Platform/2011-08-30: Difference between revisions

From MozillaWiki
Jump to navigation Jump to search
 
(21 intermediate revisions by 10 users not shown)
Line 7: Line 7:


= Notices / Schedule =
= Notices / Schedule =
* We are chemspilling for a [http://blog.mozilla.com/security/2011/08/29/fraudulent-google-com-certificate/ security issue with the DigiNotar CA].
* This means new versions of Firefox for desktop (3.6.21, 6.0.1, 7, 8, and 9) and mobile (6.0.1, 7, 8, and 9), Thunderbird (3.1.13, and 6.0.1) and SeaMonkey (2.3.2)
** The impact on the old 3.6.21, now 3.6.22 schedule are TBD


= Firefox Development =
= Firefox Development =
Line 14: Line 17:


= Firefox Developer Tools =
= Firefox Developer Tools =
* Initial style inspector patch landed on fx-team. {{bug|582596}}, improvements incoming.
* Console API storage patches landed.


= Performance =
= Performance =
* Posted a thread to dev.platform on getting rid of dev.performance
* Collection of general bugs related to [https://bugzilla.mozilla.org/buglist.cgi?type1-0-0=substring&field1-0-3=rep_platform&field0-0-0=keywords&field1-0-2=op_sys&value1-0-3=ARM&type1-0-1=equals&value1-0-1=Fennec&resolution=---&type1-0-2=equals&type1-0-3=equals&value1-0-2=Android&query_format=advanced&value1-0-0=mobile&type0-0-0=substring&value0-0-0=perf&field1-0-0=keywords&field1-0-1=product Mobile Performance]
* Collection of bugs related to [https://bugzilla.mozilla.org/buglist.cgi?quicksearch=sw:mobilestartupshrink Mobile Startup]


= GFX =
= GFX =
* We think that OpenGL ES accelerated layer composition on '''Nightly''' builds for Android is ready to be tested by technical users.
** There are still some known minor rendering glitches when pinch-zooming or panning; single frames are drawn at the wrong scale.
** Apparently, Adreno-based handhelds (like the Nexus One) freeze when using GL-accelerated layers; maybe don't try it on those phones.
** To turn it on, set <code>layers.acceleration.disabled</code> to <code>false</code> and <code>layers.acceleration.force-enabled</code> to <code>true</code> in <code>about:config</code>, and restart your browser.


= JS =
= JS =
* Type Inference has landed!
** This is Brian Hackett's work + some help from contributor Jan de Mooij
** 30% improvement on V8-v6 benchmark
** Even more on Kraken, fast on lots of other stuff too
* Formed the IonMonkey project team, already emailed about this


= Layout =
= Layout =
Line 55: Line 74:


= WebAPI =
= WebAPI =
* Progress on [[WebAPI/WebSMS|WebSMS]] (can now delete stored messages). Runs on Android and B2G
* [[WebAPI/WebUSB|WebUSB]] is coming along on windows
* Bent started working on [[WebAPI/WebTelephony|WebTelephony]]. Will only run on B2G due to API limitations in Android


= Network =
= Network =
Line 61: Line 84:


= Mobile =
= Mobile =
* Pre-honeycomb flash support (click to play) patches up.  Bug 630007.
* Pre-honeycomb flash support (click to play) patches up.  {{Bug|630007}}.
* More tablet UI has landed
** Transparency seems to cause a scrolling perf hit
* Webkit Honeycomb source are now available.
* Webkit Honeycomb source are now available.


Line 76: Line 101:


* Desktop
* Desktop
** '''#1 Windows crash''': [https://bugzilla.mozilla.org/show_bug.cgi?id=682313 Bug 682313]
** '''#1 Windows crash''': (TRUNK) [https://bugzilla.mozilla.org/show_bug.cgi?id=682313 Bug 682313], (ALL VERSIONS) [https://bugzilla.mozilla.org/show_bug.cgi?id=672144 Bug 672144]
** '''#1 Mac crash''': [https://bugzilla.mozilla.org/show_bug.cgi?id=670842 Bug 670842]
** '''#1 Mac crash''': [https://bugzilla.mozilla.org/show_bug.cgi?id=670842 Bug 670842]
* Mobile
* Mobile
** '''#1 Android crash''': https://bugzilla.mozilla.org/show_bug.cgi?id=663494 Bug 663494]
** '''#1 Android crash''': [https://bugzilla.mozilla.org/show_bug.cgi?id=663494 Bug 663494]


