W3C home > Mailing lists > Public > public-web-perf@w3.org > January 2018

Minutes (was: Re: [1/18/18] webperf group call @ 11AM PST)

From: Xiaoqian Wu <xiaoqian@w3.org>
Date: Mon, 22 Jan 2018 14:24:34 +0800
To: public-web-perf <public-web-perf@w3.org>
Message-ID: <a19abc5f9b4fd182c1a96c3b24203ae5@webmail.w3.org>
Available at:
https://www.w3.org/2018/01/18-webperf-minutes.html

Also in plain text:
----------------------
WebPerf Group Call
18 Jan 2018
Agenda

Attendees
Present
Philip, igrigorik, Yoav, Ryosuke, Tim, Todd, Nic, @@
Regrets
Chair
Todd
Scribe
igrigorik
Contents
Topics
HR-TIme
Beacon
User Timing
Resource Timing
Navigation Timing
Summary of Action Items
Summary of Resolutions
Todd: focusing on specs that we're trying to get to CR today

HR-TIme
Todd: we have some test failures: https://wpt.fyi/hr-time
... let's try to get Chrome and FF to align

<scribe> ACTION: Tim to take an AI on Chrome side

<scribe> ACTION: someone to to file bugs against Firefox

Todd: Edge 15 is ~year old, there is an open issue to get that updated
... we just asked the spec to transition to PR, we should get Chrome and 
FF to pass on the tests

Ryosuke: I think our failures are all due to us not implementing toJSON 
and timeOrigin, we have bugs open for those
... no objections to either, just haven't got to them yet

Todd: Edge should have implemented all of it, need to test on latest 
versions

Beacon
Todd: no open issues, but test status has red @ https://wpt.fyi/beacon
... need to understand these failures

Yoav: Fetch#keepalive discussions on Blink-dev may have implications for 
limits

Todd: AI here is reach out to various vendors

Ryosuke: on preview build we should be passing all the tests

Todd: edge 15 should also be passing, need to investigate why its 
failing

<scribe> ACTION: Todd to email contacts at each vendor on open issues on 
FF side, will ping Ben, Harald, Marcos on Safari, we believe everything 
should be green

for CR we need two UA's: if we have latest Edge and latest Safari pass, 
we should be good to go

User Timing
igrigorik: main question on PR is whether measure should throw a 
SyntaxError?
... do we need to update WPT tests for this update?

<scribe> ACTION: philipwalton toreview SyntaxError logic for measure()

Tim: it looks like the tests that Chrome fail are related to this; 
Safari and Chrome are not throwing

Edge: IE did not throw either

Ryosuke: I have concerns over compat if we throw

philipwalton: we could resolve not to throw, update the tests

<scribe> ACTION: philipwalton to write up summary on Github, ping FF for 
feedback

Resource Timing
Nic: we're still collecting data for some of the L1 issues

triage 
https://github.com/w3c/resource-timing/issues?q=is%3Aopen+is%3Aissue+no%3Amilestone

buffer issue: L2, exposing requests and processing time: L3

Nic: initiatorType might span L2 and L3

Yoav: let's split L2 to write tests for existing initiatorTypes defined 
in the spec

<scribe> ACTION: Nic to open new issue to track initiatorType test for 
L2

Todd: https://github.com/w3c/resource-timing/issues/70
... this was specced in NT1 but we don't have a test

Yoav: I can take this as an AI
... moving to L2

AI: Nic to sweep L2 issues for low-hanging fruit

Navigation Timing
https://wpt.fyi/navigation-timing — lots of red, we need to review

AI: review NT and RT tests to identify where failures are

Todd: we may need a few test holes, AI: Todd to followup

https://github.com/w3c/navigation-timing/issues/82

Nic: we're advising not to use navigationStart, but what's the right way 
to do it?

Ilya: use timeOrigin

AI: add note to the section pointing to HR time

Ryosuke: merging NT and RT could make sense, they're very intertwined

Todd: I generally agree that it's the right thing to do, but wondering 
about benefits relative to time investment

Yoav: ditto

Todd: https://github.com/w3c/navigation-timing/issues/78 is a long-term 
maintenance question

<scribe> ACTION: Ilya to sketch out involved work to merge the specs, 
we'll review in future call

next call: Feb 1st, 11AM PST

agenda, we'll coordinate on email

Summary of Action Items
[NEW] ACTION: Ilya to sketch out involved work to merge the specs, we'll 
review in future call
[NEW] ACTION: Nic to open new issue to track initiatorType test for L2
[NEW] ACTION: philipwalton to write up summary on Github, ping FF for 
feedback
[NEW] ACTION: philipwalton toreview SyntaxError logic for measure()
[NEW] ACTION: someone to to file bugs against Firefox
[NEW] ACTION: Tim to take an AI on Chrome side
[NEW] ACTION: Todd to email contacts at each vendor on open issues on FF 
side, will ping Ben, Harald, Marcos on Safari, we believe everything 
should be green

Summary of Resolutions
[End of minutes]

On 2018-01-18 11:38, Ilya Grigorik wrote:
> Note that we're changing from our usual time of 10AM to one hour later
> @ 11AM PST.
> 
> Hangout:
> https://hangouts.google.com/hangouts/_/chromium.org/webperf-wg [1]
> 
> Tomorrow's call will be focused on issue triage of existing specs. If
> you have a particular issues you'd like to raise to the group, or bump
> to the top of the agenda, please leave a note in the doc:
> 
> https://docs.google.com/document/d/10dz_7QM5XCNsGeI63R864lF9gFqlqQD37B4q8Q46LMM/edit#heading=h.pc0zv03mzxj0
> [2]
> 
> 
> Links:
> ------
> [1] https://hangouts.google.com/hangouts/_/chromium.org/webperf-wg
> [2]
> https://docs.google.com/document/d/10dz_7QM5XCNsGeI63R864lF9gFqlqQD37B4q8Q46LMM/edit#heading=h.pc0zv03mzxj0
Received on Monday, 22 January 2018 06:24:37 UTC

This archive was generated by hypermail 2.3.1 : Monday, 22 January 2018 06:24:38 UTC