[minutes] 20110615 Web Performance WG Teleconference #37

Meeting Summary:



1.      Discussed feedback on the Unified Timing v3

As per mailing list thread, http://lists.w3.org/Archives/Public/public-web-perf/2011Jun/0055.html, there were some concerns with the current proposal. The current proposal suggests an inheritance model, single buffer, and the use of a generic API set. We agreed that the single buffer may be too constraining and potentially not needed for certain Timing types. We also agreed that the inheritance model was a good idea. Microsoft to propose the next version of the unified timing by applying an inheritance model to the current spec. We will discuss this proposal in next week's call.



2.      Discussed feedback on Page Visibility spec

a.       Concerns with Page Visibility spec and IDLE spec

Concerns were raised on the mailing list and the call regarding clarification of behavior of the page visibility spec and potential conflicts with the IDLE spec. In order to ensure issues are closed, we agreed that all future concerns should be opened as issues via the Web Perf issue tracker, http://www.w3.org/2010/webperf/track/issues. This will help the working group track issues and converge on decisions. We will discuss the new issues on the mailing list and calls as they are opened.



b.       Simplifying the Page Visibility spec

It was raised that we should consider simplifying the Page Visibility spec by removing the document.visibilityState and relying only on the boolean, document.hidden, that describes whether the page is visible to the eyes or not. This simplification would remove some concerns with debouncing (as preview states would no longer be needed) and may separate this spec from prerendering. This issue has not yet been closed on and there will be a thread to the mailing list to discuss further.



c.        CSS animation perf statistics

This issue was raised in the last two minutes of the call. There wasn't a strong opinion on the matter, as more research was needed. Follow up on the mailing list to be done.



3.      Discuss Battery Status specification feedback.

We ran out of time and did not get a chance to discuss this spec. It will be covered in the next working group call.



4.      Discuss requestAnimationFrame open issues.

We were hoping to go through the seven open issues on requestAnimationFrame, however, time ran out. These will be reviewed in the next working group call.



We have the following action items from this meeting:

1.       Jatinder Mann: Propose a unified timing v4 with inheritance model applied to current spec.

2.       Kyle Simpson: Open feedback on Page Visibility spec as issues.

3.       Kyle Simpson: Share an IDLE spec with working group.



Detailed Notes:



Web Perf Teleconference #37 6/15/2011



IRC log: http://www.w3.org/2011/06/15-webperf-irc


Meeting Minutes: http://www.w3.org/2011/06/15-webperf-minutes.html



Attendees

Present

Jatinder Mann, JamesS, TonyG, Cameron, Phillipe, Arvind Jain, Kyle Simpson



Regrets

James Robinson, Zhiheng Wang, Anne



Scribe

Jatinder Mann



Contents

Agenda

1.       Discuss Unified Proposal.

2.       Discuss Battery Status specification feedback.

3.       Discuss Page Visibility Last Call.

4.       Discuss requestAnimationFrame open issues


--------------------------------------------------------------------------------
Discuss Unified Proposal.

Jatinder: Thank you James for putting the time into formulating this proposal. As per action items from the last conference call, we have been actively evaluating this proposal. We had responded to the mailing list with our concerns: http://lists.w3.org/Archives/Public/public-web-perf/2011Jun/0055.html
... In general, there are a few main issues. The proposal attempts to create a single interface that all Timings must use for data collection, storage and retrieval. At first glance, the unified interface looked like a very good idea. However, we are concerned that this single interface constrains current and future Timings to be generalized and risks losing, or making it difficult, to have access to functionality specific to that Timing. A widget
... The single, circular buffer is also a large concern, as mentioned in the mailing list.

JamesS: A single buffer isn't required, we can change that.

James: Having an inheritance model will take us part of the way to where I'd like this to be.

Jatinder: I can take an action item to include an inheritance model into our current specification. I will share that proposal on the mailng list.

<plh> [[

<plh> As examples, the attribute returns true when:

<plh> The User Agent is minimized.

<plh> The User Agent is not minimized, but the page is on a background tab.

<plh> The Operating System lock screen is shown.

<plh> The User Agent is minimized and a preview is shown.

<plh> ]]

<plh> (for the hidden attribute)

<plh> and also transparent window

<plh> actually, opera allows one to minimize a tab