B2G/Meeting/2012-11-13

From MozillaWiki
< B2G‎ | Meeting
Jump to: navigation, search

News

  • Welcome to Milestone 5: Prompt, pretty and popular. Focus is on bug fixes, performance optimizations and visual completeness.
  • Daily Smoketests: Using the "smoketest" keyword. If you own a bug with that keyword, should relocate to a bunker and work on nothing else until it's fixed.

Administrative

Ecosystem

  • <ateam> (mdas, jgriffin, ahal, wlach) - updated 10/30
    • This week:
    • Overview: https://etherpad.mozilla.org/b2gstatus-ateam
      • Have converted update.xml generation on dogfood update server to a web app; will make it much more flexible in dealing with future changes
      • Mochitests are running reliably green on TBPL on cedar (https://tbpl.mozilla.org/?tree=Cedar ); working on moving this over to m-c and other trunk trees. ETA end-of-week.
      • WebAPI tests are running with oranges on TBPL on cedar; working on fixing up some problems; will target moving to m-c next week.
      • Work is progressing on getting reftests running on cedar; will work on getting them into m-c after that, likely next week.
      • Work to get XPCShell tests in TPBL will lag the other unit tests, likely ETA week of Nov 12
      • Work to adapt mochitest-plain so it can handle browserChrome events (bug 798580) was much more complicated than anticipated, but work is progressing and there a couple of patches for review.
      • Working on getting Gaia integration tests running daily on a panda and reporting to autolog; ETA next week. Will look at gaia unit tests after.
      • Assisting WebQA with their automation of smoke tests and configuration of a Jenkins instance for running them on an unagi
      • Working on getting Eideticker running in a "manual" mode on a daily basis on a panda.
    • Last week:
      • Shaking out bugs in the update process with marshall_law and tchung which arose after the switch to aurora.
      • WebAPI tests just got turned on for cedar in buildbot; we're waiting for them to appear on TBPL and then will resolve any problems that show up. I hope to be able to promote them to trunk branches by end of week. Reftests and mochitests to follow.
      • We have telephony tests running again.
      • Work on mochitest-chrome is slower than anticipated, since it has never run remotely (i.e., with Fennec).
      • Need a decision on https://bugzilla.mozilla.org/show_bug.cgi?id=798580; so we can alter the way that mochitests are run on B2G in order to support tests for security checks (https://bugzilla.mozilla.org/show_bug.cgi?id=780955).
      • Investigating ways to get Gaia tests in CI.
  • <QA team> (Geo, Jason, Tony)

BLOCKERS

STATUS

    • Met w/ A-Team+Fabrice during workweek, came up with plan for mochitest framework to optionally run tests at a high privilege level to ease permissions concerns for functionality tests and centralize code for much easier maintenance.
    • Refactored permissions testing approach into three separate suites to reduce dependencies on unstable system parts: https://wiki.mozilla.org/B2G/QA/API_Permissions_Test_Plan -- tracking bug + patch for existing work on Suite 2 coming this week or early next.
    • Starting triage on existing mochitests to expand suite running on TBPL.
    • Fota testing this week, awaiting a good build for stable first
    • Apps
    • Definition of "blockers" for identity has been resolved and cleared up
    • Last week:

BLOCKERS

    • 'smoketest' keyworded now
    • Bug 805967 - [OTA update] much of my user data blown away in today's update (fixed tomorrow's build?)
    • Bluetooth transfer work
    • video recording on gaia work
    • Signing of packaged apps - really concerned about progress here

