Once determined, the propagation path cannot be changed".
So it doesn't matter if the DOM tree is changed somehow,
propagation path stays the same.
(I call it "event target chain".)


"The DOM event flow is an application of this model: the propagation 
path for a Node object is determined by its Node.parentNode  chain, and 
all events accomplish the capture and target phases"
If we are going to have defaultView in the graph above, this text should 
probably mention something about it. Basically that from document the 
event propagates to defaultView if defaultView implements
EventTarget interface.
There is the problem with load event not propagating to defaultView, but 
that is pretty HTML specific.


1.3
"...but in exceptional cases also immediately before the event is 
dispatched"
Perhaps worth to mention an example about this behavior.
<input type="checkbox">'s .checked handling comes to mind.
.checked is changed just before 'click' event is dispatched, and if
default action is prevented, .checked is set to its original value.



1.4
needs some work :)


1.6.1
What is the use case for canDispatch()?
I'd expect all the implementations to return always true.
Or what does the "implementation can generate" mean?


1.7
"The DOM will not attempt to define all possible events."
Maybe "This document does not attempt..." or something similar


I wonder what is the reason to change focus and blur to be UIEvents.


1.7.4.
"@@ does the dblclick event cancel the click event?"
no.


1.7.5
- Should this be informative?
- wheelDelta is underspecified, IMO.


I'll continue from 1.7.6


-Olli
Received on Wednesday, 8 July 2009 20:48:43 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 22 June 2012 06:14:02 GMT