===Socorro===
===Socorro===
* ADUs Friday - Migration to Data Center resulted in missing some ADUs (note the wacky numbers for Friday). The issue has corrected itself.  
* ADUs Friday - Migration of blocklist (AMO) to new Data Center resulted in missing some ADUs (note the wacky numbers for Friday). The issue has corrected itself.  
* Duplicate Reports - Algorithm for dupes in Socorro probably needs follow up.
* Duplicate Reports - Algorithm for dupes in Socorro probably needs follow up.
* Missing Bug IDs in Socorro: [https://bugzilla.mozilla.org/show_bug.cgi?id=681682 Bug 681682]
* Missing Bug IDs in Socorro: [https://bugzilla.mozilla.org/show_bug.cgi?id=681682 Bug 681682]
Line 100: Line 125:


====Other====
====Other====
* Mac 10.7 specific crash - [https://bugzilla.mozilla.org/show_bug.cgi?id=670842 Bug 670842]
* Flash stats
* Flash stats
** [https://crash-analysis.mozilla.com/rkaiser/firefox.4plus.flashsummary.html Summary of daily % of Flash hangs and crashes since 03/01/11]
** [https://crash-analysis.mozilla.com/rkaiser/firefox.4plus.flashsummary.html Summary of daily % of Flash hangs and crashes since 03/01/11]
Line 107: Line 133:
** https://crash-analysis.mozilla.com/rkaiser/2011-08-29/2011-08-29.firefox.9.0a1.flashhangdetails.html
** https://crash-analysis.mozilla.com/rkaiser/2011-08-29/2011-08-29.firefox.9.0a1.flashhangdetails.html
* [https://bugzilla.mozilla.org/show_bug.cgi?id=665775 Bug 665775] and [https://bugzilla.mozilla.org/show_bug.cgi?id=622140 Bug 622140] - blocklist issue that has been lingering since about June. It accounts for a significant volume of crashes so we are working on getting this resolved.
* [https://bugzilla.mozilla.org/show_bug.cgi?id=665775 Bug 665775] and [https://bugzilla.mozilla.org/show_bug.cgi?id=622140 Bug 622140] - blocklist issue that has been lingering since about June. It accounts for a significant volume of crashes so we are working on getting this resolved.
* Working on a couple of session proposals for the all-hands related to crashes.


==Mobile==
==Mobile==
Line 121: Line 148:


= Roundtable =
= Roundtable =
* Guest developers for test days (joe)
** Joe Drew did two videos as part of being a "[http://quality.mozilla.org/qmo-news/2011/08/25/guest-developer-for-testday-tomorrow/ guest developer]" on last Friday's test day:
*** [http://vimeo.com/28213202 Introduction to triaging and verifying an assertion bug on Windows]
*** [http://vimeo.com/28214393 An addendum to that for OS X and Linux]
** Others who would like to be test day guests are wanted! The idea is to explain things about your workflow, how to do something that might not be obvious, or any other thing you might want to show to our triagers and testers.
** Contact [mailto:mevans@mozilla.com Matt Evans] or [mailto:ahughes@mozilla.com Anthony Hughes] about participating.
* Blizzard contacting team leads for priorities and feature pages.

Latest revision as of 18:34, 6 September 2011


« previous week | index | next week »

Notices / Schedule

  • We are chemspilling for a security issue with the DigiNotar CA.
  • This means new versions of Firefox for desktop (3.6.21, 6.0.1, 7, 8, and 9) and mobile (6.0.1, 7, 8, and 9), Thunderbird (3.1.13, and 6.0.1) and SeaMonkey (2.3.2)
    • The impact on the old 3.6.21, now 3.6.22 schedule are TBD

Firefox Development

Firefox Developer Tools

  • Initial style inspector patch landed on fx-team. bug 582596, improvements incoming.
  • Console API storage patches landed.

Performance

  • Posted a thread to dev.platform on getting rid of dev.performance
  • Collection of general bugs related to Mobile Performance
  • Collection of bugs related to Mobile Startup

GFX

  • We think that OpenGL ES accelerated layer composition on Nightly builds for Android is ready to be tested by technical users.
    • There are still some known minor rendering glitches when pinch-zooming or panning; single frames are drawn at the wrong scale.
    • Apparently, Adreno-based handhelds (like the Nexus One) freeze when using GL-accelerated layers; maybe don't try it on those phones.
    • To turn it on, set layers.acceleration.disabled to false and layers.acceleration.force-enabled to true in about:config, and restart your browser.

