RE: [User Timing] New test cases

Pan,

I have reviewed the User Timing test cases.


-       test_user_timing_entry_type.html.

The current approved test cases don't cover testing the class names of the entries. This is a good addition to the test suite.



-       test_user_timing_mark_and_measure_exception_when_invoke_without_parameter.html

The current approved test cases don't test the scenario where mark and measure are called without parameters. This is a good addition as well.



-       test_user_timing_mark_with_name_of_navigation_timing_optional_attribute.html

My concern with this test case initially was that secureConnectionStart is an optional parameter. Though we can test for this, it should be clear that it is an optional test. I think the current test text makes that clear.

I recommend we move the above three test cases to the approved folder. Considering the other test cases overlap with the current approved tests, I recommend we don't take the others.

Thanks,
Jatinder

From: Deng, Pan [mailto:pan.deng@intel.com]
Sent: Sunday, February 3, 2013 11:59 PM
To: Jatinder Mann; James Simonsen; public-web-perf
Subject: RE: [User Timing] New test cases


Hi Jatinder,



Refer to another thread, I prepared three User Timing test cases that not covered by the approved, some of them are already in WebKit test set.



http://w3c-test.org/webperf/tests/submission/Intel/user-timing/test_user_timing_entry_type.html

This case tests the timing entry type "PerformanceMark" and "PerformanceMeasure"

http://w3c-test.org/webperf/tests/submission/Intel/user-timing/test_user_timing_mark_and_measure_exception_when_invoke_without_parameter.html

This case tests the invalid call of performance.mark() and performance.measure()

http://w3c-test.org/webperf/tests/submission/Intel/user-timing/test_user_timing_mark_with_name_of_navigation_timing_optional_attribute.html

This case tests mark exception of "secureConnectionStart", I think for the UA that supports "secureConnectionStart" attribute in Navigation Timing, this test is a must to fully conform to User Timing Spec.



Both chrome and IE10 pass these tests. (Ignore vendor prefix problem).

Jatinder and James, could you please review them? :)



Thanks

Pan


From: Jatinder Mann [mailto:jmann@microsoft.com]
Sent: Thursday, January 24, 2013 6:45 AM
To: Deng, Pan; James Simonsen; public-web-perf
Subject: RE: [User Timing] New test cases

I reviewed these tests cases. It seems like these tests were targeting an older version of the spec, as most of the tests are hitting script errors in IE10 trying to call window.performance.mark() and window.performance.measure(). These functions were removed from the User Timing spec prior to taking the spec to CR.

I also noticed that there is a test testing mark("secureConnectionStart"), even though secureConnectionStart is an optional attribute in Navigation Timing.

Seeing that most of these tests overlap with the already approved User Timing tests, http://w3c-test.org/webperf/tests/approved/UserTiming/, I don't think we'll need to make any updates here though.

I'd like to thank Pan for spending the time writing these tests cases. Thanks!

Jatinder

From: James Simonsen [mailto:simonjam@chromium.org]
Sent: Wednesday, October 24, 2012 10:51 AM
To: public-web-perf
Subject: [User Timing] New test cases

Hi all,

I've uploaded the User Timing tests written by Pan Deng from Intel. You can find them here:

http://w3c-test.org/webperf/tests/submission/Intel/user-timing/

Please take a look!

James