RE: Action 99: Update test_timing_attributes_order test

Cycling this back around as this is one of the remaining open issues against moving Navigation Timing forward.  I think we should keep prerequisite tests separate from functionality testing of Navigation Timing.  Everyone else?

Thanks,
Karen

From: Karen Anderson (IE) [mailto:Karen.Anderson@microsoft.com]
Sent: Tuesday, May 15, 2012 3:15 PM
To: Tony Gentilcore; public-web-perf@w3.org
Subject: RE: Action 99: Update test_timing_attributes_order test and document_readiness_exists investigations

I started working on this and wonder if we really want to tie this test into Navigation Timing.  The motivation to add the test at TPAC<http://www.w3.org/2010/webperf/track/actions/68> was to create a test to verify the dependency on HTML5.  If we tie it into NT, UAs that have not implemented Navigation Timing yet will fail the test even though they do meet the dependency requirements.  If we don't, we still have the dependency verification, but do not guarantee the validity of the data.  I think this is fine, as we have taken the precedence in other tests to trust the browser to timestamp the attributes correctly and focus on validating proper order.

My vote is to leave the test as is.  What are everyone else's thoughts?

Thanks,
Karen

From: Karen Anderson (IE) [mailto:Karen.Anderson@microsoft.com]<mailto:[mailto:Karen.Anderson@microsoft.com]>
Sent: Friday, May 11, 2012 9:39 AM
To: Tony Gentilcore
Cc: public-web-perf@w3.org<mailto:public-web-perf@w3.org>
Subject: RE: Action 99: Update test_timing_attributes_order test and document_readiness_exists investigations

That sounds like a nice value-add.  I can work on that for next week.

Thanks,
Karen

From: Tony Gentilcore [mailto:tonyg@google.com]<mailto:[mailto:tonyg@google.com]>
Sent: Monday, May 07, 2012 6:50 PM
To: Karen Anderson (IE)
Cc: public-web-perf@w3.org<mailto:public-web-perf@w3.org>
Subject: Re: Action 99: Update test_timing_attributes_order test and document_readiness_exists investigations

Hi Karen, Thanks for the updates.

http://w3c-test.org/webperf/tests/submission/Microsoft/NavigationTiming/test_timing_attributes_order.htm

Looks good to me.

http://w3c-test.org/webperf/tests/submission/Microsoft/NavigationTiming/test_document_readiness_exist.html

This test checks that document.readyState has the right value at the right time. That's fine, but isn't related to Navigation Timing. To make it relevant, we should verify that each of performance.timing.{domLoading,domInteractive,domComplete} are 0 prior to reaching each readiness phase and are >0 after reaching it.

-Tony