[minutes] Web Performance WG Teleconference #100 2013-03-06

Meeting Summary:

1.     High Resolution Time support in Web Workers

The working group discussed extending performance.now() support in Web Workers. The origin of the now() method in a worker will be associated with the start of navigation of the document that the worker is attached to. For shared workers, it would be the time of creation of the worker. For the Timing specs, it seems like workers attached to a document will report the timing of the document. Share workers should not have access to the Timing data.

I will update the Timing specs and create a High Resolution Time Level 2 spec.

2.     Beacon API

Alois and I have shared our initial draft of the Beacon API, https://dvcs.w3.org/hg/webperf/raw-file/tip/specs/Beacon/Overview.html. We'll reach out to get more feedback on it.

3.     Diagnostics and Errors

The working group has started to look into diagnostics and errors. Alois and I will share a list of topics for discussions. In the meanwhile, I am working on a draft for the Network Diagnostics/Error data spec which I will share with the working group shortly.

4.     Test Cases

Ganesh and Pan will follow up on the Resource Timing tests that are currently failing on IE.

5.     Charter

The AC reps have begun their review of the new working group charter. We are expected to be able to publish new specification as first public working drafts in about a month and a half.

6.     Agenda for the next conference call

The working group plans to discuss the prerender feature in our next conference call.

W3C Web Performance WG Teleconference #100 2013-03-06

IRC log: http://www.w3.org/2013/03/06-webperf-irc

Meeting Minutes: http://www.w3.org/2013/03/06-webperf-minutes.html


Jatinder Mann, James Simonsen, Philippe Le Hegaret, Ganesh Rao, Rob Dickenson


Jatinder Mann


1.     Discuss Test Cases issues

2.     Feedback on Existing Specifications

3.     Feedback on New Specifications


performance.now() in Web Workers

Jatinder: I thought specifying that workers support an interface is as easy as adding "WorkerGlobalScope implements Performance;" to the spec. But, if we want to exclude certain Timing interfaces from workers, we would need to do so in those spec. We may want to say that the Timing APIs will return the timing of the document context that the worker is attached to and shared workers should not support those APIs or just return an error.

James: That sounds reasonable. For shared workers, we could just use the start of creation of the shared worker at time 0.

Jatinder: I'll follow up with a proposal for this.

Philippe: Looks like we'll probably be doing a High Resolution L2.
Beacon API

Jatinder: I sent out a draft of the Beacon API based on the feedback that was given last week. Hopefully, we will get more broader feedback from the working group now that the email has been sent out.
Diagnostics and Errors

Jatinder: One of the topics that was discussed at the workshop was Network Diagnostics data. This seems like an analogous error spec to Resource Timing. I'm working on a draft that I can share with the working group soon.
Test Cases

Jatinder: The Resource Timing test cases at http://w3c-test.org/webperf/tests/submission/Intel/resource-timing/ are still failing to run for me on IE.
... There was a PHP file that Pan said he wanted to upload, I wonder if that will fix the issue.

Ganesh: We will follow up on seeing if we can get that fixed. I'll send you a mail to see which version of IE we should look at.

Philippe: AC reps are now reviewing the charter. We should be able to publish our new specs as first public working draft in about a month and half.

Received on Thursday, 7 March 2013 23:03:22 UTC