<scribe> scribe: AndersonQuach

clear the agenda

list the agenda

save the agenda

list the agenda

Zhiheng: Spec updated to include items discussed last conf call.

AndersonQuach: I see relaxed origin captured in the spec and HTML5, 5.3.1 same-origin restriction.

JasonSobel: Yes, captured logon auth -> target page scenario well.

AndersonQuach: Yeah, right now navigationStart and redirectStart, redirectEnd is lost for landing page -> autho domain -> target domain.

NicJansma: how will this be decorated, previous page and current page must set a header or script value.

JasonSobel: The group should follow-up with details offline perhaps.

NicJansma: This can be worked around now in JavaScript.

JasonSobel: This is nice to have and not critical.
... Hit small amount of people that can work around the issue.

Zhiheng: Let's follow-up offline.

TonyG: What is the status of CORS? Can we use this for solving this problem?

JasonSobel: Can we follow-up with the author annevk@opera.com with CORS and the status?

Updates on Navigation Timing and Same-Origin

<Zhiheng> sorry, got dropped call, dialing again.

JasonSobel: Let's check in with the timeline, let's understand where they are at before a go/no go with the usage of CORS.

TonyG: Last remaining question, when the timings are exposed.
... Make all the changes, to expose the key points.

Zhiheng: document.domain / relaxed origin, still want to specifiy host component and port number. We need to define clearly what we want.

Christian: why do we care about the port number?

NicJansma: W3C definition of origin is schema, host, and port.

Christian: I suggest same-origin policy to include document.domain.

http://msdn.microsoft.com/en-us/library/ms533028(VS.85).aspx

http://dev.w3.org/html5/spec/origin-0.html#relaxing-the-same-origin-restriction

Christian: At least in Firefox, document.domain is free, function to compare origin already takes it into account.

JasonSobel: same-origin policy captures the options, i. document.domain and ii. use of CORs and the timeline.

next topic

Updates on the timeline for Navigation Timing, From Working Draft to Recommendation

AndersonQuach: Target for 11/30/2010 for Last Call; and 01/07/2010 for Candidate Review

Zhiheng: Ideally we should have a Firefox reference implementation. Before last call.
... Targeting Last Call in Fb

JasonSobel: Don't understand why we have to delay for a 3rd reference implementation. If possible we should move faster.

Zhiheng: Firefox a significant part of the ua community.

Christian: Firefox wants the interface, currently not known who is working on it.
... Could take up to Jan before it can be submitted to the Firefox feature repository. It could take a while before they are ready.

TonyG: Firefox, Jonas, Christian and Sergey have not raised problems with the spec.

AndersonQuach: To make progress we should target last call for end of Nov and target conformance tests and CR for Jan.

Zhiheng: Will follow-up Jonas on the timeline and Navigation Timing spec.

AndersonQuach: I can follow-up with submitting tests and hosting tests for Navigation Timing.

TonyG: I can contribute tests for conformance testing.

AndersonQuach: Great! IE has tests to contribute as well.

Resource Timing and User Timing

Zhiheng: can we access timing by ID or URL?

NicJansma: for URL can iterate over results in the interface.
... Not everything have have an id and complications, change source image associated with id. we should have them individually captured.

Zhiheng: use case for resource id, or url, if there are many iframes and differentiate mine versus third party.

NicJansma: we see the limitations, similar API to get by id or url?

JasonSobel: the way people are going to use this is varied, allowing people to do filtering in script seems reasonable.
... don't mind to parse a blob to figure out what's interesting.
... i like that the developer does this in script.
... batched processing is the primary use.
... from onload indexed for a specific timing, is a nice to have. from facebook, the batched processing is what we already do.

TonyG: Let's get some ideas about the various scenarios and have the whole hour for it.

Summary

1. The working group will discuss in mail the technical details of using document.domain in order to disclose navigationStart with respect to a relaxed same origin policy http://dev.w3.org/html5/spec/origin-0.html#relaxing-the-same-origin-restriction

2. JasonSobel will follow up with the author of CORS annevk@opera.com on the timeline of the working draft spec.

3. Zhiheng to follow-up with Jonas on Navigation Timing timeline and spec feedback.

4. Target Timeline for Navigation Timing, Last Call for 11/30/2010 and work towards Candidate Review for 1/15/2010.

5. Anderson to follow-up offline with conformance test infrastructure to host test cases.

6. The Web Performance working group to follow-up in mail to discuss the technical trade-offs of the Resource Timing proposals.

This is scribe.perl Revision: 1.135  of Date: 2009/03/02 03:52:20  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/JasonSorbel/JasonSobel/
Found Scribe: AndersonQuach
Inferring ScribeNick: AndersonQuach
Default Present: [Microsoft], +1.415.829.aaaa, +1.415.568.aabb, +1.650.214.aacc, +1.650.390.aadd, +1.650.253.aaee, +1.650.390.aaff
Present: [Microsoft] +1.415.829.aaaa +1.415.568.aabb +1.650.214.aacc +1.650.390.aadd +1.650.253.aaee +1.650.390.aaff AndersonQuach NicJansma TonyG Christian ArvindJain JasonSobel Zhiheng
Regrets: JasonWeber
Got date from IRC log name: 03 Nov 2010
Guessing minutes URL: http://www.w3.org/2010/11/03-webperf-minutes.html
People with action items: 

WARNING: Input appears to use implicit continuation lines.
You may need the "-implicitContinuations" option.