273494 - If document.onmousedown is hooked, window.onmousedown is fired too

Status: RESOLVED INVALID

(Core :: DOM: Events, defect)

Description

[:aceman]

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.7.5) Gecko/20041107 Firefox/1.0
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.7.5) Gecko/20041107 Firefox/1.0

If document.onmousedown is hooked, window.onmousedown is fired too.
Is that the correct behaviour?

Reproducible: Always
Steps to Reproduce:
1. write the JS code:
2. document.onmousedown=your_func()
3. window.onmousedown=your_func()
4. press a mouse button on the page

Actual Results:  
The function will run twice.

Expected Results:  
I would expect it to run only once, because the document has caught the event
and should not pass (bubble) it forward to the window.

Of course, if some spec says otherwise...

The same happens with onmouseup.

For referrence, IE 5.5-6 only runs the function once, even if you hook all 4 events.

Comment 1

[Paul van Schayck]

mousdown does bubble:
http://www.w3.org/TR/2000/REC-DOM-Level-2-Events-20001113/events.html#Events-MouseEvent

"The mousedown event occurs when the pointing device button is pressed over an
element. This event is valid for most elements.

    * Bubbles: Yes
    * Cancelable: Yes"

Comment 2

[Gérard Talbot]

Reporter, please examine this page. 

http://www.webreference.com/programming/javascript/dhtmlref/chap6/2/3.html

Mozilla's event propagation is the same as NetScape 6 in the graphic 
http://www.webreference.com/programming/javascript/dhtmlref/chap6/2/fig61.gif
at that url. Note that event sent to the document are not bubbling to the window
object in MSIE.

Bubbling from the document object to the window object is not part of the DOM 2
Event interface or part of any standards. It's outside specs. So what does MSIE
and what does Mozilla is neither good or bad. For some events, like resize, the
event will be relayed to the window object in MSIE. 
FWIW, Opera 7.x often (never?) does not relay bubbling events to the window
object either.

Resolving as INVALID

Comment 3

[David Baron :dbaron: (⌚️UTC-4, no longer working on Mozilla)]

If we're breaking Web pages because we're doing something that's not required by
a spec, we should at least have a good reason for it -- and if we do, then the
proper resolution would be wontfix rather than invalid.

Comment 4

[:aceman]

comment 3: this page was optimized mainly for IE, but they probably wanted to be
sure so they hooked everything they could. They saw no harm on IE and opera, but
aren't aware of the problems on Mozilla. I will tell them about this, so don't
worry. And the function doesn't do anything serious, so the page isn't broken,
just annoying (shows things 4x instead of 1x).

You probably mean, we should have IE parity if nothing (any standard) prohibits
it. I would agree with that.

But I wanted to be sure what the standards say, so for me this bug is invalid
too. If you wish wontfix, we would have to change the summary to say '... do the
same as IE'.

Thanks guys.