STATUS

      • reminder for devs that test on device, to pull from mozilla-aurora branch
      • Methodology being used to get around permissions on mochitest (e.g., https://github.com/geoelectric/webapi-smoke-sprint/blob/master/settings/test_launcher.html) broken by recent (valid) platform changes. Need a new way to launch these, will be working to nail a supported best practice this week.
      • Separating out API permissions testing from dependencies on app install, mochitest-chrome, etc, so we can get some substantial testing sooner. Working on a revised project plan.
      • Unblocked on apps testing, can go back to testing hosted and packaged apps
  • <Release Engineering> (Aki) - Updated on 13nov
   This was a BIG week in RelEng. Lots of behind-the-scenes work came to visible completion!
   (DONE) unagi builds on tbpl, with nightlies and symbols https://bugzilla.mozilla.org/show_bug.cgi?id=800364
   (DONE) b2g builds on emulator-with-codecs; https://bugzilla.mozilla.org/show_bug.cgi?id=789652 done. 
   (DONE) b2g emulator tests:
   (DONE) webapi live and green: https://bugzilla.mozilla.org/show_bug.cgi?id=800025
   (DONE) reftests live and green: https://bugzilla.mozilla.org/show_bug.cgi?id=808085
   (DONE) mochitests live and green  https://bugzilla.mozilla.org/show_bug.cgi?id=807771
   aki/rail/armen/jgriffin working through rollout issues: https://bugzilla.mozilla.org/show_bug.cgi?id=807125 only pending xpcshell
   (DONE) b2g desktop linux64 builds https://bugzilla.mozilla.org/show_bug.cgi?id=804330
   (DONE) discovered "deliverable" of hosted git repo; working out requirements w/cjones, ehsan, agal, RelEng and IT
   test B2G-build-on-pandaboards 
   https://bugzilla.mozilla.org/show_bug.cgi?id=778249
   Will be primary build in test automation https://bugzilla.mozilla.org/show_bug.cgi?id=777530
   new kernel done: stayed on 3.0.8.x tzimmerman/jmaher: https://bugzilla.mozilla.org/show_bug.cgi?id=778248.
   physical panda board setup (for both nativefennec or b2g)
   existing 48-60 delayed coming online
   ordered 800; first batch arrived, rest in transit
  • <product team> (Chris Lee)
    • This week:
      • on the way to London
    • Last week:
      • User story audits happening this week.
      • We are meeting with TEF next week to review the overall program (Product, Engineering, UX, Marketing, Apps, BD, etc.)
      • Working with release management to prioritize all dogfood blockers to kick off the program asap
      • Coordinating with OEM partners on software updates
      • Working with the team on prioritizing remaining feature work
  • <PMO team> (Kevin Hu, Dietrich, Michael)
    • This week:
      • Driving open user-stories to zero for 19th
      • Partner meetings
  • <Release Management> (Alex Keybl, Lukas Blakk, Bhavana Bajaj)
    • This week:
    • Dogfooding:
      • FOTA update testing with Test Drivers will begin internally once there is a stable update that we can promote
      • Expansion of test program hinges on success of internal FOTA update for our current 300 testers, NOT blocking on ZTE updates
      • Filed 811347 (access to builds in b2g via LDAP) -- any objections? [lsblakk will follow up on this with b2g-release-drivers]
    • Convergence:
      • Team has been nailing down the external update situation (security or otherwise) and FOTA over the past week
      • Up-to-date schedule: https://etherpad.mozilla.org/b2g-convergence-schedule
      • 72 remaining C1 bugs (for Nov19): http://bit.ly/X5rtBE
      • 89 C2 bugs (for Dec10), but triage into this milestone is not complete: http://bit.ly/X5rzJC
      • If a C1 bug is left unfixed on Nov 20, please be prepared to provide details for an exception including ETA, risk, alternatives, and whether there are any blockers
    • Last Week:
    • Dogfooding:
      • Nailing down the bug filing pipeline for feedback
      • We've asked that users who know how to file actionable bugs help us skip the feedback collation step (taking some pressure off of SUMO/QA)
    • Convergence:
      • Working with the security team to come up with a final update proposal that keeps our users as safe as possible given OEM/Carrier requirements
      • Finalizing a convergence milestone plan
      • Crash symbols -> Socorro sounds like they're close to being done (jgriffin said bug 807005)
      • Focusing on FOTA update testing to take:
        • 802214 – Unagi phones are using an unrepresentative amount of memory, affecting testing
        • 803733 – Remove the hack around otoro orientation issue when capturing still images
  • Developer Documentation (sheppy)
    • This week:
      • not in attendence

User Documentation & Support

(Michelle has a conflict today, just read the wiki)

  • Dogfooding and sending feedback summaries
  • Mobile redeisgn is landing and looking good
  • Action on the in-product links to support from Settings, thanks for the help there
  • Eveything.me will contribute help articles and forum support to SUMO per verbal agreement last week
  • TEF cancelled our meeting this week, so no new SLA progress
  • Starting training documentation outlines and scheduling
  • Based on 'second freeze' features coming in, I'd like to plan on conducting doc reviews during the last weeks of November, so we can incorporate all changes in early Dec.

