Re: Navigation Timing implementation report

On 4/11/12 11:09 AM, Philippe Le Hegaret wrote:
> We have 3 failures on firefox, 1 failure on IE9.

The document.open issue is a bug in Firefox.

The test_timing_attributes_order failure in Firefox is due to an issue 
in the test, as I pointed out earlier in the list.  The test is making 
assumptions about the precise ordering of load event terminations that's 
not warranted given the current state of implementations and other 
specifications.  Specifically, it assumes that the firing of a load 
event for a <frame> happens after the default action phase of the load 
event for the window inside that frame.

Unless this assumption is somehow critical to navigation timing (and 
then it should be explicitly spelled out in the navigation timing spec 
itself), it would be better to not make this assumption in the test.

I haven't looked at test_navigation_type_reload in detail, but at first 
glance it looks like the same issue as test_timing_attributes_order.

-Boris

Received on Wednesday, 11 April 2012 16:56:23 UTC