Platform/2014-12-16

« previous week | index | next week »

Need To Know

(Release and system issues that may impact engineering this week.)

Notices/Schedule (lsblakk/sylvestre/lmandel)

Next Merge: January 6, 2025		Next Release: January 7, 2025
Trains
Central: 135	Aurora: 54	Beta: 134	Release: 132

• 35.0b4 looking like we'll do a build 2 with backout bug 1097321 due to startup crash issues in b3

• Holiday Beta scheduling
  • Week 3 - Dec 15-19: Beta 4 (Desktop & Mobile) and Beta 5 (Desktop) build & ship as usual
  • Week 4 - Dec 22-26: Beta 6 (Desktop & Mobile) gtb on Monday Dec 22 as usual, ship on Tues Dec 23 - no further betas
  • Week 5 - Dec 29-Jan 2: Beta 8 (Desktop & Mobile) gtb on Monday Dec 29 as usual, ship on Tues Dec 30 - no further betas
  • Week 6 - Jan 5-9: Beta 10 (Mobile) and Desktop RC 35.0 gtb on Monday Jan 5th, ship on Tues Jan 6 to beta channels

Build Changes (gps)

(Build changes of which engineers should be aware.)

RelEng (catlee)

(Repo, test, and other information for engineers from the release engineering team.)

Upcoming Outages/Upgrades

(System outages/upgrades and tree closures that impact engineering.)

• Tree Closing Window Saturday, Dec 20. Tentative time 0900-1730 PT, see tracker for details.

Quality Programs

(An opportunity to hear about status with the various quality programs that do not have a formal team structure.)

OrangeFactor (ryanvm)

• Past week's OrangeFactor: 2.29 (last week: 3.16).
• Many of the top oranges are lacking in action. I'll be pinging in bugs soon, please help if you can.
• 15 intermittent failures marked as fixed in the last week - List - Thanks!.
  • Thanks to Andreas Pehrson from Telenor for fixing a frequent B2G assertion orange.

CritSmash (dbolter)

MemShrink (njn)

<Read Only>

DMD now does cumulative heap profiling. This is great for finding places where excessive numbers of short-lived heap allocations occur.

Stability (kairo)

• Numbers look bad on desktop to some degree due to https://bugzilla.mozilla.org/show_bug.cgi?id=1097321#c26 (75% of all 35.0b3 crashes!)

Team Stand-ups

(In <2 mins, what did your team accomplish last week, on what is your team working on this week, and on what, if anything, is your team blocked? No questions during the stand-ups. All questions should be asked during the roundtable.)

A*Team (jgriffin)

• All of this was announced in dev.platform and dev.b2g but here's your chance to ask questions
• PSA: You can now look your Try pushes to your custom mozharness repository (e.g. a user repo)
  • Test mozharness changes on Try
• Old news: You can easily run mozharness jobs on your machine (more support coming)
  • Run mozharness jobs with --cfg developer_config.py

Accessibility (dbolter)

B2G Services (dougt)

Cloud Services (mmayo)

Desktop Platform (bsmedberg)

Developer Services (gps/lthomson)

Developer Tools (prouget)

DOM (jst/overholt)

<Read Only>

• We're disabling WebSocket on Workers on beta because it's not stable enough.

Electrolysis (e10s) (blassey)

• Please help test e10s in Nightly!
  • Known issues: https://wiki.mozilla.org/Electrolysis#What_to_Expect
• bug 1092630 landed. Should fix a host of problems on OSX related to plugins. Please test!

Firefox Desktop (gavin)

Firefox Mobile (snorp/blassey/mfinkle)

GFX (milan)

JS (naveed)

<Read Only>

• 2015 Layout Roadmap posted:
  • https://wiki.mozilla.org/Platform/Roadmap#JavaScript

Layout (jet/dbaron)

<Read Only>

• 2015 Layout Roadmap posted:
  • https://wiki.mozilla.org/Platform/Roadmap#Layout

Media (mreavy)

<Read Only>

1. What has your team done since last week that might impact others?

• Landed rewrite of the webrtc call negotiation code - regressions are possible, some have been fixed
• Hello Rooms fixes have been uplifting to aurora and beta.

