1243881 - Consider unshipping translateTime

Status: RESOLVED FIXED

(Core :: DOM: Core & HTML, defect)

Description

[Boris Zbarsky [:bzbarsky]]

[Tracking Requested - why for this release]: Working group can't make up its mind, and we may not want to ship an API they've decided shouldn't exist after all.

See https://github.com/w3c/hr-time/issues/22#issuecomment-176404544 -- the webperf working group can't make up its mind, again.

I think we have a few options:

1)  Back out bug 1169068 and the changes to not use the parent's timebase for
    dedicated workers until this working group gets its act together and decides
    something else.  Note that we're talking about changes on beta at this
    point.
2)  Just go ahead and ship, and then either remove later or just leave in or
    something.

I think option 1 is the best if we don't think the risk is too high for beta...  Andrea, thoughts?

Also, I think we should stop implementing stuff this working group suggests until it's in some equivalent of CR, because they have demonstrated several times now that they don't really care about past consensus and will just randomly change stuff.

Comment 1

[Kyle Huey (Exited; not receiving bugmail, old account, do not use)]

It's early in beta, so I think #1 is the way forward.

I would definitely support a strongly worded letter (TM) to this working group that we consider their specs too unstable to implement.

Comment 2

[Jonas Sicking (:sicking) No longer reading bugmail consistently]

I've only followed the discussion in [1]. That discussion is certainly indicating that we should unship by doing #1. I'm not sure if there are other discussions elsewhere though.

[1] https://github.com/w3c/hr-time/issues/22

Comment 3

[Boris Zbarsky [:bzbarsky]]

Andrea, let me know if you don't have time for this and I'll pick it up, I guess.

Comment 4

[Andrea Marchesini [:baku]]

Attachment: time.patch

Comment 5

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 8713543 [details] [diff] [review]
time.patch

No, this isn't enough.  We also need to change the timebase to the parent one for dedicated workers, because otherwise people won't be able to map times...

Comment 6

[Andrea Marchesini [:baku]]

Attachment: time2.patch

Consider this second patch too.

Comment 7

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 8714260 [details] [diff] [review]
time2.patch

r=me

Comment 8

[Boris Zbarsky [:bzbarsky]]

Comment on attachment 8713543 [details] [diff] [review]
time.patch

r=me given the other patch

Comment 9

[Pulsebot]

https://hg.mozilla.org/integration/mozilla-inbound/rev/52792f2aac93
https://hg.mozilla.org/integration/mozilla-inbound/rev/b9dba72f9e97

Comment 10

[Boris Zbarsky [:bzbarsky]]

Note that once this sticks we'll need to uplift to Aurora and Beta...

Comment 11

[Andrea Marchesini [:baku]]

Comment on attachment 8713543 [details] [diff] [review]
time.patch

Approval Request Comment
[Feature/regressing bug #]: Implement the latest time performance spec
[User impact if declined]: the working group is still changing the spec and we don't want to ship something that is not stable enough from a spec point of view.
[Describe test coverage new/current, TreeHerder]: mochitests
[Risks and why]: none
[String/UUID change made/needed]: none

Comment 12

[Andrea Marchesini [:baku]]

Comment on attachment 8714260 [details] [diff] [review]
time2.patch

See previous comment.

Comment 13

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/mozilla-central/rev/52792f2aac93
https://hg.mozilla.org/mozilla-central/rev/b9dba72f9e97

Comment 14

[Sylvestre Ledru [:Sylvestre]]

Comment on attachment 8713543 [details] [diff] [review]
time.patch

Make perfect sense, taking it. Should be in 45 beta 3.

Comment 15

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/releases/mozilla-aurora/rev/8b2e83c82646
https://hg.mozilla.org/releases/mozilla-aurora/rev/b6b30f189b79

Comment 16

[Carsten Book [:Tomcat]]

https://hg.mozilla.org/releases/mozilla-beta/rev/958dada10473
https://hg.mozilla.org/releases/mozilla-beta/rev/88b57fb122ab

Comment 17

[Kohei Yoshino]

Updated the site compat doc: https://www.fxsitecompat.com/en-CA/docs/2015/performance-now-in-workers-is-now-based-on-the-workers-creation-time/

Comment 18

[Jean-Yves Perrier [:teoli]]

I removed the mention that was in Firefox 45 for devs too.