Web Performance Working Group -- 12 Jul 2017

<igrigorik> https://github.com/w3c/web-platform-tests/pull/6428

toJSON

Ilya: we landed one of the tests
... Tobie thought it would cover what we need

Todd: what Tobie is doing is awesome but he doesn't check the serialized value/string
... in the ideal world, we'd have one

Ilya: your point is that it doesn't guarantee that the number is the right one?

Todd: correct. it's not perfect.
... we used to have a lot of bugs going through if we didn't check those values
... for each member, checking that the value is properly serialized
... the WPT aren't exhaustive though
... so it's a small gapp
... re boundaries: if you get a string that is longer that what can be carried
... one way to implement is to have a constructor for the webidl

plh: I think we have that already. just need to check that object<->JSON<->object stays consistent

Todd: I'll look to this if we can come up with something

Ilya: we should follow up with Tobie on the thread

Tobie: will do

Ilya: we'll also have someone on our side looking at this this quarter
... we landed the resource timing
... we should update all of the other specs
... for NT, the serializer has been deprecated but we have a legacy section
... how should we treat that?
... we never required toJSON in the past

Plh: remove it or upgrade it?

Tobie: window.performance.navigation.toJSON exists

Ilya: ok, in that case we should update it to toJSON
... NT, HRT, PT, LT
... ST already landed

Shubhie: paint timing?

Ilya: yes
... I'll follow up
... this was a painful because it broke our specs
... I'd like to find a way not to make that happens

<igrigorik> https://github.com/w3c/web-platform-tests/labels/wg-webperf

WPT #6241

<igrigorik> https://github.com/w3c/web-platform-tests/pull/6241

Todd: I need to finish that one
... the dialog prompt is the issue here
... I'll fiddle with the code

WPT #4976

Ilya: Xiaoqian, can you help to land that one?

xiaoqian: sure

WPT #4660

Shubhie: we're landing that one

Ilya: so, in Chrome, we do the flush before disconnect, right?

Shubhie: yes

Ilya: do we have a bug opened in blink?

Shubhie: believe so. we reopened an old bug.

Todd: as long as you guys agree, we'll follow

Ilya: ok to merge then

WPT #3220

Ilya: plh, can you lookin into this one>

plh: sure

WPT #2919

Yoav: I'll look into this one
... I'll open a new one and close #2919

<igrigorik> https://github.com/w3c/web-platform-tests/issues/552

WPT #552

Ilya: I'll ping Simon on this one

Beacon testing

Todd: I'll check the new PR before it gets submitted

Ilya: does it include for the new behavior, 64Kb

Todd: yes

<igrigorik> https://github.com/w3c/beacon/issues

Ilya: what about CORS and ArrayBuffer?

Todd: ArrayBuffer is covered but not sure if we test the content type
... I believe we have some CORS tests. I'll check.

Ilya: ok, once we land those, we can move forward

GitHub/IRC integration

https://github.com/dbaron/wgmeeting-github-ircbot/blob/master/src/config.json

example: https://github.com/w3c/csswg-drafts/issues/1537

Plh: I'll send a PR to dbaron config

Shipping specs

Todd: did you look at our states?
... would be nice to review where we are

Shubhie: would be nice to have a reviewer for the paint timing spec

Yoav: I'll look at it

Shubhie: spec text and that sort of stuff

Nic: I'll look at it too

Ilya: the README is the explainer.

Shubhie: the .bs file is the source

https://github.com/w3c/longtasks/

https://github.com/w3c/longtasks/blob/master/deploy.sh

Plh: I'll setup it for you

<yoav> https://github.com/w3c/preload/pull/107

Preload

Yoav: we can't use single quotes in HTTP headers
... people copied/pasted the example and reported the bug
... so did a very simple pull request to fix this

Ilya: will look
... sounds straightforward

<xiaoqian> https://github.com/w3c/strategy/issues/93