Status updates

Gaia

APIs/Platform Work

  • Payments+Identity (Ferjm, Badida, Jedp, Caitlyn)
    • 5 more on-device bugs to fix before ready for e2e testing
    • waiting on eta from caitlyn
  • Bluetooth support (ericchou, qDot, ginayeh) - Updated on 10/30
    • More bugfixing and cleanup mostly
    • File sending about to land
    • Still in discussions about testing
    • Platform contacts - qDot (Low Level), gina (HFP (headset) + SCO (audio)), eric (OPP (file transfer, being down with))
  • Automatic updates (Marshall, Fabrice, clee) - updated on 11/13
   Landed update interval change in settings for gaia + gecko
   Landed update URL support for %PRODUCT_MODEL% (i.e. "unagi1")
   Landed initial update tools for gecko-wrapped FOTA updates. Incremental gecko OTA update support currently in pull request (need to follow up on review)
   Setup a "fake" FOTA w/ jgriffin, verified w/ tchung. Will push out the first "real" FOTA to fix camera rotation when the next build is blessed as stable
   Patch in progress for "nice"ing the updater process (used in both OTA and FOTA updates)
   POC / demo for Peter La's new FOTA installing animation during the work week
   Tons of meetings w/ OEMs / internal release drivers to figure out our overall update plan
  • Camera (Mike Habicher) - Updated on 10/30
    • This Week:
      • bug 799586: camera app crashes starting video mode after taking a picture
        • looks like a race condition: camera driver is still servicing takePicture() when the mode switch comes in
      • bug 800203: major graphical glitch after taking a picture on unagi
      • bug 801693: need to plumb in support for reporting recorder status/error updates
      • bug 806582 - LANDED: photos taken with camera have incorrect orientation data
        • fix landed on unagi, still needs PR acceptance on otoro
      • bug 804741 - LANDED: need to call close() after passing video file handle to setOutputFile()
      • bug 800627 - LANDED: video recording not working on unagi
        • unagi was missing libOmxVidEnc.so and proper media_profiles.xml
      • bug 806055 - PATCH READY, BUT: camera saves huge (384x512) thumbnails
        • tried to make camera default to smallest non-0x0 thumbnail size, but
        • the camera driver is ignoring the settings
        • I can shrink the thumbnail sizes by reducing the quality (1% = 4.5KiB at 384x512) but the thumbnails look like something out of an Atari 2600
      • bug 807058 - PATCH READY: expose CameraCapabilities.thumbnailSizes and CameraControl.thumbnailSize
    • Last Week:
      • bug 795379 - LANDED: expose recorder profiles to JS (currently only records at 352x288)
      • bug 804359: need to expose runtime-detected recorder profiles without relying on constants defined in media/MediaProfiles.h
        • vendor implementations can and have diverged from the AOSP headers we build against
        • conflicting profiles are currently disabled
        • lower priority, since v1 doesn't need to record at high resolutions
      • bug 804802: recording in 800x480 mode can cause the camera to hang
        • if an 800x480 recording is started on a fresh camera app start, recorder works fine
        • if an 800x480 recording is starting after any other recording, the DSP fails to allocate memory and the recorder and camera hang
        • high-res recording not required for v1
      • bug 795090: limit maximum size of recorded file (else video recorder will fill your uSD card)
        • blocked on bug 801693
        • this can probably be done more intelligently in the JS layer
    • In The Queue:
      • bug 795332: "record-size" option may be deprecated in favour of "video-size"
        • not a big issue for Otoro, since the camera hw only supports one video stream, and its properties come from "preview-size"
      • bug 789067: clean up extraneous abstract ICameraControl class
        • patch posted to bug, still waiting on (increasingly obselete :) review
      • bug 786756: exposing more camera frobs
  • RIL: Cell broadcast (Vicamo) - ???
    • Bug 778093 - Cell Broadcast
      • Have to fix Gaia permission, and cooperate with WIP patch in bug 802121. (jaoo)
      • Update Gaia fixed previous permission problem.

Roundtable