W3C home > Mailing lists > Public > public-web-perf@w3.org > January 2012

[Navigation Timing] Action Item 76: Add a Navigation Timing test case to test NavigationStart

From: Karen Anderson (IE) <Karen.Anderson@microsoft.com>
Date: Mon, 16 Jan 2012 19:14:27 +0000
To: "public-web-perf@w3.org" <public-web-perf@w3.org>
Message-ID: <18D01621E7D58B4CA8BA9DDBD0D8B12C0384D08C@TK5EX14MBXC288.redmond.corp.microsoft.com>
Hi All,

After our call last week, I looked through our meeting minutes to get the details of Action Item 76: Add a Navigation Timing test case to test NavigationStart.  I've update the item with these notes:

According to meeting minutes: http://lists.w3.org/Archives/Public/public-web-perf/2012Jan/0020.html

This action item is related to the following:
18:21:08 [JatinderMann]There is another thread that discusses that we need a test that checks that navigationStart is recorded as "Immediately after the user agent prompts to unload the previous document" as stated in the documentation.
18:22:16 [JatinderMann] http://lists.w3.org/Archives/Public/public-web-perf/2011Nov/0023.html
James: Looks like a test case that we should add.

The email thread in question:
      In FF it seems navigationStart is measuring the timestamp after the previous page is loaded, not "Immediately after the user agent prompts to unload the previous document" as stated in the documentation.

      --Ricardo

Doesn't the existing check: window.performance.timing.unloadEventStart >= window.performance.timing.navigationStart cover this?

If we need this to be a tighter validation, we can collect the time just before exiting the onunload handler and compare that to navigationStart with some reasonable delta.  However, I think the point is to ensure order and placement of navigationStart which is currently present in the existing suite.  Am I missing something?

Thoughts?  If everyone agrees that this scenario is already covered, then I'll go ahead and close this action item.  Otherwise, I'm open for other feedback.

Thanks,
Karen
Received on Monday, 16 January 2012 19:15:07 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:04:32 UTC