[minutes] 20110323 Web Performance WG

Web Performance WG Teleconference #25

23 Mar 2011



See also: IRC log http://www.w3.org/2011/03/23-webperf-irc



Minutes

http://www.w3.org/2011/03/23-webperf-minutes.html



Attendees

Present

Plh, Jatinder, Mann, TonyG, Nic, Jansma, Karen, Anderson, Zhiheng, James, Cameron

Regrets

Chair

Jason Weber

Scribe

Jatinder Mann



Contents

* Topics

1. Share face to face meeting agenda.

2. Updates on changes to the test harness.

3. Feedback and discussion on what resources should be included in the Resource timing arrays.

4. Feedback and discussion on how to enable cross-domain access to Resource timing information.

5.  Feedback and discussion on Navigation timing.

6. Feedback and discussion on updates to User timing.



* Summary of Action Items

________________________________________

<scribe> Chair: Arvind Jain

<scribe> agenda: http://lists.w3.org/Archives/Public/public-web-perf/2011Mar/0081.html

<scribe> scribe: Jatinder Mann



Topic: F2f agenda

http://lists.w3.org/Archives/Public/public-web-perf/2011Mar/0096.html

Everyone agrees with the F2F agenda as is.



Topic: Updates on changes to the test harness

http://lists.w3.org/Archives/Public/public-web-perf/2011Mar/0065.html



TonyG has updated all test cases and they are passing.



Need a process to ensure when http://w3c-test.org/resources/testharness.js is updated, all dependent test cases are run to catch any regressions in the tests themselves. Need to determine the correct place to raise such concerns. I can follow up and ensure this gets fixed.



Tony: James Graham should be a good starting point.



Topic: Feedback and discussion on what resources should be included in the Resource timing arrays



Link: http://lists.w3.org/Archives/Public/public-web-perf/2011Mar/0080.html



Two things to talk about: 1) what to include in the resource timing array, 2) how to deal with cross-domain access



We agree that we should include just the downloaded resources in the resource timing array. E.g., most UAs today cache static resources after initial download. Script initiated and XHR resource downloads should be included.



We will take an action to try to update the spec to make this clear.



<plh> ACTION: Jatinder to Update RT spec to be clear on the behavior of what gets included in the RT array. [recorded in http://www.w3.org/2011/03/23-webperf-minutes.html#action03]

<trackbot> Created ACTION-9 - Update RT spec to be clear on the behavior of what gets included in the RT array. [on Jatinder Mann - due 2011-03-30].



Everyone agrees that the HTTP response solution is a complete solution. We may still want to consider a hybrid solution that uses both a HTTP response solution and a meta tag.



<scribe> ACTION: Jatinder to Update RT spec to include a HTTP response solution similar to the CORS spec for review by list. [recorded in http://www.w3.org/2011/03/23-webperf-minutes.html#action04]

<trackbot> Created ACTION-10 - Update RT spec to include a HTTP response solution similar to the CORS spec for review by list. [on Jatinder Mann - due 2011-03-30].



Topic: Feedback and discussion on Navigation timing.



<scribe> ACTION: Zhiheng to update Navigation Timing spec to make the sentence "For example, the following Javascript shows the time it take to fully load a page: " clear. [recorded in http://www.w3.org/2011/03/23-webperf-minutes.html#action05]



<trackbot> Created ACTION-11 - Update Navigation Timing spec to make the sentence "For example, the following Javascript shows the time it take to fully load a page: " clear. [on Zhiheng Wang - due 2011-03-30].



Agreement that adding requestEnd is not adding much value - especially considering the large undefined time between requestEnd and responseStart.



We agree that TYPE_RESERVED should remain in the spec as it is a future catch all.



We agree that TYPE_NAVIGATE is currently underspecified. TYPE_NAVIGATE should include form submissions and script initiated navigations.



<scribe> ACTION: Zhiheng to Update the NT spec to include form submissions and script initiated navigations as a TYPE_NAVIGATE. [recorded in http://www.w3.org/2011/03/23-webperf-minutes.html#action06]



<trackbot> Created ACTION-12 - Update the NT spec to include form submissions and script initiated navigations as a TYPE_NAVIGATE. [on Zhiheng Wang - due 2011-03-30].



Jatinder and Nic to respond to Olli on the mailing list.



<scribe> ACTION: Philippe to Follow up and ensure that futures changes to test harness will be communicated to all owners of test cases dependent on the test harness due April 6. [recorded in http://www.w3.org/2011/03/23-webperf-minutes.html#action09]



<trackbot> Created ACTION-13 - Follow up and ensure that futures changes to test harness will be communicated to all owners of test cases dependent on the test harness due April 6. [on Philippe Le Hégaret - due 2011-03-30].



Summary of Action Items



[NEW] ACTION: Jatinder to Update RT spec to be clear on the behavior of what gets included in the RT array. [recorded in http://www.w3.org/2011/03/23-webperf-minutes.html#action03]



[NEW] ACTION: Jatinder to Update RT spec to include a HTTP response solution similar to the CORS spec for review by list. [recorded in http://www.w3.org/2011/03/23-webperf-minutes.html#action04]



[NEW] ACTION: Philippe to Follow up and ensure that futures changes to test harness will be communicated to all owners of test cases dependent on the test harness due April 6. [recorded in http://www.w3.org/2011/03/23-webperf-minutes.html#action09]



[NEW] ACTION: Zhiheng to update Navigation Timing spec to make the sentence "For example, the following Javascript shows the time it take to fully load a page: " clear. [recorded in http://www.w3.org/2011/03/23-webperf-minutes.html#action05]



[NEW] ACTION: Zhiheng to Update the NT spec to include form submissions and script initiated navigations as a TYPE_NAVIGATE. [recorded in http://www.w3.org/2011/03/23-webperf-minutes.html#action06]