JS

  • Type Inference has landed!
    • This is Brian Hackett's work + some help from contributor Jan de Mooij
    • 30% improvement on V8-v6 benchmark
    • Even more on Kraken, fast on lots of other stuff too
  • Formed the IonMonkey project team, already emailed about this

Layout

  • More 3D transforms stuff landed, some more to come before we can try turning it on by default (mattwoodrow)
    • Discovered an interesting regression on Facebook where they feed -webkit properties to any browser that Modernizr says supports 3D transforms :-(
  • Jonathan Kew persuaded some authors of TeX hyphenation dictionaries to relicense their dictionaries to be a bit more liberal, so we will ship dictionaries for some more languages
  • Team status:
    • dbaron: review hell, investigating mobile text resizing
    • dholbert: CSS3 flexbox
    • matspal, tnikkel: bug fixing
    • fantasai: getting started on page-break-*
    • jfkthame: bug fixing, patches for Graphite font support
    • jwatt: converting SVG code to use display lists and be more like other layout code
    • heycam: WebIDL, reworking SVG text to use CSS layout
    • smontagu: implementing and unicode-bidi:isolate

Video

  • Team status:
    • doublec: bug fixing
    • kinetik: tidying up libcubeb sound library for review
    • karl: reviews, tuning the lighting in the Auckland office
    • cpearce: fixing remaining issues with fullscreen API (such as multiprocess support)
    • rgiles: starting on WebVTT subtitles
    • derf, jmvalin: Opus work, WebRTC standards discussion, WebRTC demo hacking
    • rjesup: WebRTC
    • roc: MediaStreams processing infrastructure

DOM

WebAPI

  • Progress on WebSMS (can now delete stored messages). Runs on Android and B2G
  • WebUSB is coming along on windows
  • Bent started working on WebTelephony. Will only run on B2G due to API limitations in Android

Network

Plugins

Mobile

  • Pre-honeycomb flash support (click to play) patches up. bug 630007.
  • More tablet UI has landed
    • Transparency seems to cause a scrolling perf hit
  • Webkit Honeycomb source are now available.

Accessibility

Multi-Process

Tree Management

Security

Stability Report

Socorro

  • ADUs Friday - Migration of blocklist (AMO) to new Data Center resulted in missing some ADUs (note the wacky numbers for Friday). The issue has corrected itself.
  • Duplicate Reports - Algorithm for dupes in Socorro probably needs follow up.
  • Missing Bug IDs in Socorro: Bug 681682

Desktop

Top Issues

Channels

  • Trunk - 9.0a1
    • Bug 682313 - Firefox 9.0a1 Crash [@ je_free | mozilla::ipc::windows::DeferredSettingChangeMessage::~DeferredSettingChangeMessage() ]
    • Bug 682916 - Firefox Crash [@ XPC_WN_NoHelper_Finalize ]
    • Bug 682974 - Resolved as duplicate of an earlier bug
  • Aurora
    • [@ RenderTargetStateHolder::PushToDevice(ID3D10Device*, CD3DSurface*) ] - appears as new signature but is the same user. Will file if volume increases.
  • Beta
    • Bug 680927 - Startup crash in mozcomp.dll with IBM Tivoli Access Manager
    • Bug 680921 - malware related

Other

Mobile

  • Naoki has a detailed report for Mobile - https://wiki.mozilla.org/QA/Fennec/crash_rep_29Aug2011.
  • Big issue - really not enough ADUs on Nightly and Aurora.
  • A few issues we need to get addressed ASAP
    • Still missing symbols for many of the mobile crashes - over 40% for 8a2 over 30% for 7 over 45% for release 6 - Bug 672287
    • Bug 679065 - empty report list returned for top crashes.
  • Crash Highlights
    • Fennec 9a1: at least 81.2 % of crashers are new (out of 186 crashes); Fennec 8 has a lot of new crashers as well.
    • Top crash in 6 & 7 - Bug 663494
    • Special crashkill meeting today at noon to talk about how to make these crashes more visible - isolate the high volume reproducible ones and get them in front of devs.

Roundtable

  • Blizzard contacting team leads for priorities and feature pages.
Retrieved from "https://wiki.mozilla.org/index.php?title=Platform/2011-08-30&oldid=347112"