W3C home > Mailing lists > Public > public-web-perf@w3.org > May 2013

RE: navigationStart in NavigationTiming

From: Jatinder Mann <jmann@microsoft.com>
Date: Wed, 8 May 2013 16:50:18 +0000
To: "Boris Zbarsky (bzbarsky@mit.edu)" <bzbarsky@mit.edu>, "'arvind@google.com'" <arvind@google.com>
CC: "'public-web-perf@w3.org'" <public-web-perf@w3.org>
Message-ID: <989d9ff06df94a7bbd58943138174a1a@BLUPR03MB065.namprd03.prod.outlook.com>
In IE10, grabbing performance.timing from the console on an about:blank page gives me the following values:

unloadEventEnd		0
unloadEventStart	0
navigationStart	1368030410528
redirectEnd	0	
redirectStart	0
fetchStart	1368030410528
domainLookupStart	1368030410528
domainLookupEnd	1368030410528
connectStart	1368030410528
connectEnd	1368030410528
requestStart	1368030410528
responseStart	1368030410528
responseEnd	1368030410528
domLoading	1368030410528
domInteractive	1368030410537
domContentLoadedEventStart	1368030410539
domContentLoadedEventEnd	1368030410539
domComplete	1368030410539
loadEventEnd	1368030410541
loadEventStart	1368030410541

Here unloadEventStart/End are zero'd out, as there is no previous navigation, and redirectStart/End are zero'd out as there are no same origin redirects. All fetchStart, DNS, connect, request, and response values are equal to navigationStart, meaning zero deltas from navigationStart. The DOM processing attributes have updated timing values. If we are to define the expected about:blank behavior, the above approach seems to work.

However, it does look like Chrome and Firefox also define values here, but the values don't seem to match each other. Seeing that these timing values aren't very useful to site developers, we can just specify that user agents MUST define the attributes, to avoid throwing an exception, but add a SHOULD clause on the expected values. 

-----Original Message-----
From: Boris Zbarsky [mailto:bzbarsky@MIT.EDU] 
Sent: Wednesday, May 1, 2013 6:36 PM
To: Arvind Jain
Cc: Jatinder Mann; public-web-perf@w3.org
Subject: Re: navigationStart in NavigationTiming

On 5/1/13 9:12 PM, Arvind Jain wrote:
> I mean to say the behavior is unspecified for "about:blank"

I frankly see no reason to leave this unspecified.  Worst-case, just define it to all 0 for initial about:blank.

-Boris
Received on Wednesday, 8 May 2013 16:54:21 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:04:35 UTC