2. What will your team do this week that might impact others?

• Preparing to land webrtc.org import of stable branch 40 (bug 1109428) - same branch Chrome 40 will use.
  • May well break --disable-webrtc builds again ;-) We'll fix it.
• k18e's team is planning to uplift some media rewrites being done by alfredo and blake wu.

3. Is anything slowing your team down or getting in their way?

• GMail/Gcal ;-)

Necko (dougt/jduell)

• HTTP/2 is scheduled to land on FF 35 (we'll advertise draft spec only)
  • we'll advertise final (non-draft) HTTP/2 support in 36
  • opportunistic encryption will be in 37

• B2G emulator still appears to be broken (pipe writes can get lost)
  • we're trying to hack around it for our fix

Performance (vladan)

<Read Only>

No update this week.

Seceng (dougt)

Shumway (tschneidereit)

WebAPI (overholt)

Roundtable

(Comments and questions that arise during the course of the meeting or otherwise do not have a section.)

Links

Friends of the Tree

Mailing List Threads

(Threads that are likely to be of interest to engineering from various mailing lists.)

Good Reads

(Links to blog posts, books, videos, etc. that you think will be of interest to others.)

irc #planning Log From This Meeting

10:52 AM <KaiRo> lmandel: looks like I found the issue behind the gfx issues on beta, see https://bugzilla.mozilla.org/show_bug.cgi?id=1097321#c26
10:53 AM <KaiRo> and lsblakk ^^^
10:55 AM <lmandel> KaiRo: Good sleuthing.
10:56 AM <KaiRo> lmandel: luckily I just stumbled over bugmail of SoftVision verifying that bug that seems to have caused the issue ;-)
10:57 AM <KaiRo> and that got me thinking "hey, we did just land unblocking something that includes Intel stuff? wait a minute..." ;-)
11:00 AM <KaiRo> lsblakk: actually, 1) sheriffs/RyanVM can do a straight backout right away if we are sure enough, that doesn't even need Jeff, and 2) I'm actually reluctant to sign off 35.0b4 with this huge startup crash issue continued
11:00 AM <•lsblakk> ah, didn't see that it was a startup crash anywhere until now
11:00 AM <•lsblakk> in that case, i'm happy to take a backout and do a build 2
11:03 AM <overholt> cpeterson, did milan_ and I get disconnected?
11:03 AM <KaiRo> lsblakk: yes, took me a bit to realize as well, I'm definitely in favor of a build2 there.
11:04 AM <cpeterson> overholt: i don't see you guys on screen
11:04 AM <overholt> cpeterson, thanks. looks like a Vidyo crash. we're back.
11:04 AM <Pike> not just that the calendars are all upset, my head thinks it's wednesday, too
11:07 AM <RyanVM|sheriffduty> KaiRo: lsblakk: backing out now
11:07 AM <KaiRo> RyanVM|sheriffduty: thanks
11:10 AM <RyanVM|sheriffduty> jduell: CC me to that bug when you file it?
11:10 AM — RyanVM|sheriffduty wonders if that's the same underlying issue we have with other emulator test stability issues
11:11 AM <jduell> RyanVM|sheriffduty: will do
11:11 AM <RyanVM|sheriffduty> thanks
11:11 AM <jesup> jduell: sorry, we don't detect it directly: #if defined(MOZ_WIDGET_GONK) && defined(DEBUG) // B2G emulator debug is very, very slow and has problems dealing with realtime audio inputs
11:11 AM <jduell> jesup: ah. Ok, thanks
11:13 AM <jduell> jesup: hmm, this looks interesting... http://mxr.mozilla.org/mozilla-central/source/gfx/thebes/gfxAndroidPlatform.cpp#117